ETSI MANO

Add to my manuals
184 Pages

advertisement

ETSI MANO | Manualzz
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
GROUP SPECIFICATION
Network Functions Virtualisation (NFV);
Management and Orchestration
Disclaimer
This document has been produced and approved by the Network Functions Virtualisation (NFV) ETSI Industry Specification
Group (ISG) and represents the views of those members who participated in this ISG.
It does not necessarily represent the views of the entire ETSI membership.
2
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Reference
DGS/NFV-MAN001
Keywords
configuration, management, network, NFV,
orchestration
ETSI
650 Route des Lucioles
F-06921 Sophia Antipolis Cedex - FRANCE
Tel.: +33 4 92 94 42 00 Fax: +33 4 93 65 47 16
Siret N° 348 623 562 00017 - NAF 742 C
Association à but non lucratif enregistrée à la
Sous-Préfecture de Grasse (06) N° 7803/88
Important notice
The present document can be downloaded from:
http://www.etsi.org
The present document may be made available in electronic versions and/or in print. The content of any electronic and/or
print versions of the present document shall not be modified without the prior written authorization of ETSI. In case of any
existing or perceived difference in contents between such versions and/or in print, the only prevailing document is the
print of the Portable Document Format (PDF) version kept on a specific network drive within ETSI Secretariat.
Users of the present document should be aware that the document may be subject to revision or change of status.
Information on the current status of this and other ETSI documents is available at
http://portal.etsi.org/tb/status/status.asp
If you find errors in the present document, please send your comment to one of the following services:
http://portal.etsi.org/chaircor/ETSI_support.asp
Copyright Notification
No part may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying
and microfilm except as authorized by written permission of ETSI.
The content of the PDF version shall not be modified without the written authorization of ETSI.
The copyright and the foregoing restriction extend to reproduction in all media.
© European Telecommunications Standards Institute 2014.
All rights reserved.
TM
TM
TM
DECT , PLUGTESTS , UMTS and the ETSI logo are Trade Marks of ETSI registered for the benefit of its Members.
TM
3GPP and LTE™ are Trade Marks of ETSI registered for the benefit of its Members and
of the 3GPP Organizational Partners.
GSM® and the GSM logo are Trade Marks registered and owned by the GSM Association.
ETSI
3
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Contents
Intellectual Property Rights ................................................................................................................................9
Foreword.............................................................................................................................................................9
Modal verbs terminology....................................................................................................................................9
1
Scope ......................................................................................................................................................10
2
References ..............................................................................................................................................10
2.1
2.2
3
3.1
3.2
4
4.1
4.2
4.3
4.4
4.5
4.5.1
4.5.2
4.5.3
4.6
4.6.1
4.6.2
4.6.3
4.7
5
5.1
5.2
5.3
5.4
5.4.1
5.4.2
5.4.3
5.4.4
5.4.5
5.4.6
5.4.7
5.5
5.5.1
5.5.2
5.5.3
5.6
5.6.1
5.6.2
5.6.3
5.6.4
5.7
5.7.1
5.7.2
5.7.3
5.7.4
5.7.5
5.7.6
5.7.7
Normative references ....................................................................................................................................... 10
Informative references ...................................................................................................................................... 10
Definitions and abbreviations .................................................................................................................11
Definitions ........................................................................................................................................................ 11
Abbreviations ................................................................................................................................................... 11
NFV Management and Orchestration: objectives and concepts .............................................................13
Overview .......................................................................................................................................................... 13
Management and Orchestration aspects of Network Functions Virtualisation Infrastructure .......................... 14
Management and Orchestration aspects of Virtualised Network Functions ..................................................... 15
Management and Orchestration aspects of Network Services .......................................................................... 16
Other management and orchestration aspects of NFV framework ................................................................... 16
Fault and performance management ........................................................................................................... 16
Policy Management .................................................................................................................................... 17
Testing aspects of Network Services .......................................................................................................... 17
Relation of NFV management and orchestration with existing operations and management systems ............. 17
Overview .................................................................................................................................................... 17
NFV-MANO Interworking with OSS/BSS to deliver NFV business benefits............................................ 18
Key Challenges and Considerations ........................................................................................................... 18
Administrative Domains................................................................................................................................... 19
Management and Orchestration architectural framework ......................................................................20
Overview .......................................................................................................................................................... 20
Principles of NFV-MANO ............................................................................................................................... 20
NFV-MANO architectural framework overview ............................................................................................. 21
NFV-MANO architectural framework functional blocks ................................................................................. 24
NFV Orchestrator (NFVO) ......................................................................................................................... 24
VNF manager (VNFM) .............................................................................................................................. 25
Virtualised infrastructure manager (VIM) .................................................................................................. 26
NS Catalogue .............................................................................................................................................. 27
VNF Catalogue ........................................................................................................................................... 27
NFV Instances repository ........................................................................................................................... 27
NFVI Resources repository ........................................................................................................................ 27
Other functional blocks .................................................................................................................................... 27
Element management (EM) ........................................................................................................................ 28
Operations Support System/Business Support System (OSS/BSS) ............................................................ 28
Network Functions Virtualisation Infrastructure (NFVI) ........................................................................... 28
Network Controllers ......................................................................................................................................... 28
Overview .................................................................................................................................................... 28
NFVO, VIM and Network Controllers ....................................................................................................... 29
Roles and responsibilities ........................................................................................................................... 31
NFVI network EMS/NMS vs. Network Controllers ................................................................................... 32
NFV-MANO reference points .......................................................................................................................... 32
Os-Ma-nfvo................................................................................................................................................. 32
Ve-Vnfm-em ............................................................................................................................................... 33
Ve-Vnfm-vnf .............................................................................................................................................. 33
Nf-Vi ........................................................................................................................................................... 34
Or-Vnfm ..................................................................................................................................................... 34
Or-Vi ........................................................................................................................................................... 34
Vi-Vnfm...................................................................................................................................................... 35
ETSI
4
5.8
5.8.1
5.8.2
6
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Interfaces description approach ........................................................................................................................ 35
Overview .................................................................................................................................................... 35
Producer-consumer paradigm ..................................................................................................................... 36
NFV management and orchestration information elements ...................................................................37
6.1
Introduction ...................................................................................................................................................... 37
6.2
Network Service information elements ............................................................................................................ 40
6.2.1
Network Service Descriptor (nsd) ............................................................................................................. 41
6.2.1.1
nsd base element .................................................................................................................................. 41
6.2.1.2
Connection Point (nsd:connection_point)................................................................................ 42
6.2.1.3
Service deployment flavour (nsd:service_deployment_flavour) ....................................... 42
6.2.1.3.1
Base element.................................................................................................................................... 42
6.2.1.3.2
Constituent VNF (nsd:service_deployment_flavour:constituent_vnf) ............ 43
6.2.2
Network Service Record (nsr) .................................................................................................................. 43
6.2.2.1
nsr base element .................................................................................................................................. 43
6.2.2.2
Connection Point (nsr:connection_point)................................................................................ 43
6.3
Virtualised Network Function information elements ....................................................................................... 44
6.3.1
VNF Descriptor (vnfd) ............................................................................................................................. 44
6.3.1.1
vnfd base information elements .......................................................................................................... 44
6.3.1.2
Virtual Deployment Unit (vnfd:vdu) ................................................................................................ 45
6.3.1.2.1
vnfd:vdu base elements ............................................................................................................... 45
6.3.1.2.2
vnfd:vdu information elements related to CPUs ......................................................................... 46
6.3.1.2.3
vnfd:vdu information elements related to memory ..................................................................... 48
6.3.1.2.4
vnfd:vdu information elements related to security...................................................................... 48
6.3.1.2.5
vnfd:vdu information elements related to hypervisors ................................................................ 49
6.3.1.2.6
vnfd:vdu information elements related to PCIe .......................................................................... 49
6.3.1.2.7
vnfd:vdu information elements related to network interfaces ..................................................... 50
6.3.1.2.8
vnfd:vdu information elements related to virtual switches ......................................................... 51
6.3.1.2.9
vnfd:vdu information elements related to general reliability and availability ............................ 51
6.3.1.2.10
vnfd:vdu information elements related to storage ....................................................................... 52
6.3.1.3
VNF internal Virtual Link (vnfd:virtual_link)......................................................................... 52
6.3.1.4
Connection Point (vnfd:connection_point) ............................................................................. 52
6.3.1.5
Deployment flavour element (vnfd:deployment_flavour) ...................................................... 53
6.3.1.5.1
Constituent VDU (vnfd:deployment_flavour:constituent_vdu) ............................ 53
6.3.2
VNF Record (vnfr)................................................................................................................................... 53
6.3.2.1
vnfr base elements .............................................................................................................................. 54
6.3.2.2
VNF internal Virtual Link (vnfr:virtual_link)......................................................................... 55
6.3.2.3
Connection Point (vnfr:connection_point) ............................................................................. 55
6.3.2.4
Virtual Deployment Unit (vnfr:vdu) ................................................................................................ 56
6.3.2.4.1
vnfr:vdu base element ................................................................................................................ 56
6.3.2.4.2
VNFC instance (vnfr:vdu:vnfc_instance) ........................................................................ 56
6.3.2.4.3
Connection Point (vnfr:vdu: vnfc_instance:connection_point) ............................ 57
6.4
Virtual Link information elements ................................................................................................................... 57
6.4.1
Virtual Link Descriptor (vld) .................................................................................................................... 58
6.4.2
Virtual Link Record (vlr) ......................................................................................................................... 58
6.4.3
Relation between internal Virtual Links and Virtual Links ........................................................................ 59
6.5
Virtualised Network Function Forwarding Graph information elements ......................................................... 60
6.5.1
VNF Forwarding Graph Descriptor (vnffgd) .......................................................................................... 62
6.5.1.1
vnffgd base element........................................................................................................................... 62
6.5.1.2
Network Forwarding Path (vnffgd:network_forwarding_path) .......................................... 63
6.5.2
VNF Forwarding Graph Record (vnffgr) ............................................................................................... 63
6.5.2.1
vnffgr base element........................................................................................................................... 63
6.5.2.2
Network Forwarding Path (vnffgr:network_forwarding_path) .......................................... 63
6.6
Physical Network Function information elements ........................................................................................... 63
6.6.1
PNF Descriptor (pnfd) .............................................................................................................................. 64
6.6.1.1
pnfd base element ............................................................................................................................... 64
6.6.1.2
Connection Point (pnfd:connection_point) ............................................................................. 64
6.6.2
PNF Record (pnfr) ................................................................................................................................... 64
6.6.2.1
pnfr base element ............................................................................................................................... 64
ETSI
5
6.6.2.2
6.7
6.8
7
7.1
7.1.1
7.1.1.1
7.1.1.2
7.1.2
7.1.2.1
7.1.2.2
7.1.3
7.1.3.1
7.1.3.2
7.1.4
7.1.4.1
7.1.4.2
7.1.5
7.1.5.1
7.1.5.2
7.2
7.2.1
7.2.1.1
7.2.1.2
7.2.2
7.2.2.1
7.2.2.2
7.2.3
7.2.3.1
7.2.3.2
7.2.4
7.2.4.1
7.2.4.2
7.2.5
7.2.5.1
7.2.5.2
7.2.6
7.2.6.1
7.2.6.2
7.2.7
7.2.7.1
7.2.7.2
7.2.8
7.2.8.1
7.2.8.2
7.3
7.3.1
7.3.1.1
7.3.1.2
7.3.2
7.3.2.1
7.3.2.2
7.3.3
7.3.3.1
7.3.3.2
7.3.4
7.3.4.1
7.3.4.2
7.3.5
7.3.5.1
7.3.5.2
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Connection Point (pnfr:connection_point) ............................................................................. 65
VNF Instantiation input parameter ................................................................................................................... 65
Network Service Instantiation Input Parameters .............................................................................................. 65
NFV-MANO interfaces ..........................................................................................................................66
Interfaces concerning Network Services .......................................................................................................... 66
Network Service Descriptor management .................................................................................................. 66
Description ............................................................................................................................................ 66
Operations ............................................................................................................................................. 67
Network Service lifecycle management ..................................................................................................... 67
Description ............................................................................................................................................ 67
Operations ............................................................................................................................................. 68
Network Service lifecycle change notification ........................................................................................... 69
Description ............................................................................................................................................ 69
Operations ............................................................................................................................................. 69
Network Service performance management ............................................................................................... 70
Description ............................................................................................................................................ 70
Operations ............................................................................................................................................. 70
Network Service fault management ............................................................................................................ 71
Description ............................................................................................................................................ 71
Operations ............................................................................................................................................. 71
Interfaces concerning Virtualised Network Functions ..................................................................................... 71
VNF Package management ......................................................................................................................... 71
Description ............................................................................................................................................ 71
Operations ............................................................................................................................................. 72
VNF software image management.............................................................................................................. 72
Description ............................................................................................................................................ 72
Operations ............................................................................................................................................. 72
VNF lifecycle operation granting ............................................................................................................... 73
Description ............................................................................................................................................ 73
Operations ............................................................................................................................................. 73
VNF lifecycle management ........................................................................................................................ 74
Description ............................................................................................................................................ 74
Operations ............................................................................................................................................. 75
VNF lifecycle change notification .............................................................................................................. 76
Description ............................................................................................................................................ 76
Operations ............................................................................................................................................. 76
VNF configuration ...................................................................................................................................... 77
Description ............................................................................................................................................ 77
Operations ............................................................................................................................................. 77
VNF performance management .................................................................................................................. 78
Description ............................................................................................................................................ 78
Operations ............................................................................................................................................. 78
VNF fault management ............................................................................................................................... 79
Description ............................................................................................................................................ 79
Operations ............................................................................................................................................. 79
Interfaces concerning virtualised resources ...................................................................................................... 80
Virtualised resources catalogue management ............................................................................................. 80
Description ............................................................................................................................................ 80
Operations ............................................................................................................................................. 80
Virtualised resources capacity management ............................................................................................... 81
Description ............................................................................................................................................ 81
Operations ............................................................................................................................................. 81
Virtualised resources management ............................................................................................................. 82
Description ............................................................................................................................................ 82
Operations ............................................................................................................................................. 83
Virtualised resources performance management ........................................................................................ 84
Description ............................................................................................................................................ 84
Operations ............................................................................................................................................. 84
Virtualised resources fault management ..................................................................................................... 85
Description ............................................................................................................................................ 85
Operations ............................................................................................................................................. 85
ETSI
6
7.4
7.4.1
7.4.2
7.5
7.5.1
7.5.2
7.6
7.6.1
7.6.2
7.7
7.7.1
7.8
7.8.1
7.8.2
7.9
Policy administration interface ......................................................................................................................... 85
Description.................................................................................................................................................. 85
Operations ................................................................................................................................................... 86
Network Forwarding Path management interface ............................................................................................ 86
Description.................................................................................................................................................. 86
Operations ................................................................................................................................................... 86
NFVI hypervisor management interface .......................................................................................................... 87
Description.................................................................................................................................................. 87
Operations ................................................................................................................................................... 88
NFVI compute management interface .............................................................................................................. 90
Description.................................................................................................................................................. 90
NFVI networking management interface ......................................................................................................... 90
Description.................................................................................................................................................. 90
Operations ................................................................................................................................................... 91
Interfaces exposed between different service providers ................................................................................... 91
Annex A (informative):
A.1
VNF Instance management and orchestration case study .........................93
IMS MRF management and orchestration case study ............................................................................93
A.1.1
A.1.2
A.2
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
IMS MRF on-boarding ..................................................................................................................................... 95
IMS MRF instance provisioning and configuration ......................................................................................... 96
Network Service fault Management case study ...................................................................................100
Annex B (informative):
VNF lifecycle management .........................................................................102
B.1
Introduction ..........................................................................................................................................102
B.2
VNF Package on-boarding flows .........................................................................................................103
B.2.1
B.2.2
B.2.3
B.2.4
B.2.5
B.2.6
B.3
On-board VNF Package flow ......................................................................................................................... 104
Disable VNF Package flow ............................................................................................................................ 105
Enable VNF Package flow ............................................................................................................................. 105
Update VNF Package flow ............................................................................................................................. 106
Query VNF Packages flow ............................................................................................................................. 107
Delete VNF Package flow .............................................................................................................................. 107
VNF instantiation flows .......................................................................................................................108
B.3.1
B.3.1.1
B.3.1.2
B.3.1.3
B.3.1.4
B.3.1.5
B.3.1.6
B.3.1.7
B.3.2
B.3.2.1
B.3.2.2
B.4
VNF instantiation flows with resource allocation done by NFVO ................................................................. 108
VNF Check Feasibility ............................................................................................................................. 109
VNF instantiation flow ............................................................................................................................. 111
NFVO: validation ..................................................................................................................................... 112
NFVO: request to VNF Manager to instantiate the VNF.......................................................................... 113
VNF Manager: request validation and processing .................................................................................... 113
NFVO: pre-allocation processing ............................................................................................................. 113
Orchestration: resource allocation (compute, storage and network) and interconnection setup ............... 114
VNF instantiation flows with resource allocation done by VNF Manager..................................................... 115
VNF instantiation from EM ...................................................................................................................... 115
VNF instantiation from NFVO ................................................................................................................. 116
VNF instance scaling flows..................................................................................................................117
B.4.1
B.4.2
B.4.3
B.4.4
B.4.4.1
B.4.4.2
B.4.4.3
B.4.4.4
Detecting need to scale ................................................................................................................................... 117
Determining scaling action ............................................................................................................................. 118
Scaling flow with resource allocation done by NFVO ................................................................................... 119
Scaling flows with resource allocation done by VNF Manager ..................................................................... 121
Automatic VNF expansion triggered by VNF performance measurement results .................................... 121
EM initiated VNF expansion .................................................................................................................... 123
Automatic VNF contraction triggered by VNF performance measurement results .................................. 124
EM initiated VNF contraction .................................................................................................................. 125
B.5
VNF instance termination flows...........................................................................................................127
B.6
NFV fault management ........................................................................................................................128
Annex C (informative):
Network Service lifecycle management flows............................................131
ETSI
7
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
C.1
Introduction ..........................................................................................................................................131
C.2
Network Service on-boarding flows.....................................................................................................131
C.2.1
C.2.2
C.2.3
C.2.4
C.2.5
C.2.6
On-board Network Service Descriptor flow ................................................................................................... 132
Disable Network Service Descriptor flow ...................................................................................................... 133
Enable Network Service Descriptor flow ....................................................................................................... 133
Update Network Service Descriptor flow ...................................................................................................... 134
Query Network Service Descriptor flow ........................................................................................................ 135
Delete Network Service Descriptor flow........................................................................................................ 135
C.3
Network Service instantiation flows ....................................................................................................137
C.4
Network Service instance scaling.........................................................................................................139
C.4.1
C.4.2
Network Service instance scale-out ................................................................................................................ 139
Network Service instance scale-in.................................................................................................................. 141
C.5
Network Service instance update flows due to VNF instance modification ........................................142
C.6
Network Service instance termination flows........................................................................................145
C.7
VNF Forwarding Graph lifecycle management flows .........................................................................147
C.7.1
C.7.2
C.7.3
C.7.4
Create VNF Forwarding Graph ...................................................................................................................... 147
Update VNF Forwarding Graph ..................................................................................................................... 148
Query VNF Forwarding Graph ...................................................................................................................... 149
Delete VNF Forwarding Graph ...................................................................................................................... 149
Annex D (informative):
Orchestration flows......................................................................................151
D.1
Introduction ..........................................................................................................................................151
D.2
NFVI-PoP setup and configuration ......................................................................................................151
D.3
Resources provisioning/de-provisioning ..............................................................................................152
Annex E (informative):
E.1
VNFD/NSD representations using TOSCA ...............................................153
Describing IMS MRF with TOSCA .....................................................................................................153
E.1.1
E.1.2
E.1.2.1
E.1.2.2
E.1.2.3
TOSCA meta-model ....................................................................................................................................... 153
IMS MRF representation in TOSCA .............................................................................................................. 153
IMS MRF NSD ......................................................................................................................................... 153
MRB VNFD.............................................................................................................................................. 155
MRF VNFD .............................................................................................................................................. 156
Annex F (informative):
YANG/XML VNFD & NSD model ............................................................158
F.1
Use Cases of Network Service with Physical Network Functions .......................................................158
F.2
Examples ..............................................................................................................................................158
F.2.1
F.2.2
F.2.3
F.2.4
F.2.5
F.3
F.3.1
F.3.2
F.3.3
NSD for Gi LAN Network Service ................................................................................................................ 159
VNFD for Virtual Firewall ............................................................................................................................. 160
VNFD for Virtual Carrier Grade NAT ........................................................................................................... 163
VNFD for Virtual Video Cache ..................................................................................................................... 165
VLDs .............................................................................................................................................................. 168
YANG schema .....................................................................................................................................168
YANG schema for NSD ................................................................................................................................. 168
YANG schema for VNFD .............................................................................................................................. 170
YANG schema for VLD................................................................................................................................. 174
Annex G (informative):
TM Forum SID service model ....................................................................176
Annex H (informative):
Open Virtualisation Format .......................................................................179
Annex I (informative):
Other information ........................................................................................180
I.1
I.1.1
I.1.2
OpenStack ............................................................................................................................................180
Introduction .................................................................................................................................................... 180
NFV-related features in OpenStack community............................................................................................. 180
ETSI
8
Annex J (informative):
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Authors & contributors ...............................................................................182
History ............................................................................................................................................................184
ETSI
9
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Intellectual Property Rights
IPRs essential or potentially essential to the present document may have been declared to ETSI. The information
pertaining to these essential IPRs, if any, is publicly available for ETSI members and non-members, and can be found
in ETSI SR 000 314: "Intellectual Property Rights (IPRs); Essential, or potentially Essential, IPRs notified to ETSI in
respect of ETSI standards", which is available from the ETSI Secretariat. Latest updates are available on the ETSI Web
server (http://ipr.etsi.org).
Pursuant to the ETSI IPR Policy, no investigation, including IPR searches, has been carried out by ETSI. No guarantee
can be given as to the existence of other IPRs not referenced in ETSI SR 000 314 (or the updates on the ETSI Web
server) which are, or may be, or may become, essential to the present document.
Foreword
This Group Specification (GS) has been produced by ETSI Industry Specification Group (ISG) Network Functions
Virtualisation (NFV).
Modal verbs terminology
In the present document "shall", "shall not", "should", "should not", "may", "may not", "need", "need not", "will",
"will not", "can" and "cannot" are to be interpreted as described in clause 3.2 of the ETSI Drafting Rules (Verbal forms
for the expression of provisions).
"must" and "must not" are NOT allowed in ETSI deliverables except when used in direct citation.
ETSI
10
1
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Scope
The present document describes the management and orchestration framework required for the provisioning of
virtualised network functions (VNF), and the related operations, such as the configuration of the virtualised network
functions and the infrastructure these functions run on. The objectives are to define this framework, provide
requirements for management and orchestration, identify topics that may serve in later gap analysis against current
standards, identify best practices and provide guidance on how to address identified new topics. The focus of the
present document is on aspects of management and orchestration that are specific to NFV.
The present document addresses the following topics of management and orchestration: architecture framework for
management and orchestration of NFV, information elements, interfaces, provisioning, configuration, and operational
management, including interworking with existing operations and management systems.
2
References
2.1
Normative references
References are either specific (identified by date of publication and/or edition number or version number) or
non-specific. For specific references, only the cited version applies. For non-specific references, the latest version of the
referenced document (including any amendments) applies.
Referenced documents which are not found to be publicly available in the expected location might be found at
http://docbox.etsi.org/Reference.
NOTE:
While any hyperlinks included in this clause were valid at the time of publication, ETSI cannot guarantee
their long term validity.
The following referenced documents are necessary for the application of the present document.
Not applicable.
2.2
Informative references
References are either specific (identified by date of publication and/or edition number or version number) or
non-specific. For specific references, only the cited version applies. For non-specific references, the latest version of the
referenced document (including any amendments) applies.
NOTE:
While any hyperlinks included in this clause were valid at the time of publication, ETSI cannot guarantee
their long term validity.
The following referenced documents are not necessary for the application of the present document but they assist the
user with regard to a particular subject area.
[i.1]
ETSI GS NFV 001: "Network Functions Virtualisation (NFV); Use Cases".
[i.2]
ETSI GS NFV 002: "Network Functions Virtualisation (NFV); Architectural Framework".
[i.3]
ETSI GS NFV 003: "Network Functions Virtualisation (NFV); Terminology for Main Concepts in
NFV".
[i.4]
ETSI GS NFV-INF 001: "Network Functions Virtualisation; Infrastructure Overview".
[i.5]
ETSI GS NFV-INF 003: "Network Functions Virtualisation (NFV); Infrastructure; Compute
Domain".
[i.6]
ETSI GS NFV-INF 004: "Network Functions Virtualisation (NFV); Infrastructure; Hypervisor
Domain".
ETSI
11
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
[i.7]
ETSI GS NFV-INF 005: "Network Functions Virtualisation (NFV); Infrastructure; Network
Domain".
[i.8]
ETSI GS NFV-SWA 001: "Network Functions Virtualisation (NFV); Virtual Network Function
Architecture".
[i.9]
OpenStack: Cloud Software. [Online].
NOTE:
[i.10]
NOTE:
Available at: http://www.openstack.org.
OpenStack Havana. [Online].
Available at: https://www.openstack.org/software/havana/.
3
Definitions and abbreviations
3.1
Definitions
For the purposes of the present document, the terms and definitions given in ETSI GS NFV 003 [i.3] and the following
apply:
administrative domain: collection of systems and networks operated by a single organization or administrative
authority
NOTE:
The components which make up the domain are assumed to interoperate with a significant degree of
mutual trust among them based on a stable trust relationship, while a transient, specific trust relationship
shall be established for interoperating with components in other domains.
connection point: information element representing the virtual and/or physical interface that offers the network
connections between instances of NS, VNF, VNFC (based on the VDU information element), PNF and a VL
NOTE:
Some examples of virtual and/or physical interfaces are a virtual port, a virtual NIC address, a physical
port, a physical NIC address or the endpoint of an IP VPN.
consumer: role played by a functional block that consumes certain functions exposed by another functional block
deployment flavour: template that describes a specific deployment (of a Network Service or VNF) supporting specific
KPIs (such as capacity and performance)
infrastructure domain: administrative domain that provides virtualised infrastructure resources such as compute,
network, and storage or a composition of those resources via a service abstraction to another Administrative Domain,
and is responsible for the management and orchestration of those resources
producer: role played by a functional block that produces certain functions, and exposes them externally through
public interfaces to other functional blocks
resource orchestration: subset of NFV Orchestrator functions that are responsible for global resource management
governance
3.2
Abbreviations
For the purposes of the present document, the abbreviations given in ETSI GS NFV 002 [i.2] and the following apply:
ADC
API
BRAS
BSS
CIM
CMS
CPE
CPU
Application Delivery Controller
Application Programming Interface
Broadband Remote Access Server
Business Support System
Common Information Model
Cloud Management System
Customer Premise Equipment
Central Processing Unit
ETSI
12
CSAR
DMTF
DPDK
DSL
DSLAM
E2E
ECC
EM
EMS
FCAPS
GB
GRE
HA
HW
ID
IGMP
IMS
IOMMU
IOTLB
IP
IPMI
ISG
KPI
KQI
KVM
LAN
LRO
LSO
MAC
MEF
MPLS
MRB
MRF
MTU
NCT
NF
NFP
NFV
NFVI
NFVI-PoP
NFV-MANO
NFVO
NIC
NMS
N-PoP
NS
NSD
NSR
NVGRE
OSS
OVF
PCI
PNF
PNFD
PNFR
PRS
PXE
QoS
RCA
RDMA
RSS
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Cloud Service Archive
Distributed Management Task Force
Data Plane Development Kit
Digital Subscriber Line
DSL Access Multiplexer
End to end
Error Correcting Code
Element Management
Element Management System
Fault Management, Configuration Management, Accounting Management, Performance
Management, and Security Management
Gigabyte
Generic Routing Encapsulation
High Availability
Hardware
Identifier
Internet Group Management Protocol
IP Multimedia System
Input/Output Memory Management Unit
Input/Output Translation Lookaside Buffer
Internet Protocol
Intelligent Platform Management Interface
Industry Specification Group
Key Performance Indicator
Key Quality Indicator
Kernel Virtual Machine
Local Area Network
Large Receive Offload
Large Segmentation Offload
Media Access Control
Metro Ethernet Forum
Multi-Protocol Label Switching
Media Resource Broker
Media Resource Function
Maximum Transmission Unit
Network Connection Topology
Network Function
Network Forwarding Path
Network Functions Virtualisation
Network Functions Virtualisation Infrastructure
NFVI Point of Presence
NFV Management and Orchestration
Network Functions Virtualisation Orchestrator
Network Interface Card
Network Management System
Network Point of Presence
Network Service
Network Service Descriptor
Network Service Record
Network Virtualization using Generic Routing Encapsulation
Operations Support System
Open Virtualisation Format
Peripheral Component Interconnect
Physical Network Function
Physical Network Function Descriptor
Physical Network Function Record
Problem Resolution Standard
Preboot Execution Environment
Quality of Service
Root Cause Analysis
Remote Direct Memory Access
Receive Side Scaling
ETSI
13
RTT
SDN
SHA
SID
SLA
SMT
SR-IOV
S-CSCF
SWA
TCP
TCP/IP
TLB
TOSCA
URI
VDU
VEPA
vEPC
VIM
VL
VLAN
VLD
VLR
VM
VNF
VNFC
VNFD
VNFFG
VNFFGD
VNFFGR
VNFM
VNFR
vPGW
VPN
VXLAN
WAN
WIM
XML
XSD
4
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Round Trip Time
Software-Defined Networking
Secure Hash Algorithm
Information Framework
Service Level Agreement
Simultaneous Multithreading
Single Root Input/Output Virtualization
Service Call Session Control Function
Software Architecture Work group
Transmission Control Protocol
Transmission Control Protocol / Internet Protocol
Translation Lookaside Buffer
Topology and Orchestration Specification for Cloud Applications
Uniform Resource Identifier
Virtualisation Deployment Unit
Virtual Ethernet Port Aggregator
Virtual Evolved Packet Core
Virtualised Infrastructure Manager
Virtual Link
Virtual LAN
Virtual Link Descriptor
Virtual Link Record
Virtual Machine
Virtualised Network Function
Virtual Network Function Component
Virtualised Network Function Descriptor
VNF Forwarding Graph
VNFFG Descriptor
VNFFG Record
VNF Manager
VNF Record
Virtual Packet Gateway
Virtual Private Network
Virtual eXtensible LAN
Wide Area Network
WAN Infrastructure Manager
Extensible Markup Language
XML Schema Definition
NFV Management and Orchestration: objectives and
concepts
This clause provides an overview of the NFV management and orchestration objectives.
4.1
Overview
Network Functions Virtualisation (NFV) adds new capabilities to communications networks and requires a new set of
management and orchestration functions to be added to the current model of operations, administration, maintenance
and provisioning. In legacy networks, Network Function (NF) implementations are often tightly coupled with the
infrastructure they run on. NFV decouples software implementations of Network Functions from the computation,
storage, and networking resources they use. The virtualisation insulates the Network Functions from those resources
through a virtualisation layer.
The decoupling exposes a new set of entities, the Virtualised Network Functions (VNFs), and a new set of relationships
between them and the NFV Infrastructure (NFVI). VNFs can be chained with other VNFs and/or Physical Network
Functions (PNFs) to realize a Network Service (NS).
ETSI
14
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Since Network Services (including the associated VNF Forwarding Graphs (VNFFGs), Virtual Links (VLs), Physical
Network Functions (PNFs), VNFs, NFVI and the relationships between them did not exist before the emergence of
NFV, their handling requires a new and different set of management and orchestration functions. The Network
Functions Virtualisation Management and Orchestration (NFV-MANO) architectural framework has the role to manage
the NFVI and orchestrate the allocation of resources needed by the NSs and VNFs. Such coordination is necessary now
because of the decoupling of the Network Functions software from the NFVI.
The virtualisation principle stimulates a multi-vendor ecosystem where the different components of NFVI, VNF
software, and NFV-MANO architectural framework entities are likely to follow different lifecycles (e.g. on
procurement, upgrading, etc.). This requires interoperable standardised interfaces and proper resource abstraction
among them.
The present document focuses primarily on the differences introduced by the Network Functions Virtualisation process,
and not on aspects of the Network Functions that remained constant (i.e. the latter continue to be the focus of preexisting entities). Those differences can be described by grouping them in the following hierarchy:
•
virtualised infrastructure;
•
virtualised network function;
•
Network Service.
The elaboration of NFV-MANO aspects below is structured according to this hierarchy. The management of end to end
services is not in the scope of the present document.
NOTE:
4.2
The concept of reusable VNF groups is not supported in the present document.
Management and Orchestration aspects of Network
Functions Virtualisation Infrastructure
Network Functions Virtualisation Infrastructure (NFVI) resources under consideration are both virtualised and
non-virtualised resources, supporting virtualised network functions and partially virtualised network functions.
Virtualised resources in-scope are those that can be associated with virtualisation containers, and have been catalogued
and offered for consumption through appropriately abstracted services, for example:
•
Compute including machines (e.g. hosts or bare metal), and virtual machines, as resources that comprise both
CPU and memory.
•
Storage, including: volumes of storage at either block or file-system level.
•
Network, including: networks, subnets, ports, addresses, links and forwarding rules, for the purpose of
ensuring intra- and inter-VNF connectivity.
The management and orchestration of virtualised resources should be able to handle NFVI resources (e.g. in NFVI
Nodes), in NFVI Points of Presence (NFVI-PoPs). Management of non-virtualised resources is restricted to
provisioning connectivity to PNFs, necessary when a NS instance includes a PNF that needs to connect to a VNF, or
when the NS instance is distributed across multiple NFVI-PoPs or N-PoPs.
The virtualised resources are leveraged for providing VNFs with the resources they need. Resource allocation in the
NFVI is a potentially complex task because a lot of requirements and constraints may need to be met at the same time.
Particularly requirements for network allocation add new complexity compared to known resource allocation strategies
for computing resources in virtualised environments. For example, some VNFs require low latency or high bandwidth
links to other communication endpoints.
Allocation and release of resources is a dynamic process, in response to consumption of those services by other
functions. While the management and orchestrations function for virtualised infrastructure are VNF-unaware, resource
allocations and releases may be needed throughout the VNF lifetime. An advantage of NFV is that with increasing load
VNFs can dynamically consume services that allocate additional resource when scaling-out is triggered.
Services exposing virtualised resources include (non-exhaustive list):
•
discovery of available services;
ETSI
15
•
management of virtualised resources availability/allocation/release;
•
virtualised resource fault/performance management.
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
In the case of virtualised resources distributed across multiple NFVI-PoPs, those services could either be exposed
directly by the management and orchestration functions for each individual NFVI-PoP, or via a higher-level service
abstraction presenting the virtualised resources across multiple NFVI-PoPs. Both types of services could be exposed to
the consuming functions. In the case of the higher level service abstraction previously mentioned, the management and
orchestration of virtualised resources and non-virtualised networking resources across those NFVI-PoPs falls under the
responsibility of the management and orchestration of the virtualised infrastructure that may in turn use the services
exposed directly by the management and orchestration functions of a single or across multiple NFVI-PoPs. In order to
provide those services, the management and orchestration of the virtualised infrastructure consumes services provided
by the NFVI.
The NFV management and orchestration functions that coordinate virtualised resources in a single NFVI-PoP and/or
across multiple NFVI-PoPs need to ensure exposure of services that support accessing these resources in an open, well
known abstracted manner. These services can be consumed by other authenticated and properly authorized NFV
management and orchestration functions (e.g. functions that manage and orchestrate virtualised network functions).
4.3
Management and Orchestration aspects of Virtualised
Network Functions
Management and orchestration aspects of a VNF include traditional Fault Management, Configuration Management,
Accounting Management, Performance Management, and Security Management (FCAPS), but the focus in the present
document is on the newer aspect introduced by NFV. The decoupling of Network Functions from the physical
infrastructure they use results in a new set of management functions focused on the creation and lifecycle management
of the needed virtualised resources for the VNF, collectively referred to as VNF Management.
VNF Management functions are responsible for the VNF's lifecycle management including operations such as:
•
Instantiate VNF (create a VNF using the VNF on-boarding artefacts).
•
Scale VNF (increase or reduce the capacity of the VNF).
•
Update and/or Upgrade VNF (support VNF software and/or configuration changes of various complexity).
•
Terminate VNF (release VNF-associated NFVI resources and return it to NFVI resource pool).
The deployment and operational behaviour requirements of each VNF is captured in a deployment template, and stored
during the VNF on-boarding process in a catalogue, for future use. The deployment template describes the attributes
and requirements necessary to realize such a VNF and captures, in an abstracted manner, the requirements to manage its
lifecycle. The VNF Management functions perform the lifecycle management of a VNF based on the requirements in
the template. At instantiation NFVI resources are assigned to a VNF based on the requirements captured in the
deployment template but also taking into consideration specific requirements, constraints, and policies that have been
pre-provisioned or are accompanying the request for instantiation. The inclusion of lifecycle management requirements
in the deployment template allows the VNF Management functions to handle similarly a very simple VNF with a single
component or a highly complex VNF with multiple components and inter-dependencies - affording flexibility to the
VNF provider.
During the lifecycle of a VNF, the VNF Management functions may monitor KPIs of a VNF, if such KPIs were
captured in the deployment template. The management functions may use this information for scaling operations.
Scaling may include changing the configuration of the virtualised resources (scale up, e.g. add CPU, or scale down,
e.g. remove CPU), adding new virtualised resources (scale out, e.g. add a new VM), shutting down and removing VM
instances (scale in), or releasing some virtualised resources (scale down).
The VNF Management performs its services by maintaining the virtualised resources that support the VNF
functionality, without interfering with the logical functions performed by the VNFs, and its functions are exposed in an
open, well known abstracted manner as services to other functions.
The services provided by VNF Management can be consumed by authenticated and properly authorized NFV
management and orchestration functions (e.g. functions that manage Network Services).
ETSI
16
4.4
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Management and Orchestration aspects of Network
Services
The Network Service Orchestration is responsible for the Network Service lifecycle management including operations
such as:
•
On-board Network Service, i.e. register a Network Service in the catalogue and ensure that all the templates
describing the NS are on-boarded.
•
Instantiate Network Service, i.e. create a Network Service using the NS on-boarding artefacts.
•
Scale Network Service, i.e. grow or reduce the capacity of the Network Service.
•
Update Network Service by supporting Network Service configuration changes of various complexity such as
changing inter-VNF connectivity or the constituent VNF instances.
•
Create, delete, query, and update of VNFFGs associated to a Network Service.
•
Terminate Network Services, i.e. request the termination of constituent VNF instances, request the release of
NFVI resources associated to NSs, and return them to NFVI resource pool if applicable.
The deployment and operational behaviour requirements of each Network Service is captured in a deployment template,
and stored during the Network Service on-boarding process in a catalogue, for future selection for instantiation. The
deployment template fully describes the attributes and requirements necessary to realize such a Network Service.
Network Service Orchestration coordinates the lifecycle of VNFs that jointly realize a Network Service. This includes
(not limited to) managing the associations between different VNFs, and when applicable between VNFs and PNFs, the
topology of the Network Service, and the VNFFGs associated with the Network Service.
During the Network Service lifecycle, the Network Service Orchestration functions may monitor KPIs of a Network
Service if such requirements were captured in the deployment template, and may report this information to support
explicit request for such operations from other functions.
The Network Service Orchestration performs its services by using the VNF Management services and by orchestrating
the NFV Infrastructure that supports the interconnection between VNFs functionality, and its functions are exposed in
an open, well known abstracted manner as services to other functions. In order to fulfil its responsibilities, the Network
Service Orchestration functions consume services exposed by other functions (e.g. Virtualised Infrastructure
Management functions).
The services provided by Network Service Orchestration can be consumed by authenticated and properly authorized
other functions (e.g. Operations Support System (OSS), Business Support System (BSS)).
4.5
Other management and orchestration aspects of NFV
framework
This clause provides a high level description of some additional management and orchestration aspects of NFV, that
apply to management of virtualised resources, VNFs, and NSs.
4.5.1
Fault and performance management
Fault and performance management are functionalities that shall exist in any NFV framework. They support the
assurance aspects of the lifecycle of any Network Service instance or VNF instance, in order to provide Service Level
Agreement (SLA) management. Fault and performance management functionality is typically distributed over different
functional blocks that are dedicated among others to fault and performance measurement, resultscalculation and
aggregation, fault correlation, and fault resolution.
Fault notifications may be the result of several sources of faults: physical infrastructure (compute, storage, and
networking related faults); virtualised infrastructure (e.g. VM-related faults), and application logic (i.e. VNF instance
related faults).
ETSI
17
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Fault correlation and root-cause analysis are processes that determine the reason for faults conditions and the impact of
fault conditions. Once correlated and analysed, the correlated fault information helps determining the necessary
corrective actions, and it helps triggering such actions at one or more fault resolution points, within the NFV
Framework or outside the NFV Framework (e.g. OSS).
Fault correlation can be centralized or distributed among multiple functional blocks, and no assumption is made here
regarding either. Ideally, faults should be analysed and resolved as soon as possible, hence at the functional block that
has sufficient information to perform the root-cause analysis and correlation and to determine the necessary corrective
action.
4.5.2
Policy Management
NFV policy management refers to the management of rules governing the behaviour of NFV-MANO functions
(e.g. management of VNF or NS scaling operations, access control, resource management, fault management, etc.).
Policies are defined with conditions and corresponding actions. For example, a scaling policy may state to execute the
related actions if the required conditions (e.g. VNF low CPU usage) were to manifest during runtime. Different actions
defined on the policy can be mutually exclusive resulting in process of selecting a particular action (or set of actions) to
be executed. Once declared, a policy may be bound to one or more NS instances, VNF instances, and NFVI resources.
NFV-MANO needs to support the execution of policies both automatically and manually, (e.g. by requiring manual
intervention from other operation systems, e.g. OSS/BSS).
4.5.3
Testing aspects of Network Services
As part of the NS instantiation process, as well as at any other times during the NS lifecycle, there may be a need to
perform testing (performance, operational, functional, etc.) for various reasons. For example, performance testing is a
key step of network engineering across all levels, and its main goal is to detect problems before a service goes live.
Additionally, in NFV deployments, testing has a subtle yet critical role - tuning the particular NFVI configuration. Any
sufficiently complex Network Service needs to be tested across multiple aspects, and across different network paths
(subsets or end-to-end); it can be tested as a whole, in parts (segments), or in parallel. The test plan for a Network
Service should be integrated and created as part of the design process of the service. The different ways to integrate a
test plan in a NS design are deliberately not covered by the present document.
The testing of a single VNF is a subset case of testing of an entire Network Service. In particular it is a necessary step
before placing a VNF "in service", e.g. on a new NFVI and in cases where it is determined that the VNF fails to perform
at the expected levels. Healing a VNF is a process that in some cases may require functional testing, before determining
the appropriate healing procedure.
NS testing should not impact the services provided by the NS.
NS testing procedures may be described in appropriate templates, supporting automation via the Network Service
Orchestration functions, or administered manually.
4.6
Relation of NFV management and orchestration with
existing operations and management systems
4.6.1
Overview
NFV-MANO functions play an important role in delivering the business benefits envisioned by the NFV ISG. These
benefits are achieved through interworking with other functional components in the NFV framework, as well as external
entities, such as Operations and Business Support Systems (OSS/BSS).
OSS/BSS include the collection of systems and management applications that service providers use to operate their
business. NFV is disruptive to many OSS/BSS functions, as the success of such a technology enabled business
transformation is predicated on changes in mind-sets, skillsets and toolsets to support corresponding process
re-engineering efforts.
ETSI
18
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
NFV is a paradigm shift in how the networks that underpin today's service provider infrastructures are built and
operated, and how the services they deliver are managed. New degrees of freedom are introduced to the network and its
management as resources now may be added, changed, and removed dynamically. This change opens up a wave of new
business opportunities. However, a new and highly agile operational approach is needed to take full advantage of these
opportunities.
4.6.2
NFV-MANO Interworking with OSS/BSS to deliver NFV business
benefits
NFV-MANO alone cannot deliver all the NFV business benefits; it needs to integrate and interwork with other
management entities for this purpose (e.g. OSS, BSS), using interfaces offered by those entities and offering its own
interfaces to be used by external entities.
In order for service providers to achieve the full benefit of NFV, NFV-MANO solutions should be considered
holistically alongside OSS/BSS integration and management requirements. Simply extending existing OSS/BSS models
to account for virtualisation will not be sufficient, because this approach will not support the new value-added
capabilities and services provided by NFV. Efforts are needed to ensure that NFV-MANO and OSS/BSS evolution is
coordinated so as to jointly support the following:
•
Open and consistent interfaces, to facilitate automation, self-service operations at service and product level
that can respond with the speed and agility required by changing business needs.
•
Adaptive automation, where service usage drives on-demand resource requirements, triggering feedback from
the system that the management functions analyse and make changes to, enabling the infrastructure to provide
the resources and services needed at that point in time.
•
Orchestration, where policies (and other mechanisms) can guide the decisions required to change all or part of
the system to perform a given function.
•
Personalized services that are easily configured, by the operator and/or end-user at the service and/or network
resource layers, to fit individual customer preferences and requirements.
•
Technology-driven innovation, where rapid development, continuous integration, deployment, and
experimentation, meet business and service operations agility and enable the migration to next generation
operations.
4.6.3
Key Challenges and Considerations
Management of end-to-end services by OSS/BSS requires convergence on a common approach to presenting
management services and management information from both legacy network systems and NFV based systems, so that
accurate end-to-end management views can be derived.
Standards are essential to achieve this convergence process. However, experience has taught the industry that this is
already a major issue for converged fixed and mobile network management including applications and services. It will
be even more acute when managing end-to-end services across a mixture of NFV functions, infrastructure, and legacy
interconnected network systems, in a highly dynamic NFV environment.
Some design patterns are used for converged fixed and mobile network operations. These patterns are shared between
interfaces, and are exposed by diverse network management systems. This enables a common way to integrate an
end-to-end view across multiple resources:
•
To describe the semantics and relationships of information exchanged across management interfaces, a
federated information model that provides common and consensually defined terms, concepts, and objects is
required.
•
To lower integration complexity, a set of common interface operations are required for covering request and
response patterns, as well as the behaviour and sequencing of those requests and responses for each of fault,
configuration, accounting, performance, inventory, and security management.
•
To improve consistency in representing information in end-to-end applications, common data types and
common vocabularies are required. For example: specification of fault codes and location codes used across
multiple resources.
ETSI
19
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
This pattern based approach, developed from prior converged network management studies, can help to inform and
guide future NFV standards, open source and proof of concept project development and their integration approaches.
More importantly, it may help the industry in managing converged networks (fixed and mobile) with a similar
methodology and complementary functional realizations.
Another key challenge is automation. Operators need to speed up the ability to dynamically incorporate change. The
agile management of network functions, applications, and services provides new opportunities for revenue while
reducing costs. Automation may be supported by an aligned approach between the OSS/BSS and the NFV-MANO
functionality based on, for example:
•
Open and standardized interfaces.
•
Common information models.
•
Mappings to different data models.
•
Ability to incorporate real-time data analytics.
The key challenge in this area is to provide a flexible and dynamic policy-based management approach.
Furthermore, the need for real-time processing (in conjunction with offline processing) is an additional key challenge
that plays an increasingly important role. It is desirable to support real-time processing by automation.
The support of the processing of a huge amount of data, including data analytics, based on several data sources
(e.g. structured, semi-structured, and unstructured data from the infrastructure as well as other sources), also in real
time, is a further key challenge in the OSS/BSS and NFV context.
To address the need for dynamic operation of NFV management and orchestration, the management architecture shall
support run-time integration of management functions and processes. This is different from today's static integration, as
it implies that interfaces shall be adaptive to support business agility.
In other words, the interfaces in NFV-MANO may not need to mandate a specific list of end points based on the
information being exchanged; in contrast, they may instead allow for different entities to consume the information
based on service necessity. A loosely-coupled set of interfaces will allow services and applications, as well as
associated business and operational processes, to be data-driven and controlled by policies. This adaptive approach
further enables a steady evolution of existing legacy systems to integrate with NFV.
4.7
Administrative Domains
In a typical scenario for NFV there will not be a single organization controlling and maintaining a whole NFV system.
The organizations may either be departments of the same root organization e.g. a network department and an IT
datacentre department or they may be different companies providing different functional blocks of the NFV
Architectural Framework. The basic use cases for these scenarios are described in ETSI GS NFV 001 [i.1].
Administrative Domains can be mapped to different organizations and therefore can exist within a single service
provider or distributed among several service providers. Administrative Domains can be collapsed, resulting in different
deployment options.
For the purpose of the present document, the following administrative types of domains are identified:
•
Infrastructure Domain.
•
Tenant Domain.
NOTE:
The present document only focuses on two types of Administrative Domains; additional domain types
may be identified later to support different deployment scenarios.
An Infrastructure Domain may be defined by different criteria (e.g. by organization, by type of resource such as
networking, compute and storage as in traditional datacentre environments, by geographical location, etc.), and multiple
Infrastructure Domains may co-exist. An Infrastructure Domain may provide infrastructure to a single or multiple
Tenant Domains. The Infrastructure Domain is application agnostic and thus has no insight of what is executed within a
VNF.
ETSI
20
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
A Tenant Domain may be defined by different criteria (e.g. organization, by type of Network Service, etc.), and
multiple Tenant Domains may co-exist. A Tenant Domain may use infrastructure in a single or multiple Infrastructure
Domains.
While the VNFs and Network Services reside in the Tenant Domain, in order to materialize they need the NFVI
resources from the Infrastructure Domains. Therefore the Tenant Domain consumes resources from one or more
Infrastructure Domains using the Infrastructure Domain orchestration functionality to orchestrate and operate virtual
infrastructure resources required by VNFs and Network Services built using those VNFs. Typical management tasks
within the Tenant Domain are VNF FCAPS. Typical management and orchestration tasks requiring the involvement of
both the Tenant Domain and the Infrastructure Domain are on-boarding of VNFs, instantiation of VNFs and scaling of
VNFs. The Tenant Domain is application aware, in the sense that functional blocks in this Administrative Domain
understand and support the logical functionality of the VNFs.
Other typical relationships between a Tenant Domain and an Infrastructure Domain include (non-exhaustive):
•
A Tenant Domain may use VNFs belonging to that Tenant Domain and/or to other Tenant Domains, in order
to realize the Tenant Domain's Network Services.
•
A Tenant Domain and an Infrastructure Domain could be mapped against the same or different organizations
(in a particular in the case when the Infrastructure Domain is offering infrastructure to a Tenant Domain).
•
While typically an Infrastructure Domain provides infrastructure services to a Tenant Domain, it may also
provide infrastructure services to one or more other Infrastructure Domains.
The identification of different Administrative Domains leads to the requirement to use different management and
orchestration functionality in those domains. The cross-relationships between the different Administrative Domains
leads to the identification of the requirement that NFV-MANO functionality is not monolithic, but rather needs to be
layered: management and orchestration functionality in an Administrative Domain may provide services that are
consumed by management and orchestration functionality in a different Administrative Domain.
However, the specific mapping of Administrative Domains to NFV-MANO functions in different layers is deliberately
left out of scope for the present document, since many different mappings can be supported, depending on the specifics
of the deployment environment.
5
Management and Orchestration architectural
framework
5.1
Overview
This clause expands on the NFV Architectural Framework (ETSI GS NFV 002 [i.2]) by providing a functional
architecture with more granular distribution and a more detailed description of the NFV Management and Orchestration
(NFV-MANO) functionality, as well as a description of the reference points between NFV-MANO functional blocks
and other functional blocks in the E2E NFV reference architecture. The NFV-MANO architectural framework focuses
on the aspects brought into the operator's networks by NFV, hence mainly on the new functional blocks and reference
points between those functional blocks. Other functional blocks and reference that may be necessary for a better
understanding of the NFV-MANO architectural framework may be included and briefly described, but the detailed
description for those are in-scope for other NFV documents, which the present document references.
5.2
Principles of NFV-MANO
The NFV-MANO architectural framework relies on a set of principles that support the combination of the concepts of
distinct Administrative Domains and layered management and orchestration functionality in each of those domains:
•
The architecture needs to include the possibility of multiple NFV-MANO functional blocks in the Tenant
Domain leveraging resources in the same Infrastructure Domain, and there shall a general architectural
principle of horizontal equality between those NFV-MANO functional blocks.
ETSI
21
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
•
Orchestration is provided in multiple functional blocks and it is important to emphasize that there is no one
functional block that has primacy over others. There shall be a general architectural principle of equality used
in distribution of orchestration functionality.
•
The NFV-MANO functional blocks in the Infrastructure Domain should offer abstracted services to functional
blocks in the Tenant Domain. The aspects of resource management corresponding to these abstracted services
should be fully embedded as selectable resource policy within them.
•
It should be possible to leverage best cloud management practices such assignment of resources, according to
the availability of the abstract services presented by the Infrastructure Domain to the Tenant Domain.
•
The NFV-MANO functionality may be realized in different ways, for example, as a monolithic single instance,
as a scalable system with multiple load-sharing instances, as a system of distributed cooperating instances, or
as a functionally decomposed and distributed system. It may be realized as an extension of a cloud/network
management system or as a separate system interacting with cloud/network management systems for realizing
NFV.
The present document does not mandate any specific realization of the NFV-MANO architectural framework, and
instead it recommends that the following best practices are leveraged and the NFV-MANO architectural framework
should:
•
lend itself to virtualisation and be implementable in software only;
•
lend itself to possible distribution and scaling across the NFVI in order to improve service availability and
support different locations;
•
lend itself to full automation (ability to react to events in real-time without human intervention, and execute
actions associated with those events, based on pre-provisioned templates and policies);
•
lend itself to implementations that do not contain any single points of failures with the potential to endanger
service continuity;
•
lend itself to an implementation with an open architecture approach and should expose standard or "de-facto"
standard interfaces;
•
support and help realize the feasible decoupling of VNF software from the hardware;
•
support management and orchestration of VNFs and Network Services using NFVI resources from a single or
across multiple NFVI-PoPs;
•
support modelling of NFVI resource requirements of VNFs in an abstracted way;
•
support modelling of NFVI resources in a way that an abstraction of them can be exposed by functionality in
one layer, to functionality in a different layer.
5.3
NFV-MANO architectural framework overview
The following entities from the NFV architectural framework (ETSI GS NFV 002 [i.2]) are considered within scope of
the NFV-MANO architectural framework:
a)
Functional blocks identified as belonging to NFV Management and Orchestration (NFV-MANO).
b)
Other functional blocks that interact with NFV-MANO via reference points.
c)
All reference points that enable communications to, from, and within NFV-MANO.
The NFV-MANO architectural framework represented in figure 5.1 is at a functional level and it does not imply any
specific implementation. Multiple functional blocks may be merged internalizing the reference point between them.
Finally, reference points drawn in bold and continuous lines and labelled are considered new and critical for
NFV-MANO architectural framework, representing potential targets for development in open source projects and/or
later standardization.
ETSI
22
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
The NFV-MANO architectural framework follows the principles enumerated in clause 5.1. Each of the functional
blocks has a well-defined set of responsibilities and operates on well-defined entities, using management and
orchestration as applicable within the functional block, as well as leveraging services offered by other functional blocks.
The NFV-MANO architectural framework identifies the following NFV-MANO functional blocks:
•
Virtualised Infrastructure Manager (VIM).
•
NFV Orchestrator (NFVO).
•
VNF Manager (VNFM).
NFV-MANO architectural framework identifies the following data repositories:
•
NS Catalogue.
•
VNF Catalogue.
•
NFV Instances repository.
•
NFVI Resources repository.
The NFV-MANO architectural framework identifies the following functional blocks that share reference points with
NFV-MANO:
•
Element Management (EM).
•
Virtualised Network Function (VNF).
•
Operation System Support (OSS) and Business System Support functions (BSS).
•
NFV Infrastructure (NFVI).
The NFV-MANO architectural framework identifies the following main reference points:
•
Os-Ma-nfvo, a reference point between OSS/BSS and NFVO.
•
Ve-Vnfm-em, a reference point between EM and VNFM.
•
Ve-Vnfm-vnf, a reference point between VNF and VNFM.
•
Nf-Vi, a reference point between NFVI and VIM.
•
Or-Vnfm, a reference point between NFVO and VNFM.
•
Or-Vi, a reference point between NFVO and VIM.
•
Vi-Vnfm, a reference point between VIM and VNFM.
Table 5.1 illustrates the mapping between the named reference points defined in ETSI GS NFV 002 [i.2] and the named
reference points defined in the present document.
ETSI
23
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Table 5.1
Reference
point in ETSI
GS NFV 002
[i.2]
Os-Ma
Reference point
in the present
document
Os-Ma-nfvo
Or-Vnfm
Ve-Vnfm
Or-Vnfm
Ve-Vnfm-em,
Ve-Vnfm-vnf
Vi-Vnfm
Or-Vi
Nf-Vi
VnNf
Vi-Vnfm
Or-Vi
Nf-Vi
Vn-Nf
Notes
The Os-Ma reference point in ETSI GS NFV 002 [i.2] terminates at the NFV-MANO
boundary. In the present document, the Os-Ma-nfvo reference point is derived from
Os-Ma, and terminates at the NFVO boundary. In the present document, no other
reference points have been identified between OSS/BSS and NFV-MANO functional
blocks.
Unchanged
The Ve-Vnfm reference point in ETSI GS NFV 002 [i.2] terminates at a boundary
comprising multiple functional blocks. In the present document, the Ve-Vnfm-em
and the Ve-Vnfm-vnf reference points are derived from Ve-Vnfm, and Ve-Vnfm-em
terminates at the EM boundary, while Ve-Vnfm-vnf terminates at the VNF boundary.
Unchanged
Unchanged
Unchanged
Unchanged
Other unnamed reference points have been added between NFV Orchestrator and the VNF Catalogue, VNFM and the
VNF Catalogue, NFV Orchestrator and the NFV Service Catalogue, NFV Orchestrator and the NFV Instances
repository, as well as between NFV Orchestrator and NFVI Resources repository.
Figure 5.1: The NFV-MANO architectural framework with reference points
The following clauses define the functional blocks and the reference points in-scope for the NFV-MANO architectural
framework.
ETSI
24
5.4
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
NFV-MANO architectural framework functional blocks
This clause describes the NFV-MANO architectural framework functional blocks and reference points.
5.4.1
NFV Orchestrator (NFVO)
The NFV Orchestrator has two main responsibilities:
•
the orchestration of NFVI resources across multiple VIMs, fulfilling the Resource Orchestration functions
described in clause 4.2;
•
the lifecycle management of Network Services, fulfilling the Network Service Orchestration functions, as
described in clause 4.4.
NOTE:
In the present document, for simplicity reasons, the two responsibilities are kept within one functional
block and sharing a common information base represented by the NFV Instances and NFV Resources
repository. To support different multi-vendor deployments and/or different mappings of functionality to
Administrative Domains, the two responsibilities may be separated in future. The NFVO uses the
Network Service Orchestration functions to coordinate groups of VNF instances as Network Services that
jointly realize a more complex function, including joint instantiation and configuration, configuring
required connections between different VNFs, and managing dynamic changes of the configuration,
e.g. for scaling the capacity of the Network Service. The Network Service Orchestration function uses the
services exposed by the VNF Manager function and by the Resource Orchestration function.
The following list expresses the non-exhaustive set of capabilities provided by the NFVO, via its Network Service
Orchestration functions. These capabilities may be exposed by means of interfaces consumed by NFV-MANO
functional blocks or by authorised external entities:
•
Management of Network Services deployment templates and VNF Packages (e.g. on-boarding new Network
Services and VNF Packages). During on-boarding of NS and VNF, a validation step is required. To support
subsequent instantiation of a NS, respectively a VNF, the validation procedure needs to verify the integrity and
authenticity of the provided deployment template, and that all mandatory information is present and consistent.
In addition, during the on-boarding of VNFs, software images provided in the VNF Package for the different
VNF components are catalogued in one or more NFVI-PoPs, using the support of VIM.
•
Network Service instantiation and Network Service instance lifecycle management, e.g. update, query, scaling,
collecting performance measurement results, event collection and correlation, termination.
•
Management of the instantiation of VNF Managers where applicable.
•
Management of the instantiation of VNFs, in coordination with VNF Managers.
•
Validation and authorization of NFVI resource requests from VNF Managers, as those may impact Network
Services (granting of the requested operation needs to be governed by policies).
•
Management of the integrity and visibility of the Network Service instances through their lifecycle, and the
relationship between the Network Service instances and the VNF instances, using the NFV Instances
repository.
•
Management of the Network Service instances topology (e.g. create, update, query, delete VNF Forwarding
Graphs).
•
Network Service instances automation management (e.g. trigger automatic operational management of NS
instances and VNF instances, according to triggers and actions captured in the on-boarded NS and VNF
deployment templates and governed by policies applicable to those NS and VNF instances).
•
Policy management and evaluation for the Network Service instances and VNF instances (e.g. policies related
with affinity/anti-affinity, scaling, fault and performance, geography, regulatory rules, NS topology, etc.).
The NFVO uses the Resource Orchestration functionality to provide services that support accessing NFVI resources in
an abstracted manner independently of any VIMs, as well as governance of VNF instances sharing resources of the
NFVI infrastructure.
ETSI
25
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
The following list expresses the non-exhaustive set of functionalities performed by the Resource Orchestration function.
These functionalities may be exposed by means of interfaces and consumed by NFV-MANO functional blocks or by
authorised external entities:
•
Validation and authorization of NFVI resource requests from VNF Manager(s), as those may impact the way
the requested resources are allocated within one NFVI-PoP or across multiple NFVI-PoPs (granting of the
requested resources is governed by policies, and may require prior reservation).
•
NFVI resource management across operator's Infrastructure Domains including the distribution, reservation
and allocation of NFVI resources to Network Service instances and VNF instances by using an NFVI
resources repository, as well as locating and/or accessing one or more VIMs as needed and providing the
location of the appropriate VIM to the VNFM, when required.
•
Supporting the management of the relationship between the VNF instances and the NFVI resources allocated
to those VNF instances by using NFVI Resources repository and information received from the VIMs.
•
Policy management and enforcement for the Network Service instances and VNF instances (e.g. NFVI
resources access control, reservation and/or allocation policies, placement optimization based on affinity
and/or anti-affinity rules as well as geography and/or regulatory rules, resource usage, etc.).
•
Collect usage information of NFVI resources by VNF instances or groups of VNF instances, for example, by
collecting information about the quantity of NFVI resources consumed via NFVI interfaces and then
correlating NFVI usage records to VNF instances.
5.4.2
VNF manager (VNFM)
The VNF Manager is responsible for the lifecycle management of VNF instances as described in clause 4.3. Each VNF
instance is assumed to have an associated VNF Manager. A VNF manager may be assigned the management of a single
VNF instance, or the management of multiple VNF instances of the same type or of different types.
Most of the VNF Manager functions are assumed to be generic common functions applicable to any type of VNF.
However, the NFV-MANO architectural framework needs to also support cases where VNF instances need specific
functionality for their lifecycle management, and such functionality may be specified in the VNF Package.
The following list expresses the non-exhaustive set of functions performed by the VNF Manager function. These
functionalities may be exposed by means of interfaces and consumed by other NFV-MANO functional blocks or by
authorised external entities:
•
VNF instantiation, including VNF configuration if required by the VNF deployment template (e.g. VNF initial
configuration with IP addresses before completion of the VNF instantiation operation).
•
VNF instantiation feasibility checking, if required.
•
VNF instance software update/upgrade.
•
VNF instance modification.
•
VNF instance scaling out/in and up/down.
•
VNF instance-related collection of NFVI performance measurement results and faults/events information, and
correlation to VNF instance-related events/faults.
•
VNF instance assisted or automated healing.
•
VNF instance termination.
•
VNF lifecycle management change notification.
•
Management of the integrity of the VNF instance through its lifecycle.
•
Overall coordination and adaptation role for configuration and event reporting between the VIM and the EM.
ETSI
26
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
The deployment and operational behaviour of each VNF is captured in a template called Virtualised Network Function
Descriptor (VNFD) that is stored in the VNF catalogue. NFV-MANO uses a VNFD to create instances of the VNF it
represents, and to manage the lifecycle of those instances. A VNFD has a one-to-one correspondence with a VNF
Package, and it fully describes the attributes and requirements necessary to realize such a VNF. NFVI resources are
assigned to a VNF based on the requirements captured in the VNFD (containing resource allocation criteria, among
others), but also taking into consideration specific requirements, constraints, and policies that have been pre-provisioned
or are accompanying the request for instantiation and may override certain requirements in the VNFD (e.g. operator
policies, geo-location placement, affinity/anti-affinity rules, local regulations).
The information elements to be handled by the NFV-MANO, including the VNFD among others, need to guarantee the
flexible deployment and portability of VNF instances on multi-vendor and diverse NFVI environments, e.g. with
diverse computing resource generations, diverse virtual network technologies, etc. To achieve this, hardware resources
need to be properly abstracted and the VNF requirements be described in terms of such abstraction.
The VNFM has access to a repository of available VNF Packages and different versions of them, all represented via
their associated VNFDs. Different versions of a VNF Package may correspond to different implementations of the same
function, different versions to run in different execution environments (e.g. on different hypervisors, dependent on
NFVI resources availability information, etc.), or different release versions of the same software. The repository may be
maintained by the NFVO or another external entity.
5.4.3
Virtualised infrastructure manager (VIM)
The Virtualised Infrastructure Manager (VIM) is responsible for controlling and managing the NFVI compute, storage
and network resources, usually within one operator's Infrastructure Domain (e.g. all resources within an NFVI-PoP,
resources across multiple NFVI-POPs, or a subset of resources within an NFVI-PoP) as described in clause 4.2. A VIM
may be specialized in handling a certain type of NFVI resource (e.g. compute-only, storage-only, networking-only), or
may be capable of managing multiple types of NFVI resources (e.g. in NFVI-Nodes).
EXAMPLE:
A VIM exposes northbound open interfaces that support management of NFVI virtualised
compute, storage, and networking resources.
The southbound interfaces interface with a variety of hypervisors and Network Controllers in order to perform the
functionality exposed through its northbound interfaces. Other VIM implementations may directly expose the interfaces
exposed by such compute, storage, Network Controllers as specialized VIMs. A particular example of a specialized
VIM is a WAN Infrastructure Manager (WIM), typically used to establish connectivity between PNF endpoints in
different NFVI-PoPs. The WIM and the overall topic of network connectivity management are described in clause 5.6.
The VIM implementation is out of scope for NFV-MANO; however the interfaces exposed by VIMs are in-scope.
The following list expresses the set functions performed by the VIM. These functionalities may be exposed by means of
interfaces consumed by other NFV-MANO functional blocks or by authorised external entities:
•
Orchestrating the allocation/upgrade/release/reclamation of NFVI resources(including the optimization of such
resources usage), and managing the association of the virtualised resources to the physical compute, storage,
networking resources. Therefore, the VIM keeps an inventory of the allocation of virtual resources to physical
resources, e.g. to a server pool (ETSI GS NFV-INF 004 [i.6]).
•
Supporting the management of VNF Forwarding Graphs (create, query, update, delete), e.g. by creating and
maintaining Virtual Links, virtual networks, sub-nets, and ports, as well as the management of security group
policies to ensure network/traffic access control (ETSI GS NFV-INF 005 [i.7]).
•
Managing in a repository inventory related information of NFVI hardware resources (compute, storage,
networking) and software resources (e.g. hypervisors), and discovery of the capabilities and features
(e.g. related to usage optimization) of such resources.
NOTE 1: Inventory repositories of NFVI hardware and software resources managed by the VIM are not explicitly
shown in the NFV-MANO architectural framework.
•
Management of the virtualised resource capacity (e.g. density of virtualised resources to physical resources),
and forwarding of information related to NFVI resources capacity and usage reporting.
ETSI
27
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
•
Management of software images (add, delete, update, query, copy) as requested by other NFV-MANO
functional blocks (e.g. NFVO). While not explicitly shown in the NFV-MANO architectural framework, the
VIM maintains repositories for software images, in order to streamline the allocation of virtualised computing
resources. A validation step, performed by VIM, is required for software images before storing the image
(e.g. VNF package on-boarding and update). Image validation operations during run-time, e.g. during
instantiation or scaling, are outside the scope of the current version of the present document.
•
Collection of performance and fault information (e.g. via notifications) of hardware resources (compute,
storage, and networking) software resources (e.g. hypervisors), and virtualised resources (e.g. VMs); and
forwarding of performance measurement results and faults/events information relative to virtualised resources.
•
Management of catalogues of virtualised resources that can be consumed from the NFVI. The elements in the
catalogue may be in the form of virtualised resource configurations (virtual CPU configurations, types of
network connectivity (e.g. L2, L3), etc.), and/or templates (e.g. a virtual machine with 2 virtual CPUs and
2 GB of virtual memory).
NOTE 2: The virtualised resources catalogues are not explicitly shown in the NFV-MANO architectural
framework.
5.4.4
NS Catalogue
The NS Catalogue represents the repository of all of the on-boarded Network Services, supporting the creation and
management of the NS deployment templates (Network Service Descriptor (NSD), Virtual Link Descriptor (VLD), and
VNF Forwarding Graph Descriptor (VNFFGD) via interface operations exposed by the NFVO. Clause 6 describes in
detail the type of information elements captured in the mentioned descriptors.
5.4.5
VNF Catalogue
The VNF Catalogue represents the repository of all of the on-boarded VNF Packages, supporting the creation and
management of the VNF Package (VNF Descriptor (VNFD), software images, manifest files, etc.) via interface
operations exposed by the NFVO. Both NFVO and VNFM can query the VNF Catalogue for finding and retrieving a
VNFD, to support different operations (e.g. validation, checking instantiation feasibility). Clause 6 describes in detail
the type of information elements captured in the VNFD.
5.4.6
NFV Instances repository
The NFV Instances repository holds information of all VNF instances and Network Service instances. Each VNF
instance is represented by a VNF record, and each NS instance is represented by an NS record. Those records are
updated during the lifecycle of the respective instances, reflecting changes resulting from execution of NS lifecycle
management operations and/or VNF lifecycle management operations. This supports NFVO's and VNFM's
responsibilities in maintaining the integrity and visibility of the NS instances, respectively VNF instances, and the
relationship between them. Clause 6 describes in detail the type of information elements captured in the NS record and
VNF record.
5.4.7
NFVI Resources repository
The NFVI Resources repository holds information about available/reserved/allocated NFVI resources as abstracted by
the VIM across operator's Infrastructure Domains, thus supporting information useful for resources reservation,
allocation and monitoring purposes. As such, the NFVI Resources repository plays an important role in supporting
NFVO's Resource Orchestration and governance role, by allowing NFVI reserved/allocated resources to be tracked
against the NS and VNF instances associated with those resources (e.g. number of VMs used by a certain VNF instance
at any time during its lifecycle).
5.5
Other functional blocks
The functional blocks described in this clause are not considered part of NFV-MANO, but are described because they
exchange information with NFV-MANO functional blocks.
ETSI
28
5.5.1
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Element management (EM)
The Element Management is responsible for FCAPS management functionality for a VNF. This includes:
•
Configuration for the network functions provided by the VNF.
•
Fault management for the network functions provided by the VNF.
•
Accounting for the usage of VNF functions.
•
Collecting performance measurement results for the functions provided by the VNF.
•
Security management for the VNF functions.
The EM may be aware of virtualisation and collaborate with the VNF Manager to perform those functions that require
exchanges of information regarding the NFVI Resources associated with the VNF.
5.5.2
Operations Support System/Business Support System (OSS/BSS)
The OSS/BSS are the combination of the operator's other operations and business support functions that are not
otherwise explicitly captured in the present architectural framework, but are expected to have information exchanges
with functional blocks in the NFV-MANO architectural framework. OSS/BSS functions may provide management and
orchestration of legacy systems and may have full end to end visibility of services provided by legacy network
functions in an operator's network.
5.5.3
Network Functions Virtualisation Infrastructure (NFVI)
Within the scope of the present document, the NFVI encompasses all the hardware (e.g. compute, storage, and
networking) and software (e.g. hypervisors) components that together provide the infrastructure resources where VNFs
are deployed.
The NFVI may also include partially virtualised NFs. Examples of such partially virtualised network functions are
related to "white box" switches, hardware load balancers, DSL Access Multiplexers (DSLAMs), Broadband Remote
Access Server (BRAS), Wi-Fi access points, CPEs, etc., for which a certain part of the functionality is virtualised and is
in scope of NFV-MANO while other parts are built in silicon (PNF) either due to physical constraints (e.g. digital
interfaces to analogue physical channels) or vendor design choices. The present document does not cover the
management of PNFs and it is assumed here that it is being taken care of by some other entity, for example the
OSS/BSS or an Network Controller. The present document covers the configuration of the connectivity between
adjacent VNF(s) and PNF(s) that are comprised in the same Network Service.
5.6
Network Controllers
5.6.1
Overview
This clause describes the roles and responsibilities of Network Controllers, the VIM and the NFVO specifically
regarding the orchestration of the intra-NFVI-PoP and inter-NFVI-PoP connectivity between components of a given
VNF or various VNFs in a VNF Forwarding Graph as well as connectivity to the PNFs.
The Network Controllers, when present, provide intuitive programmatic interfaces along the lines of the network
interfaces called out in clause 7 and thus can be best categorised as an abstraction layer below the VIM for a given
NFVI-PoP. Each NFVI-PoP or Administrative Domain may include a Network Controller (e.g. SDN controller)
responsible for providing the programmable network interfaces that enable the establishment of connectivity within the
domain.
There are various types of Network Controllers categorized by:
•
the type of network which it is virtualising (NFVI-PoP Network Controller vs. WAN Network Controller,
etc.);
•
how the virtualisation is performed (overlays vs. partitioning);
ETSI
29
•
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
the underlying technology and its southbound interfaces.
A given NFVO Resource Orchestration function may consume "Network as a Service" interfaces being exposed by
multiple VIMs to provision end-to-end virtual networks. The virtual network is a service provided by the NFVI. Virtual
networks establish connectivity between components of a VNF or are used to implement a VNF Forwarding Graph
(VNFFG). They have well defined service characteristics.
In order to facilitate VIMs support of multiple Network Controllers from multiple vendors in their Administrative
Domain (via southbound interfaces), it is necessary to develop well-defined interfaces between VIMs and Network
Controllers.
5.6.2
NFVO, VIM and Network Controllers
Network Controllers may form a hierarchy in a client/server relationship where each "server" Network Controller
exposes an interface to request connectivity services, i.e. virtual networks, to its clients. At the lowest layer, Network
Controllers have visibility into the devices they control directly. At the highest layer, Network Controllers provide
connectivity services to an application and provide abstraction of the underlying resources. Each layer in the hierarchy
provides a different level of abstraction and may establish connectivity services by configuring the forwarding tables of
the Network Functions within its domain directly or by requesting connectivity from "server" Network Controllers or a
combination of both.
Figure 5.2 shows a hybrid network environment example illustrating the goal of NFV to have fully programmatic open
interfaces for service and Resource Orchestration within and across NFVI-PoPs. The figure shows:
•
An example NFVI network that provides the necessary connectivity services to establish an end-to-end service
between Endpoint 1 and Endpoint 2 through VNF 1 and VNF 2 as shown by the VNFFG representation.
•
A model where the WAN between NFVI-PoPs provides Virtual Links based on network connectivity services
between NFVI-PoPs, and where the NFVO may request Virtual Links based on network connectivity services
through the Or-Vi reference point. This requires Resource Orchestration, an NFVO function, to have an
abstracted view of the network topology and interfaces to underlying VIMs and WAN Infrastructure Manager
(WIM) to request connectivity services.
•
One VIM per NFVI-PoP but this is not mandated as there could be multiple VIMs in a single NFVI-PoP, each
responsible for an Administrative Domain. Similarly, a single VIM could be responsible for multiple NFVIPoPs. There could be multiple WIMs. Each VIM/WIM provides connectivity services within its domain and
provides an abstraction of the underlying resources through a well-defined interface.
ETSI
30
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Or-Vnfm
Or-Vi
Or-Vi
Or-Vi
Figure 5.2: Network Controller example
For this example, multiple connectivity services are requested by the Resource Orchestration component of the NFVO:
•
Virtual network between gateway in NFVI-PoP 1 and VNF1.
•
Virtual network between gateway in NFVI-PoP 2 and VNF2.
•
Virtual network between gateway in NFVI-PoP 1 and gateway in NFVI-PoP 2.
•
Physical connectivity service between the physical Endpoint 1 and a gateway in NFVI-PoP 1.
•
Physical connectivity service between the physical Endpoint 2 and a gateway in NFVI-PoP 2.
This requires each NFVI-PoP VIM to establish a virtual network within its NFVI-PoP, the WIM to establish a virtual
network between the NFVI-PoP 1 gateway and the NFVI-PoP 2 gateway, the WIM to establish connectivity between
the PNF endpoint1 and NFVI-PoP 1 gateway, and the NFVI-PoP 2 VIM to establish connectivity between PNF
endpoint 2 and the NFVI-PoP 2 gateway.
A VIM and/or a WIM can either:
•
Interface with the underlying Network Controllers to request virtual connectivity services. In such a case, the
VIMs/WIMs interface with the Network Controllers over the Nf-Vi reference point. There could be multiple
Network Controllers under a VIM (for example, if different virtual network partitioning techniques are used
within the domain); in this case, the VIM is responsible to request virtual networks from each underlying
Network Controller and setup the interworking function between them. This acknowledges the fact that there
might be existing Network Controllers in the NFVI already deployed and in use for connectivity prior to the
instantiation of the VIM/NFVO.
•
Establish the connectivity services directly by configuring the forwarding tables of the underlying Network
Functions, hence becoming the Network Controller part of the VIM (although it is not explicitly shown in the
figure). In such a case, the VIM directly controls some elements, for example, a software switch (a.k.a.
vSwitch). This could also be the case if there are Network Functions that do not fall within the Administrative
Domain of a Network Controller.
ETSI
31
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
The result is a hierarchy of control and orchestration of network resources to provide the connectivity services, from the
bottom up: Network Function, Network Controllers, VIM/WIM and NFVO Resource Orchestration. Each level in the
hierarchy provides abstraction of resources, a northbound interface to request connectivity services and the connectivity
services themselves. The higher level Network Controllers have higher level abstractions while the lower level Network
Controllers have lower levels of abstraction.
The Resource Orchestration functions of the NFVO interfaces the VIM/WIMs through well-defined interfaces
comprised in the Or-Vi reference point for requesting creation and management of virtual networks.
5.6.3
Roles and responsibilities
In the example presented in figure 5.2, the Resource Orchestration function of the NFVO is responsible for the
following aspects related to NFVI connectivity services:
•
Global view of the network characteristics of the various logical links.
•
Orchestrate the VNFs required for a NS in a manner best suited to match the network constraints specified by
the Network Service Descriptor (NSD), see clause 6.
•
Orchestrate connectivity over a combination of PNFs and VNFs.
•
Monitor the abstracted end-to-end Network Service utilization aspects such as bandwidth, latency, etc.
between the various NFVI-PoPs.
•
Support and invoke the underlying VIM/WIM northbound interfaces to request individual connectivity
services to construct the end-to-end service. Example connectivity services include E-LINE, E-LAN and
E-TREE services as defined in ETSI GS NFV-INF 005 [i.7]. Connectivity services include internal
connectivity between components of a VNF making up each VNF, external connectivity between the VNFs in
a VNF Forwarding Graph and connectivity between the VNFs and PNFs.
The VIMs are responsible for the following aspects related to NFVI connectivity services over NFVI-POPs:
•
Provide a "Network as a Service" northbound interface to the higher layers, e.g. NFVO and VNF Manager.
•
Abstract the various southbound interfaces and overlay/network partitioning mechanisms exposed by the
underlying NFVI network.
•
Invoke the underlying NFVI network southbound interfaces, whether they are Network Controllers or Network
Functions, to construct the service within the domain.
The WIMs provide connectivity services between the NFVI-POPs and connectivity to Physical Network Functions.
This functional block may be added as a specialized VIM, in particular for new greenfield virtualised deployments.
Alternatively, a WIM can also be an existing component of the OSS/BSS functional block. The WIMs are responsible
for the following aspects related to the NFVI connectivity services:
•
Path computation based on quality assurance factors such as jitter, RTT, delay & bandwidth calendaring.
•
Establish connectivity over the physical network (e.g. set of MPLS tunnels).
•
Provide a northbound interface to the higher layers, e.g. NFVO, to provide connectivity services between
NFVI-PoPs or to physical network functions.
•
Invoke the underlying NFVI network southbound interfaces, whether they are Network Controllers or Network
Functions, to construct the service within the domain.
The Network Controller function is responsible for the following aspects related to NFVI connectivity services:
•
Resource management and tracking of network resources and attributes such as bandwidth, jitter, delay; etc.
•
Implement southbound interfaces to the compute and network resources providing the connectivity services to
create overlay tunnels (e.g. VXLAN, NVGRE, MPLS over GRE) or network partitions (for example using
OpenFlow).
•
Abstract the information exposed by the underlying NFVI network via various southbound interfaces.
ETSI
32
•
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Invoke the underlying NFVI network southbound interfaces, whether they are Network Controllers or Network
Functions, to construct the service within the domain.
5.6.4
NFVI network EMS/NMS vs. Network Controllers
While figure 5.2 illustrates an overall NFV goal, it is not expected that the first NFV deployments will use Network
Controllers with programmatic/open interfaces for all network segments. It is quite possible that the NFVI network will
have Network Controllers for some domains but not for others. Figure 5.3 shows a similar example to the one described
in clause 5.6.2 but illustrates an example starting point where parts of the NFVI require an EMS/NMS to configure the
resources. This may be the case when the resources are not reconfigured regularly, i.e. for static provisioning, or when a
Network Controller with programmatic interfaces is not available for these resources. In that case, the configuration is
either done manually or through a proprietary interface.
Os-Ma-nfvo
Or-Vnfm/Vi-Vnfm
Figure 5.3: EMS/NMS for the NFVI network
5.7
NFV-MANO reference points
This clause describes reference points between NFV-MANO functional blocks and between those and other functional
blocks.
5.7.1
Os-Ma-nfvo
This reference point is used for exchanges between OSS/BSS and NFV Orchestrator, and supports the following:
•
Network Service Descriptor and VNF package management.
•
Network Service instance lifecycle management:
-
Network Service instantiation.
ETSI
33
•
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
-
Network Service instance update (e.g. update a VNF instance that is comprised in the Network Service
instance).
-
Network Service instance query (e.g. retrieving summarised information about NFVI resources
associated to the Network Service instance, or to a VNF instance within the Network Service instance).
-
Network Service instance scaling.
-
Network Service instance termination.
VNF lifecycle management:
-
For VNF lifecycle management, the NFV Orchestrator identifies the VNF Manager and forwards such
requests (see Or-Vnfm description).
•
Policy management and/or enforcement for Network Service instances, VNF instances and NFVI resources
(for authorization/access control, resource reservation/placement/allocation, etc.).
•
Querying relevant Network Service instance and VNF instance information from the OSS/BSS.
•
Forwarding of events, accounting and usage records and performance measurement results regarding Network
Service instances, VNF instances, and NFVI resources to OSS/BSS, as well as and information about the
associations between those instances and NFVI resources, e.g. number of VMs used by a certain VNF
instance.
5.7.2
Ve-Vnfm-em
This reference point is used for exchanges between EM and VNF Manager, and supports the following:
•
VNF instantiation.
•
VNF instance query (e.g. retrieve any run-time information).
•
VNF instance update (e.g. update configuration).
•
VNF instance scaling out/in, and up/down.
•
VNF instance termination.
•
Forwarding of configuration and events from the EM to the VNFM.
•
Forwarding of configuration and events regarding the VNF from the VNFM to the EM.
NOTE:
5.7.3
This reference point is only used if the EM is aware of virtualisation.
Ve-Vnfm-vnf
This reference point is used for exchanges between VNF and VNF Manager, and supports the following:
•
VNF instantiation.
•
VNF instance query (e.g. retrieve any run-time information).
•
VNF instance update (e.g. update configuration).
•
VNF instance scaling out/in, and up/down.
•
VNF instance termination.
•
Forwarding of configuration and events from the VNF to the VNFM.
•
Forwarding of configuration, events, etc. regarding VNF, from the VNFM to the VNF.
•
Verification that the VNF is still alive/functional.
ETSI
34
5.7.4
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Nf-Vi
This reference point is used for exchanges between Virtualisation Infrastructure Manager and NFV Infrastructure, and
supports the following:
•
Aallocate VM with indication of compute/storage resource.
•
Update VM resources allocation.
•
Migrate VM.
•
Terminate VM.
•
Create connection between VMs.
•
Configure connection between VMs.
•
Remove connection between VMs.
•
Forwarding of configuration information, failure events, measurement results, and usage records regarding
NFVI (physical, software, and virtualised resources) to the VIM.
5.7.5
Or-Vnfm
This reference point is used for exchanges between NFV Orchestrator and VNF Manager, and supports the following:
•
NFVI resources authorization/validation/reservation/release for a VNF.
•
NFVI resources allocation/release request for a VNF.
•
VNF instantiation.
•
VNF instance query (e.g. retrieve any run-time information).
•
VNF instance update (e.g. update configuration).
•
VNF instance scaling out/in, and up/down.
•
VNF instance termination.
•
VNF package query.
•
Forwarding of events, other state information about the VNF that may impact also the Network Service
instance.
5.7.6
Or-Vi
This reference point is used for exchanges between the NFV Orchestrator and the VIM, and supports the following:
•
NFVI resource reservation/release.
•
NFVI resource allocation/release/update.
•
VNF software image addition/deletion/update.
•
Forwarding of configuration information, events, measurement results, usage records regarding NFVI
resources to the NFV Orchestrator.
ETSI
35
5.7.7
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Vi-Vnfm
This reference point is used for exchanges between the VNF Manager and the Virtualised Infrastructure Manager, and
supports the following:
•
NFVI resources reservation information retrieval.
•
NFVI resources allocation/release.
•
Exchanges of configuration information between reference point peers, and forwarding to the VNF Manager
such information for which the VNFM has subscribed to (e.g. events, measurement results, and usage records
regarding NFVI resources used by a VNF).
5.8
Interfaces description approach
5.8.1
Overview
This clause describes the interface design approach in NFV-MANO architectural framework. The detailed description
of the interfaces is the subject of clause 7.
While reference points are a good way to identify peer-to-peer relationships between functional blocks, descriptions of
the interfaces provide a deeper understanding of how capabilities provided by "producer" functional blocks are exposed
to "consumer" functional blocks. This approach allows to:
•
Facilitate analysis of requirements against existing standardized interfaces, leading to identification of gaps
and subsequent recommendations on how to address them.
•
Lead to exposure of the appropriate level of abstraction via the interfaces, in such a way that it:
-
Provides guidance to bring consistency among existing and future interfaces to be standardized.
-
Provides the functional definitions, without limiting implementation choices of the functional blocks.
-
Emphasizes contract-based and producer-consumer interface paradigm to support the dynamic nature of
the interfaces among functional blocks, without limiting the design and implementation options.
-
Illustrates that the same function may be provided by a functional block to several other functional
blocks without restricting the communication paths between functional blocks, within the boundaries of
the NFV architectural framework.
-
Illustrates that different functional blocks can expose the same functionality in an abstracted manner to
support different implementation options.
•
Help define functional blocks' capabilities in such a way that orchestration of different entities is provided in
multiple functional blocks, if appropriate; those functional blocks should offer abstracted services to other
functional blocks.
•
Emphasise the value of designing granular functions that can be performed via atomic transactions using the
exposed interfaces.
•
Recommend the use of configuration and authorization policies to play a role in implementing the path that an
operational flow may take, when multiple alternatives are possible.
•
Enable configuration and re-configuration of the levels of granularity, frequency and types of information
being collected for fault, performance management and event data at runtime.
ETSI
36
5.8.2
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Producer-consumer paradigm
A functional block may play the role of a producer, or the role of a consumer or both. Interfaces represent a published
contract offered by a producer functional block, exposing externally the functions produced. They can be consumed by
any other consumer functional block that is authenticated and authorized to access and consume the functions offered;
authentication mechanisms and authorization policies are defined to control access to and consumption of functions
produced by a functional block, to the level of granularity desired by the owner of the producer functional block. A
producer functional block may offer different functions, and may offer different levels of granularity within each
function; authorization policies should reflect such granularity because different access and consumption rights may be
granted to different consumer functional blocks.
A consumer functional block may also have a choice of sending requests to several producer functional blocks offering
the same function; in some cases the function offered may be completely identical, in other cases one producer
functional block may offer the function with optional extensions. Depending on implementation choices, configuration
policies may be used to determine which producer functional block may be invoked by a consumer functional block to
provide the desired function.
For notification interfaces, the general paradigm used is a publish/subscribe where the publishers issue notifications
through the interface without knowledge of what if any subscribers there may be. Subscribers using the notification
interface express interest in some notifications or groups of notifications and only receive those of interest. For a
notification interface, the publisher plays the producer role and the subscriber the consumer role.
Between the choices dictated by configuration policies available at the consuming functional block, the implementation
choices of manufacturers offering the functions and the choices dictated by authorization policies at the producing
functional block, the ability to support some preferred operational flows, and disallow other operational flows in the
service provider's environment becomes straightforward.
Functional Block X
(consumes A)
Functional Block Y
(consumes A&B)
1a.1
Configuration
policy
determines
the request
path
i/f A
1b
i/f B
Functional Block Z
(consumes A&C,
produces A&B)
1a.2
i/f A
Authorization
policy
determines
the access
rights
Authorization
policy
determines
the access
rights
i/f C
Functional Block V
(produces A&C)
Figure 5.4: Interface concept in a Producer-Consumer paradigm
Figure 5.4 illustrates the notion of producer and consumer functional blocks, and the interfaces that the producer
functional blocks expose, through which the consumer functional blocks consumes the functions provided. The
following are noticeable:
•
The same interface may be exposed by different producer functional blocks (e.g. interface A). It is also
possible (not shown) that functional block Z may expose interface A with additional extensions or additional
behaviours (e.g. which can be provided by producer functional block Z, but perhaps not by producer functional
block V).
•
A producer functional block may serve multiple consumer functional blocks via the same interface
(e.g. producer functional blocks Z and V can serve multiple consumers via interface A).
ETSI
37
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
•
A functional block can be at the same time a producer and a consumer of the same interface (e.g. functional
block Z both produces and consumes interface A, or of different interfaces (e.g. functional block Z produces
interface A and consumes interface C).
•
The design supports multiple flow alternatives that can be directed by the use of policies. In the figure,
consumer functional block X may be configured to request a certain function from either producer functional
block Z (via flow 1a) or from producer functional block V (via flow 1b). Access to either of the functions is
controlled by authorization policies associated with interface A on each of producer functional blocks Z and V.
In operational flow 1.a, producer functional block Z processes the request and adds value (it could offer
functional extensions or additional behaviour (e.g. implementing some global policies) via interface A, or it
could provide additional services based on other information it has access to, e.g. policies, which may not be
available to producer functional block V) before sub-contracting the function to producer functional block V
for completion.
•
Each producer functional block in NFV-MANO may maintain a set of policy enforcement rules. These rules
may enumerate the list of other entities that are authorised to consume and perform specific operations over an
interface; additionally, these policies may be defined on a per VNF basis. The actual development and
implementation of policy rules are highly implementation specific and out of scope for the present document.
6
NFV management and orchestration information
elements
This clause describes relevant information elements for the on-boarding and lifecycle management of VNF and NS.
6.1
Introduction
The Network Service describes the relationship between VNFs and possibly PNFs that it contains and the links needed
to connect VNFs that are implemented in the NFVI network. Links are also used to interconnect the VNFs to PNFs and
endpoints. Endpoints provide an interface to the existing network, including the possibility of incorporating Physical
Network Functions to facilitate evolution of the network.
Figure 6.1: Top-level information elements
This clause describes several categories of information:
•
Information that resides in descriptors. These are deployment templates that contain relatively static
information used in the process of on-boarding VNFs and NSs.
ETSI
38
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
•
Information that resides in records. These contain relatively dynamic run-time data representing e.g. VNF or
NS instances; this data is maintained throughout the lifetime of the instance.
•
Information that may be passed between functional blocks at these interfaces.
The information is structured into information elements. An information element contains specific information. It may
contain a single value only, but may also contain further information elements. Hence, information elements can build a
tree structure specifying a containment tree of information elements. The information element can be classified into
three different types: leafs, reference elements, and sub elements which are basically an element itself. Each of these
information elements has a unique name along the whole path in the tree that lead to that element from the root of the
tree.
A leaf is a single information element that specifies a value within the scope of the present document. The data type of
the value is dependent on the information it should carry. The number of occurrences of the same leaf element inside its
parent element is specified by the cardinality of the element.
A reference element is an information element that carries a reference to another information element. This may be
represented by an URI but depends on the concrete implementation of the information model. The number of
occurrences of the same reference element inside its parent element is specified by the cardinality of the element.
A sub element is an information element itself that specifies another level in the tree. The number of occurrences of the
same sub element inside its parent element is specified by the cardinality of the element.
As explained above each of these elements has a name that is unique among the branches of a tree. The name of the
element is represented as part of the clause headers of the clauses below. The clauses itself contain tables with four
columns. The first column contains the name of the information element, the second one gives its type. The third
column gives the cardinality of the information element. If the cardinality is a positive integer x then the element occurs
exactly x times. If the cardinality is given as a range then this number of occurrences shall not exceed that range. A
range starting with "0" implies that the element may be omitted. The last column describes the content of the
information element. If the information element is a sub element the elements contained can be found by a subsequent
section whose heading is the name of this information element. For some elements and additional example column is
provided explaining the meaning of the element by an example.
To describe a Network Service and the components comprising the Network Service, information elements representing
these components are introduced. There are four information elements defined apart from the top level Network Service
(NS) information element:
•
Virtualised Network Function (VNF) information element.
•
Physical Network Function (PNF) information element.
•
Virtual Link (VL) information element.
•
VNF Forwarding Graph (VNFFG) information element.
A set of VLs in a Network Service form a Network Connection Topology (NCT) as described in
ETSI GS NFV-SWA 001 [i.8]. A VNFFG can reference other information elements in the Network Service such as
PNFs, VLs and VNFs. A VNFFG also contains a Network Forwarding Path (NFP) element. Figure 6.1 depicts the high
level structure of the information elements. NS, VNF and PNF information elements include Connection Point
descriptions.
The information elements can be used in two different contexts: as descriptors in a catalogue or template context or as
instance records in a runtime context. Figure 6.2 depicts the usage of the information elements in different contexts.
A Network Service Descriptor (NSD) is a deployment template for a Network Service referencing all other descriptors
which describe components that are part of that Network Service.
A VNF Descriptor (VNFD) is a deployment template which describes a VNF in terms of its deployment and operational
behaviour requirements. It is primarily used by the VNFM in the process of VNF instantiation and lifecycle
management of a VNF instance. The information provided in the VNFD is also used by the NFVO to manage and
orchestrate Network Services and virtualised resources on the NFVI. The VNFD also contains connectivity, interface
and KPIs requirements that can be used by NFV-MANO functional blocks to establish appropriate Virtual Links within
the NFVI between its VNFC instances, or between a VNF instance and the endpoint interface to the other Network
Functions.
ETSI
39
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
A VNF Forwarding Graph Descriptor (VNFFGD) is a deployment template which describes a topology of the Network
Service or a portion of the Network Service, by referencing VNFs and PNFs and Virtual Links that connect them.
A Virtual Link Descriptor (VLD) is a deployment template which describes the resource requirements that are needed
for a link between VNFs, PNFs and endpoints of the Network Service, which could be met by various link options that
are available in the NFVI. The NFVO can select an option following consultation of the VNFFG to determine the
appropriate NFVI to be used based on functional (e.g. dual separate paths for resilience) and other needs
(e.g. geography and regulatory requirements).
A Physical Network Function Descriptor (PNFD) describes the connectivity, Interface and KPIs requirements of Virtual
Links to an attached Physical Network Function. This is needed if a physical device is incorporated in a Network
Service to facilitate network evolution.
The NFVO on-boards all descriptors. NSD, VNFFGD, and VLD are "on-boarded" into a NS Catalogue; VNFD is
on-boarded in a VNF Catalogue, as part of a VNF Package. The present document does not specify how or where
PNFD is on-boarded; it is assumed that this is out of scope since that information is not specific to NFV, and can be
readily made available in current deployments. The integrity and authenticity of NSD, VNFFGD, VLD, VNFD and
PNFD are validated during on-boarding procedure. Multiple versions of a descriptor may exist in a catalogue, and the
NFV Orchestrator may be required to update the deployed network configuration to match an updated version of the
descriptor. The triggering of updates would be subject to business policies outside the scope of the present document.
On instantiation, the NFV Orchestrator or VNFM receive instantiation parameters from the entity initiating the
instantiation operation. Instantiation input parameters are used to customize a specific instantiation of a NS or VNF.
Instantiation input parameters contain information that identifies a deployment flavour to be used and may refer to
existing instances of VNF/PNF, which shall be incorporated in the instantiation process.
As a result of the instantiation operation records are created to represent the newly created instances. The Network
Service Record (NSR), the VNF Record (VNFR), the Virtual Link Record (VLR) and the VNFFG Record (VNFFGR)
are created based on the information given in the descriptors together with additional runtime information related to the
component instances. A PNF Record (PNFR) represents an instance related to a pre-existing PNF which is part of a
Network Service and contains a set of runtime attributes regarding PNF information, including connectivity relevant to
the NFVO. Figure 6.2 gives an overview how descriptors and records relate to each other.
The NSR, VNFR, VNFFGR and VLR information elements provide a collection of data items that may be needed to
model the state of instances of NS, VNF, VNFFG or VL, respectively; they are a result of combining NSD, VNFD,
VNFFGD and VLD information elements with input/output parameters exchanged via different interfaces produced
and/or consumed by NFV-MANO functional blocks, and do not imply any particular implementation.
ETSI
40
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Figure 6.2: Information elements in different context
6.2
Network Service information elements
The Network Service Descriptor (NSD) consists of static information elements as defined below. It is used by the NFV
Orchestrator to instantiate a Network Service, which would be formed by one or more VNF Forwarding Graphs, VNFs,
PNFs and VLs. The NSD also describes deployment flavours of Network Service.
The following is not a complete list of information elements constituting the NSD but a minimum subset needed to onboard the Network Service (NS).
ETSI
41
6.2.1
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Network Service Descriptor (nsd)
6.2.1.1
nsd base element
Identifier
Id
vendor
version
vnfd
Type
Leaf
Leaf
Leaf
Reference
Cardinality
1
1
1
1...N
vnffgd
Reference
0...N
vld
Reference
0…N
lifecycle_event
Leaf
0…N
vnf_dependency
Leaf
0…N
monitoring_
parameter
Leaf
0...N
Description
ID of this Network Service Descriptor.
Provider or vendor of the Network Service.
Version of the Network Service Descriptor.
VNF which is part of the Network Service, see clause 6.3.1.
This element is required, for example, when the Network
Service is being built top-down or instantiating the member
VNFs as well.
VNFFG which is part of the Network Service, see clause 6.5.1.
A Network Service might have multiple graphs, for example, for:
1. Control plane traffic.
2. Management-plane traffic.
3. User plane traffic itself could have multiple NFPs based
on the QOS etc. The traffic is steered amongst 1 of these
NFPs based on the policy decisions.
Virtual Link which is part of the Network Service, see
clause 6.4.1.
Defines NS functional scripts/workflows for specific lifecycle
events (e.g. initialization, termination, scaling).
Describe dependencies between VNF. Defined in terms of
source and target VNF i.e. target VNF "depends on" source
VNF. In other words a source VNF shall exist and connect to
the service before target VNF can be initiated/deployed and
connected. This element would be used, for example, to define
the sequence in which various numbered network nodes and
links within a VNF FG should be instantiated by the NFV
Orchestrator.
Represents a monitoring parameter which can be tracked for
this NS.
These can be network service metrics that are tracked for the
purpose of meeting the network service availability contributing
to SLAs (e.g. NS downtime).
These can also be used for specifying different deployment
flavours for the Network Service in Network Service Descriptor,
and/or to indicate different levels of network service availability.
Examples include specific parameters such as calls-per-second
(cps), number-of-subscribers, no-of-rules, flows-per-second,
etc.
1 or more of these parameters could be influential in
determining the need to scale-out.
ETSI
42
Identifier
service_
deployment_
flavour
Type
Element
auto_scale_policy
Leaf
connection_point
Element
pnfd
nsd_security
Reference
Leaf
6.2.1.2
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Cardinality
Description
1...N
Represents the service KPI parameters and its requirement for
each deployment flavour of the NS being described,
see clause 6.2.1.3. For example, there could be a flavour
describing the requirements to support a vEPC with 300k calls
per second. There could be another flavour describing the
requirements to support a vEPC with 500k calls persecond.
0...N
Represents the policy meta data, which may include the criteria
parameter & action-type. The criteria parameter should be a
supported assurance parameter.
An example of such a descriptor could be:
•
Criteria parameter: calls-per-second.
•
Action-type: scale-out to a different flavour ID.
1…N
This element describes a Connection Point which acts as an
endpoint of the Network Service, see clause 6.2.1.2.
This can, for example, be referenced by other elements as an
endpoint.
0...N
PNFs which are part of the Network Service, see clause 6.6.1.
0...1
This is a signature of nsd to prevent tampering. The particular
hash algorithm used to compute the signature, together with the
corresponding cryptographic certificate to validate the signature
should also be included.
Connection Point (nsd:connection_point)
Identifier
id
type
6.2.1.3
Type
Leaf
Leaf
Cardinality
1
1
Description
ID of the Connection Point.
This may be for example a virtual port, a virtual NIC
address, a physical port, a physical NIC address or the
endpoint of an IP VPN enabling network connectivity.
Service deployment flavour (nsd:service_deployment_flavour)
6.2.1.3.1
Base element
Identifier
id
flavour_key
Type
Leaf
Leaf
constituent_vnf
Element
Cardinality
Description
1
ID of the deployment flavour.
1
Assurance parameter against which this flavour is being
described.
The key could be a combination of multiple assurance
parameters with a logical relationship between them.
The parameters should be present as a monitoring_parameter
supported in clause 6.2.1.1.
For example, a flavour of a virtual EPC could be described in
terms of the assurance parameter "calls per second" (cps).
1…N
Represents the characteristics of a constituent flavour element,
see clause 6.2.1.3.2.
ETSI
43
6.2.1.3.2
Constituent VNF (nsd:service_deployment_flavour:constituent_vnf)
Identifier
vnf_reference
Type
Reference
Cardinality
1
vnf_flavour_id_
reference
Reference
1
redundancy_model
Leaf
0...1
affinity
Leaf
0...1
capability
Leaf
0...1
number_of_instances
Leaf
1
6.2.2
Description
Reference to a VNFD declared as vnfd in the Network
Service via vnf:id.
References a VNF flavour
(vnfd:deployment_flavour:id) to be used for this
service flavour, see clause 6.2.1.3.1.
Represents the redundancy of instances, for example,
"active" or "standby".
Specifies the placement policy between this instance
and other instances, if any.
Represents the capabilities of the VNF instances.
An example of capability is instance capacity
(e.g. capability = 50 %* NS capacity).
Number of VNF instances satisfying this service
assurance.
For a Gold flavour of the vEPC Network Service that
needs to satisfy an assurance of 96K cps, 2 instances of
the vMME VNFs will be required.
Network Service Record (nsr)
6.2.2.1
nsr base element
Identifier
id
auto_scale_policy
connection_point
monitoring_parameter
service_deployment_flavour
Type
Leaf
Leaf
Element
Leaf
Reference
Cardinality
1
0...N
1...N
1...N
1
vendor
version
vlr
vnfr
lifecycle_event
vnf_dependency
vnffgr
pnfr
descriptor_reference
Leaf
Leaf
Reference
Reference
Leaf
Leaf
Reference
Reference
Reference
1
1
0...N
1...N
0...N
0...N
0...N
1...N
1
resource_reservation
Leaf
0...N
runtime_policy_info
Leaf
0...N
status
notification
Leaf
Leaf
1
1...N
lifecycle_event_history
audit_log
Leaf
Leaf
0...N
0…N
6.2.2.2
Description
ID of the Network Service instance.
See clause 6.2.1.1.
See clause 6.2.1.2.
Monitoring parameter used in this instance.
References the
nsd:service_deployment_flavour used to
instantiate this Network Service instance.
See clause 6.2.1.1.
See clause 6.2.1.1.
See clause 6.4.2.
See clause 6.3.2.
See clause 6.2.1.1.
See clause 6.2.1.1.
See clause 6.5.2.
See clause 6.6.5.
The reference to the Network Service Descriptor used
to instantiate this Network Service.
Resource reservation information identification
(potentially per individual VIM) for NFVI resources
reserved for this NS instance.
Generic placeholder for input information related to NS
orchestration and management policies to be applied
during runtime of a specific NS instance (e.g. for NS
prioritization, etc.).
Flag to report status of the Network Service.
System that has registered to received notifications of
status changes.
Record of significant Network Service lifecycle events.
Record of detailed operational events.
Connection Point (nsr:connection_point)
Identifier
id
type
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Type
Leaf
Leaf
Cardinality
Description
1
ID of the Connection Point instance.
1
See clause 6.2.1.2.
ETSI
44
6.3
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Virtualised Network Function information elements
The following is not a complete list of information elements constituting the VNF information elements, but rather a
minimum representative subset considered necessary to on-board the VNF.
6.3.1
VNF Descriptor (vnfd)
A VNF Descriptor (VNFD) is a deployment template which describes a VNF in terms of deployment and operational
behaviour requirements. The VNFD also contains connectivity, interface and KPIs requirements that can be used by
NFV-MANO functional blocks to establish appropriate Virtual Links within the NFVI between VNFC instances, or
between a VNF instance and the endpoint interface to other Network Functions.
Connection Points of a VNF can be mapped to interfaces as described in ETSI GS NFV-SWA 001 [i.8].
6.3.1.1
vnfd base information elements
Identifier
Id
vendor
descriptor_version
version
Type
Leaf
Leaf
Leaf
Leaf
Cardinality
1
1
1
1
vdu
Element
1…N
virtual_link
Element
0...N
connection_point
Element
1...N
lifecycle_event
Leaf
0...N
dependency
Leaf
0…N
monitoring_parameter
Leaf
0...N
deployment_flavour
Element
1...N
auto_scale_policy
Leaf
0...N
Description
ID (e.g. name) of this VNFD.
The vendor generating this VNFD.
Version of the VNF Descriptor.
Version of VNF software, described by the descriptor
under consideration.
This describes a set of elements related to a particular
VDU, see clause 6.3.1.2.
Represents the type of network connectivity mandated
by the VNF vendor between two or more Connection
Points, see clause 6.3.1.3.
This element describes an external interface exposed
by this VNF enabling connection with a VL, see
clause 6.3.1.4 (see note).
Defines VNF functional scripts/workflows for specific
lifecycle events (e.g. initialization, termination, graceful
shutdown, scaling out/in, update/upgrade, VNF state
management related actions to support service
continuity).
Describe dependencies between VDUs. Defined in
terms of source and target VDU, i.e. target VDU
"depends on" source VDU. In other words sources VDU
shall exists before target VDU can be initiated/deployed.
Monitoring parameters, which can be tracked for this
VNF.
Can be used for specifying different deployment
flavours for the VNF in a VNFD, and/or to indicate
different levels of VNF service availability.
These parameters can be an aggregation of the
parameters at VDU level e.g. memory-consumption,
CPU-utilisation, bandwidth-consumption, etc.
They can be VNF specific as well such as calls-persecond (cps), number-of-subscribers, no-of-rules, flowsper-second, VNF downtime, etc.
One or more of these parameters could be influential in
determining the need to scale.
Represents the assurance parameter(s) and its
requirement for each deployment flavour of the VNF
being described, see clause 6.3.1.5.
Represents the policy meta data, which may include the
criteria parameter and action-type. The criteria
parameter should be a supported assurance parameter
(vnf:monitoring_parameter).
Example of such a descriptor could be:
calls-per-second.
•
Criteria parameter
scale-out to a different flavour ID, if
•
Action-type
exists.
ETSI
45
Identifier
manifest_file
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Cardinality
Description
0…1
The VNF package may contain a file that lists all files in
the package. This can be useful for auditing purposes
or for enabling some security features on the package.
manifest_file_security
Leaf
0…N
The manifest file may be created to contain a digest of
each file that it lists as part of the package. This digest
information can form the basis of a security mechanism
to ensure the contents of the package meet certain
security related properties.
If the manifest file contains digests of the files in the
package, then the manifest file should also note the
particular hash algorithm used to enable suitable
verification mechanisms. Examples of suitable hash
algorithms include, but are not limited to SHA-256,
SHA-384, SHA-512, and SHA-3.
In conjunction with an appropriate security signing
mechanism, which may include having a security
certificate as part of the VNF package, the digest
information can be used to help ensure the contents of
the VNF package have not been tampered with.
NOTE:
The connection between the VNF and the VL is expressed by the VLD referencing this Connection
Point. The Connection Point may also be attached to internal Virtual Vinks (vnfd:virtual_link:id).
6.3.1.2
Type
Leaf
Virtual Deployment Unit (vnfd:vdu)
Information elements concerning the VDU are defined in the table below. For structuring the table, it is split into
several parts with each clause below covering one of it.
6.3.1.2.1
vnfd:vdu base elements
id
Identifier
Type
Leaf
vm_image
Leaf
computation_requirement
Leaf
virtual_memory_resource_element
Leaf
virtual_network_bandwidth_resource Leaf
lifecycle_event
Leaf
constraint
high_availability
Leaf
Leaf
scale_in_out
Leaf
vnfc
Element
Cardinality
Description
1
A unique identifier of this VDU within the scope
of the VNFD, including version functional
description and other identification information.
This will be used to refer to VDU when defining
relationships between them.
0...1
This provides a reference to a VM image
(see note).
1
Describe the required computation resources
characteristics (e.g. processing power, number
of virtual CPUs, etc.), including Key Quality
Indicators (KQIs) for performance and
reliability/availability.
1
This represents the virtual memory needed for
the VDU.
1
This represents the requirements in terms of the
virtual network bandwidth needed for the VDU.
0...N
Defines VNF component functional
scripts/workflows for specific lifecycle events
(e.g. initialization, termination, graceful
shutdown, scaling out/in).
0...1
Placeholder for other constraints.
0...1
Defines redundancy model to ensure high
availability examples include:
•
ActiveActive: Implies that two instance of
the same VDU will co-exists with
continuous data synchronization.
•
ActivePassive: Implies that two instance of
the same VDU will co-exists without any
data synchronization.
0...1
Defines minimum and maximum number of
instances which can be created to support scale
out/in.
1...N
Contains information that is distinct for each
VNFC created based on this VDU.
ETSI
46
Identifier
monitoring_parameter
NOTE:
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Cardinality
Description
0...N
Monitoring parameter, which can be tracked for a
VNFC based on this VDU.
Examples include: memory-consumption,
CPU-utilisation, bandwidth-consumption, VNFC
downtime, etc.
A cardinality of zero allows for creating empty virtualisation containers as per (ETSI GS NFV-SWA 001 [i.8]).
6.3.1.2.1.1
Type
Leaf
VNFC (vnfd:vdu:vnfc)
id
Identifier
Type
Leaf
connection_point
Element
6.3.1.2.1.2
Connection Point (vnfd:vdu:vnfc:connection_point)
Identifier
id
virtual_link_reference
type
6.3.1.2.2
Cardinality
Description
1
Unique VNFC identification within the namespace of a
specific VNF.
1...N
Describes network connectivity between a VNFC
instance (based on this VDU) and an internal Virtual
Link.
Type
Cardinality
Description
Leaf
1
ID of the Connection Point.
Reference 1
References an internal Virtual Link
(vnfd:virtual_link:id, see clause 6.3.1.3) to
which other VNFCs can connect.
Leaf
1
This may be, for example, a virtual port, a virtual NIC
address, a physical port, a physical NIC address or the
endpoint of an IP VPN enabling network connectivity.
vnfd:vdu information elements related to CPUs
An implementation may have particular processor requirements. This could be for many reasons such as code related
dependencies such as the use of processor instructions, via tool suite generated dependencies such as compiler based
optimizations targeting a specific processor, or validation related dependencies as it was tested on a particular
processor. By specifying the processor characteristics in the VDU descriptor, VNFs may be able to be deployed on
suitably equipped platforms thereby facilitating a more efficient, higher performing and/or more reliable deployment.
However, note that careful consideration should be given to the use of these elements as they could limit portability of a
VNF. A VNF should be able to be deployed and instantiated by NFV-MANO even if these characteristics are not
specified in the VDU descriptor.
Identifier
cpu_instruction_set_extension
Type
Leaf
cpu_model
Leaf
cpu_model_specification_binding
Leaf
Cardinality
Description
0...N
Instruction sets are often enhanced with
instruction set extensions. This element
represents instruction set extensions that the
VDU has been developed, optimized or tested
with.
0...N
The CPU model for which the VDU has been
developed, compiled with CPU model specific
optimisations, optimized or validated on.
0...1
VDUs may be developed, compiled, optimized
or validated on particular CPU models. Some
deployments may wish to permit the VDU to
be deployed on a platform with the specified
CPU only, or with an alternative CPU with the
same architecture, instruction set, and if
specified, instruction set extensions, or with a
CPU of equivalent or greater capability.
ETSI
47
Identifier
cpu_min_clock_speed
Type
Leaf
cpu_core_reservation
Leaf
cpu_simultaneous_multi_threading_
hw_thread_specification
Leaf
cpu_core_oversubscription_policy
Leaf
cpu_core_and_hw_thread_allocation_ Leaf
topology_policy
cpu_last_level_cache_size
Leaf
cpu_direct_io_access_to_cache
Leaf
cpu_translation_look_aside_buffer_ Leaf
parameter
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Cardinality
Description
0...1
The minimum CPU clock speed may be one of
the elements that the development and
validation of the VDU has been considered
with. This may be in conjunction with some of
the other CPU elements such as CPU Model.
Requiring a minimum clock speed may be part
of a deployment requirement necessary to
help ensure particular performance or timing
related characteristics are met in the
deployment.
0...1
The number of CPU cores allocated to the
VDU. This may be necessary to help ensure
particular performance or timing related
characteristics are met in the deployment.
0...1
The use of Simultaneous Multi-Threading HW
is an efficient way to increase the compute
capacity of a platform. SMT HW threads share
some CPU core resources. In some VDU
implementations, it may be necessary to very
explicitly control the HW thread allocation on a
platform. This could be to help ensure locality
in data caches or as a mechanism to enhance
determinism.
0...1
The VDU may co-exist on a platform with
multiple VDUs or VMs and as such will be
sharing CPU core resources available in the
platform. It may be necessary to specify the
CPU core oversubscription policy in terms of
virtual cores to physical cores/threads on the
platform. This policy could be based on
required VDU deployment characteristics such
as high performance, low latency, and /or
deterministic behaviour.
0...1
The VDU may be designed to use a specific
mapping of virtual CPUs to HW threads or
cores with a specific allocation topology in
order to ensure locality in data caches and
maximize performance. The VDU will not
specify which physical resources to use, but
may specify if virtual CPUs shall be coupled
together as HW threads belonging to the same
core, or as belonging to the same processor.
0...1
The size of the last level cache may impact the
performance of the VDU, particularly for cache
intensive workloads.
0...1
The ability of an I/O device to have direct
access to the CPU cache enables
considerable memory access savings and for
I/O intensive workloads can offer significant
performance benefits.
0...N
The Translation Look-aside Buffer (TLB) is a
cache for address translation typically used by
a hardware based memory management units.
The Input/Output TLB (IOTLB) is a cache for
address translation related to remapping
hardware. The availability of a TLB and an
IOTLB can significantly improve the
performance of a virtual machine.
A number of parameters of the TLBs impact
the performance potential. These include:
•
TLB Size.
•
TLB Large Page Support.
•
IOTLB Size.
•
IOTLB Large Page Support.
ETSI
48
cpu_hot_add
Identifier
Type
Leaf
cpu_support_accelerator
Leaf
6.3.1.2.3
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Cardinality
Description
0...1
Hot add CPU is the ability to dynamically add
CPUs to a running system. The new CPU can
immediately replace a failing CPU via
migration or be brought on-line later.
0...N
This capability refers to support by the CPU
and associated chipsets of a data processing
accelerator framework, together with its
libraries and drivers.
vnfd:vdu information elements related to memory
An implementation may have particular memory requirements. This could be for many reasons such as intensive
memory needs and/or a design requirement and/or a validation requirement as the implementation could have been
validated on platforms with particular memory characteristics. By specifying the memory characteristics in the VDU
descriptor, the VNF may be able to be deployed on suitably equipped platforms thereby facilitating a more efficient,
higher performing and/or more reliable deployment. However, note that careful consideration should be given to the use
of these elements as they could limit portability of a VNF. A VNF should be able to be deployed and instantiated by
NFV-MANO even if these characteristics are not specified in the VDU descriptor.
Identifier
memory_parameter
Type
Leaf
memory_hot_add
Leaf
6.3.1.2.4
Cardinality
Description
0...N
There are a number of memory related parameters that can have a
significant impact on the performance and/or reliability of the VDU.
These include:
•
Memory Type.
•
Memory Speed
•
Number of memory channels.
•
Size of available memory.
•
Reliability characteristics such as Memory Error Correction
codes.
•
Memory oversubscription policy.
•
Memory bandwidth required per VDU.
•
Number of large pages required per VDU
•
Non-Uniform Memory Architecture (NUMA) Allocation Policy,
i.e. in NUMA architecture how you specify memory allocation
that is cognisant of the relevant process/core allocation. This
applies also to allocation of huge pages.
0...1
Hot add memory is the ability to add physical memory while the system
is running. Added memory can immediately replace failing memory via
migration or be brought on-line later.
vnfd:vdu information elements related to security
An implementation may have particular platform security requirements. This could be for many reasons such as a
sensitive nature of the implementation or due to requirements such as needed enhanced security. By specifying the
platform security characteristics in the VDU descriptor, the VNF may be able to be deployed on suitably equipped
platforms thereby facilitating a more efficient, higher performing and/or more secure, reliable deployment. However,
note that careful consideration should be given to the use of these elements as they could limit portability of a VNF. A
VNF should be able to be deployed and instantiated by NFV-MANO even if these characteristics are not specified in
the VDU descriptor.
ETSI
49
Identifier
platform_security_parameter
6.3.1.2.5
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Type Cardinalit
y
Leaf
0...N
The VDU may require some element of platform security in
order to be deployed. The security parameters could
include the availability of features such as:
•
The ability to generate true random numbers. The
availability of true random numbers can be
fundamental to the security of some protocols.
•
The availability of a Measure Launch Environment
(MLE). An MLE can provide the basis for a trusted
platform solution and help protect the platform
against attacks on vectors such as integrity,
confidentiality, reliability, and availability.
vnfd:vdu information elements related to hypervisors
An implementation may have particular hypervisor related requirements. This could be for many reasons such as being
developed on a particular hypervisor, or requiring particular hypervisor capabilities such as support for large TLB
pages. By specifying the hypervisor characteristics in the VDU descriptor, the VNF may be able to be deployed on
suitably equipped platforms thereby facilitating a more efficient, higher performing and/or more reliable deployment.
However, note that careful consideration should be given to the use of these elements as they could limit portability of a
VNF. A VNF should be able to be deployed and instantiated by NFV-MANO even if these characteristics are not
specified in the VDU descriptor.
Identifier
hypervisor_parameter
Cardinality
Description
1...N
There are a number of hypervisor related parameters that can have a
significant impact on the deployment and performance of the VDU.
These include:
•
Hypervisor type (see note 1)
•
Hypervisor version as a VDU may be validated with a particular
version.
•
Hypervisor Address Translation support parameters including:
o
Second Level Address Translation (see note 2).
o
Second Level Address Translation with Large page
support.
o
Second Level Address Translation for I/O.
o
Second Level Address Translation for I/O with Large page.
support. Where "Large" is considered to be 1 GB or
greater.
o
Support for interrupt remapping, i.e. supporting the IOMMU
in the hypervisor.
o
Support of data processing acceleration libraries in the
hypervisor, i.e. for acceleration libraries which require
hypervisor support for high performance.
NOTE 1: "BARE" could be included in this list if the network function component was to be deployed on a non-virtualised
system, but still managed via the Virtualised Infrastructure Manager.
NOTE 2: This needs to be included as the platform may support this feature but the hypervisor may not.
6.3.1.2.6
Type
Leaf
vnfd:vdu information elements related to PCIe
An implementation may have particular PCIe related requirements. This could be for many reasons such as requiring a
particular PCIe device, or a particular performance capability related to the PCIe bus. By specifying the PCIe
characteristics in the VDU descriptor, the VNF may be able to be deployed on suitably equipped platforms thereby
facilitating a more efficient, higher performing and/or more reliable deployment. However, note that careful
consideration should be given to the use of these elements as they could limit portability of a VNF. A VNF should be
able to be deployed and instantiated by NFV-MANO even if these characteristics are not specified in the VDU
descriptor.
ETSI
50
Identifier
platform_pcie_parameter
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Cardinality
Description
0...N
There are a number of PCIe related parameters that can
have a significant impact on the deployment and
performance of the VDU. These include:
•
PCIe generational capabilities.
•
PCIe bandwidth.
•
PCIe Device Pass-through.
•
PCIe SR-IOV as the VDU may require that an SRIOV virtual vunction from the specified PCIe
device can be allocated to the VM.
•
PCIe Device Assignment Affinity. The VDU may
require for performance reasons the ability to
allocate a partitionable PCIe Device capability
such as a NIC port, an entire NIC or a NIC virtual
function to the VDU while also ensuring that the
selected device is locally connected to the same
processor.
(see note).
pcie_advanced_error_reporting Leaf
0...1
Detecting and reporting correctable and un-correctable
(fatal and non-fatal) PCIe errors to software for error
handling and remediation.
platform_acceleration_device Leaf
0...N
The VDU may have been developed, optimized or
tested with an acceleration device such as a crypto
accelerator that may typically be accessed over a PCIe
bus.
NOTE:
This may be relevant for use cases that are prescriptive about the locality of CPU HW Thread (vCPU)
allocations, and would perform better with locally attached devices for efficiency/performance reasons such
as avoiding unnecessary inter-socket communications.
6.3.1.2.7
Type
Leaf
vnfd:vdu information elements related to network interfaces
An implementation may have particular network interface related requirements. This could be for many reasons such as
requiring particular NIC offload features, or a particular driver model for performance reasons. By specifying the
network interface related characteristics in the VDU descriptor, the VNF may be able to be deployed on suitably
equipped platforms thereby facilitating a more efficient, higher performing and/or more reliable deployment. However,
note that careful consideration should be given to the use of these elements as they could limit portability of a VNF. A
VNF should be able to be deployed and instantiated by NFV-MANO even if these characteristics are not specified in
the VDU descriptor.
ETSI
51
Identifier
network_interface_card_capability
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Cardinality
Description
0...N
The VDU may have been developed, optimized or
tested with certain NIC capabilities to benefit items
such as performance or scalability. These include:
•
TCP Large Segmentation Offload (LSO) for
offload the segmentation of large TCP
messages into MTU sized packets from the
CPU to the NIC.
•
Large Receive Offload (LRO), i.e. the
inverse of LSO by coalescing incoming
TCP/IP packets into larger segments for
processing in the CPU.
•
Checksum Offload.
•
Receive Side Scaling (RSS), for packet
distribution between cores.
•
Flow Director, for more fine grained (than
RSS) packet distribution between cores.
•
Mirroring of packets between interfaces.
•
Availability of Independent Rx/Tx queues for
VM so that queue pairs in the NIC can be
allocated to the VMs.
•
Jumbo Frame support.
•
VLAN tag stripping.
•
RDMA support.
•
SR-IOV support.
•
Data processing acceleration software
library support, e.g. DPDK® - see note.
network_interface_bandwidth
Leaf
0...N
The network speed/bandwidth to be guaranteed
per requested NIC.
data_processing_acceleration_libra Leaf
0...N
Name and version of the data processing
ry
acceleration library used. Orchestration can match
any NIC that is known to be compatible with the
specified library.
NOTE:
DPDK® is an example of a suitable product available commercially. This information is given for the
convenience of users of the present document and does not constitute an endorsement by ETSI of this
product.
6.3.1.2.8
Type
Leaf
vnfd:vdu information elements related to virtual switches
An implementation may have particular virtual switch related requirements. This could be for many reasons such as
virtual switch feature set or configuration interface, or for performance characteristic reasons. By specifying the virtual
switch related characteristics in the VDU descriptor, the VNF may be able to be deployed on suitably equipped
platforms thereby facilitating a more efficient, higher performing and/or more reliable deployment. However, note that
careful consideration should be given to the use of these elements as they could limit portability of a VNF. A VNF
should be able to be deployed and instantiated by NFV-MANO even if these characteristics are not specified in the
VDU descriptor.
Identifier
vswitch_capability
6.3.1.2.9
Type Cardinality
Description
Leaf
0...N
The VDU may have been developed, optimized or tested with a
particular vSwitch and may require specifying the vSwitch type, version
and key features such as overlay tunnel termination support.
vnfd:vdu information elements related to general reliability and availability
An implementation may have particular reliability and availability related requirements. This could be for many reasons
such as having an implementation that wishes to be aware of corrected or uncorrected errors in the platform for
reliability reasons. By specifying the reliability and availability related characteristics in the VDU descriptor, the VNF
may be able to be deployed on suitably equipped platforms thereby facilitating a more reliable deployment. However,
note that careful consideration should be given to the use of these elements as they could limit portability of a VNF. A
VNF should be able to be deployed and instantiated by NFV-MANO even if these characteristics are not specified in
the VDU descriptor.
ETSI
52
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Identifier
Type Cardinality
Description
corrected_error_notification
Leaf
0...N
For tracking system CORRECTABLE errors.
uncorrected_error_notification Leaf
0...N
Detects UNCORRECTED system bus errors, ECC
errors, parity errors, cache errors, and translation
lookaside buffer errors - raising exceptions.
6.3.1.2.10
vnfd:vdu information elements related to storage
An implementation may have particular storage requirements. This could be for many reasons such as requiring storage
size, type or performance characteristics. By specifying the storage characteristics in the VDU descriptor, the VNF may
be able to be deployed on suitably equipped platforms thereby facilitating a more efficient, higher performing and/or
more reliable deployment. However, note that careful consideration should be given to the use of these elements as they
could limit portability of a VNF. A VNF should be able to be deployed and instantiated by NFV-MANO even if these
characteristics are not specified in the VDU descriptor.
Identifier
storage_requirement
Type Cardinality
Description
Leaf
0...1
Required storage characteristics (e.g. size), including Key Quality
Indicators (KQIs) for performance and reliability/availability.
rdma_support_bandwitdh Leaf
0...1
The VDU may have been developed, optimized or tested with a
storage supporting RDMA over a given bandwidth.
6.3.1.3
VNF internal Virtual Link (vnfd:virtual_link)
Identifier
id
connectivity_type
Type
Leaf
Leaf
connection_points_references
Reference
root_requirement
Leaf
1
leaf_requirement
Leaf
0...1
qos
Leaf
0...N
test_access
Leaf
0...1
6.3.1.4
Cardinality
1
1
2...N
Description
Unique identifier of this internal Virtual Link.
Connectivity type (e.g. E-Line, E-LAN or
E-Tree).
References to Connection Points
(vnfd:vdu:vnfc:connection_point:id,
vnfd:connection_point:id),
e.g. of type E-Line, E-Tree, or E-LAN.
Describes required throughput of the link
(e.g. bandwidth of E-Line, root bandwidth of
E-Tree, and aggregate capacity of E_LAN).
Describes required throughput of leaf
connections to the link (for E-Tree and E-LAN
branches).
Describes the QoS options to be supported on
the VL e.g. latency, jitter, etc.
Describes the test access facilities to be
supported on the VL (e.g. none, passive
monitoring, or active (intrusive) loopbacks at
endpoints.
Connection Point (vnfd:connection_point)
Identifier
id
virtual_link_reference
Type
Leaf
Reference
type
Leaf
Cardinality
1
0...1
1
ETSI
Description
ID of the Connection Point.
References an internal Virtual Link
(vnfd:virtual_link:id, see clause
6.3.1.3) to which other VDUs, NFs, and other
types of endpoints can connect.
This may be for example a virtual port, a virtual
NIC address, a physical port, a physical NIC
address or the endpoint of an IP VPN enabling
network connectivity.
53
6.3.1.5
Deployment flavour element (vnfd:deployment_flavour)
Identifier
id
flavour_key
Type
Leaf
Leaf
constraint
Leaf
0...N
constituent_
vdu
Element
1...N
6.3.1.5.1
Cardinality
1
1
Description
ID of the VNF flavour
Monitoring parameter & its value against which this flavour is being
described. The parameters should be present as a
vnfd:monitoring_parameter, see clause 6.3.1.1.
An example is a flavour of a virtual PGW could be described in
terms of the parameter "calls per second".
There could be a flavour describing what it takes to support a
VPGW with 10k calls per second.
Constraint that this deployment flavour can only meet the
requirements on certain hardware.
Represents the characteristics of a constituent flavour element, see
clause 6.3.1.5.1.
Examples include Control-plane VDU & Data-plane VDU & Load
Balancer VDU Each needs a VDU element to support the
deployment flavour of 10k calls-per-sec of vPGW, Control-plane
VDU may specify 3 VMs each with 4 GB vRAM, 2 vCPU, 32 GB
virtual storage, etc.
Data-plane VDU may specify 2 VMs each with 8 GB vRAM, 4
vCPU, 64 GB virtual storage, etc.
Constituent VDU (vnfd:deployment_flavour:constituent_vdu)
Identifier
vdu_reference
Type
Reference
Cardinality
1
number_of_instances
constituent_vnfc
Leaf
Reference
1
1...N
6.3.2
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Description
References a VDU which should be used for this
deployment flavour by vnfd:vdu:id, see
clause 6.3.1.2.1.
Number of VDU instances required
References VNFCs which should be used for this
deployment flavour by vnfd:vdu:vnfc:id
VNF Record (vnfr)
Following instantiation, a Virtualised Network Function Record will be created to index the virtualised resources
allocated to each VNF instance, and allow traceability of associated links across multiple systems forming the NFVMANO architectural framework. This record will include sufficient information to allow future changes to the deployed
VNF instance in the light of a scalability update. VNFC instances related to a VDU are part of the VNF Record.
ETSI
54
6.3.2.1
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
vnfr base elements
Identifier
auto_scale_policy
connection_point
dependency
deployment_flavour
Type
Leaf
Element
Leaf
Reference
Cardinality
0...N
1...N
0...N
1
id
lifecycle_event
localization
Leaf
Leaf
Leaf
1
0...N
0...1
monitoring_parameter
vdu
Leaf
Element
1...N
1...N
vendor
version
virtual_link
Leaf
Leaf
Element
1
1
0...N
parent_ns
Reference
1...N
descriptor_reference
Reference
1
vnfm_id
Leaf
1
connected_external_virtual_link
Reference
1...N
vnf_address
Leaf
1...N
status
Leaf
1
notification
Leaf
1...N
lifecycle_event_history
Leaf
0...N
audit_log
Leaf
0...N
runtime_policy_info
Leaf
0...N
ETSI
Description
See clause 6.3.1.1
See clause 6.3.1.1
See clause 6.3.1.1
Reference to selected deployment flavour
(vnfd:deployment_flavour:id)
ID of the VNF instance
See clause 6.3.1.1
A language attribute may be specified to identify
default localisation/language
Active monitoring parameters
VDU elements describing the VNFC-related
relevant information, see clause 6.3.1.1
See clause 6.3.1.1
See clause 6.3.1.1
Internal Virtual Links instances used in this VNF,
see clause 6.3.1.1
Reference to records of Network Service
instances (nsr:id) that this VNF instance is
part of
The reference to the VNFD used to instantiate
this VNF
The identification of the VNFM entity managing
this VNF
Reference to a VLR (vlr:id) used for the
management access path or other internal and
external connection interface configured for use
by this VNF instance
A network address (e.g. VLAN, IP) configured
for the management access or other internal
and external connection interface on this VNF
Flag to report status of the VNF (e.g. 0=Failed,
1= normal operation, 2= degraded operation,
3= offline through management action)
Listing of systems that have registered to
received notifications of status changes
Record of significant VNF lifecycle event
(e.g. creation, scale up/down, configuration
changes)
Record of detailed operational event, (e.g. VNF
boot, operator logins, alarms sent)
Generic placeholder for input information related
to VNF orchestration and management policies
to be applied during runtime of a specific VNF
instance (e.g. for VNF prioritization, etc.)
55
6.3.2.2
VNF internal Virtual Link (vnfr:virtual_link)
Identifier
id
connectivity_type
connection_point_reference
Type
Leaf
Leaf
Reference
root_requirement
Leaf
leaf_requirement
Leaf
qos
test_access
Leaf
Leaf
6.3.2.3
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Cardinality
Description
1
ID of the internal Virtual Link instance
1
See clause 6.3.1.3
2...N
Connection Points attached to this Virtual Link
(vnfr:connection_point,
vnfr:vdu:vnfc_instance:connection_point)
e.g. of type E-Line, E-Tree, or E-LAN.
1
Available throughput of the link (e.g. bandwidth of
E-Line, root bandwidth of E-Tree, and aggregate
capacity of E_LAN)
0…1
Available throughput of leaf connections to the link (for
E-Tree and E-LAN branches)
0…N
QoS options available on the VL, e.g. latency, jitter, etc.
0…1
Test access facilities available on the VL (e.g. none,
passive monitoring, or active (intrusive) loopbacks at
endpoints
Connection Point (vnfr:connection_point)
Identifier
id
virtual_link_reference
Type
Leaf
Reference
type
Leaf
Cardinality
Description
1
ID of the Connection Point instance
1
References an internal Virtual Link
(vnfr:virtual_link:id, see clause 6.3.2.2)
to which other NFs can connect
1
See clause 6.3.1.4
ETSI
56
6.3.2.4
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Virtual Deployment Unit (vnfr:vdu)
6.3.2.4.1
vnfr:vdu base element
Identifier
computation_requirement
constraint
lifecycle_event
high_avaliability
number_of_instances
scale_in_out
virtual_network_bandwidth_
resource
vm_image
vnfc_instance
Type
Leaf
Leaf
Leaf
Leaf
Leaf
Leaf
Leaf
Cardinality
0...1
0...N
0...N
0...1
1
0...1
1
Leaf
Element
1
1...N
parent_vdu
Reference
cpu_instruction_set_extension
cpu_model
cpu_model_specification_binding
cpu_min_clock_speed
cpu_core_reservation
cpu_simultaneous_multi_
threading_hw_thread_
specification
cpu_core_oversubscription_
policy
cpu_core_and_hw_thread_
allocation_topology_policy
cpu_last_level_cache_size
cpu_direct_io_access_to_cache
cpu_translation_look_aside_
buffer_parameter
cpu_hot_add
memory_parameter
memory_hot_add
platform_security_parameter
hypervisor_parameter
platform_pcie_parameter
pcie_advanced_error_reporting
platform_acceleration_device
network_interface_card_
capability
polled_mode_driver
vswitch_capability
corrected_error_notification
uncorrected_error_notification
storage_requirement
Leaf
Leaf
Leaf
Leaf
Leaf
Leaf
0...N
0...N
0...1
0...1
0...1
0...1
See clause 6.3.1.2.1
Contains information that is distinct for each
VNFC instance created based on this VDU
Reference to the VDU (vnfd:vdu:id) used to
instantiate this element.
See clause 6.3.1.2.2
See clause 6.3.1.2.2
See clause 6.3.1.2.2
See clause 6.3.1.2.2
See clause 6.3.1.2.2
See clause 6.3.1.2.2
Leaf
0...1
See clause 6.3.1.2.2
Leaf
0...1
See clause 6.3.1.2.2
Leaf
Leaf
Leaf
0...1
0...1
0...N
See clause 6.3.1.2.2
See clause 6.3.1.2.2
See clause 6.3.1.2.2
Leaf
Leaf
Leaf
Leaf
Leaf
Leaf
Leaf
Leaf
Leaf
0...1
0...N
0...1
0...N
1...N
0...N
0...1
0...N
0...N
See clause 6.3.1.2.2
See clause 6.3.1.2.3
See clause 6.3.1.2.3
See clause 6.3.1.2.4
See clause 6.3.1.2.5
See clause 6.3.1.2.6
See clause 6.3.1.2.6
See clause 6.3.1.2.6
See clause 6.3.1.2.7
Leaf
Leaf
Leaf
Leaf
Leaf
0...N
0...N
0...N
0...N
0...1
See clause 6.3.1.2.7
See clause 6.3.1.2.8
See clause 6.3.1.2.9
See clause 6.3.1.2.9
See clause 6.3.1.2.10
6.3.2.4.2
1
Description
See clause 6.3.1.2.1
See clause 6.3.1.2.1
See clause 6.3.1.2.1
See clause 6.3.1.2.1
See clause 6.3.1.2.1
See clause 6.3.1.2.1
See clause 6.3.1.2.1
VNFC instance (vnfr:vdu:vnfc_instance)
id
Identifier
Type
Leaf
vim_id
Leaf
vc_id
Leaf
connection_point
Element
Cardinality
Description
1
Unique VNFC identification within the namespace of a
specific VNF instance.
1
VIM instance identification that manages the virtualisation
container associated with this VNFC instance.
1
Unique virtualisation container identification (e.g. VM id)
within the namespace managed by the corresponding VIM
instance.
1...N
See clause 6.3.1.2.11.
ETSI
57
6.3.2.4.3
Connection Point (vnfr:vdu: vnfc_instance:connection_point)
Identifier
id
virtual_link_reference
type
6.4
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Type
Cardinality
Description
Leaf
1
ID of the Connection Point.
Reference 1
References an internal Virtual Link
(vnfd:virtual_link:id, see clause 6.3.1.3) to which
other VNFCs can connect.
Leaf
1
This may be for example a virtual port, a virtual NIC address,
a physical port, a physical NIC address or the endpoint of an
IP VPN enabling network connectivity.
Virtual Link information elements
Exposure of connectivity options to the NFVO will require specified Virtual Link Descriptors.
The NFVO may require information from a VNFFG to determine what connectivity facilities are needed to interconnect
VNFs, and this is likely to be with a "technology neutral" abstraction. Data will be required to be passed to a lower level
system to enable logical configuration of pre-existing hardware and software networking components to fulfil a
connectivity request.
Figure 6.3: Network connection topology of a Network Service using VNFs, VLs,
and Connection Points
In ETSI GS NFV-INF 001 [i.4] three example service types have been identified for VNF connections, aligned with the
MEF E-Line, E-LAN, and E-Tree services. Use cases for these could be as follows:
•
E-Line - For a simple point to point connection between a VNF and the existing network.
•
E-LAN - where a VNF instance needs to exchange information with all other VNF instances within a Network
Service to ensure synchronisation.
•
E-Tree - where traffic from a trunk interface needs to be routed to a specific branch, such as in a load
balancing application or management system connections.
ETSI
58
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Implementation of these services within the Infrastructure network may be dependent on the physical locations of the
link end points. For example, an E-Line link between two VNFs sharing a hypervisor on one server could be connected
using a Virtual Switch (vSwitch), a switch function in the Network Interface Controller (NIC), or a Virtual Ethernet
Port Aggregator (VEPA) on an external Ethernet switch.
Likewise, ETSI GS NFV-INF [i.7] also describes Layer 3 Services, such as Layer 3 Provider Provisioned VPN
(L3VPN) services and others.
Infrastructure vendors may need to make all options available to accommodate Service Provider requirements for
performance and Security. Similarly, options involving L2 in L3 tunnelling may be required where the VNFs need to
established on physically separated servers.
6.4.1
Virtual Link Descriptor (vld)
The VLD provides a description of each Virtual Link. This type of information can be used by the NFVO to determine
the appropriate placement of a VNF instance, and by the VIM responsible for managing the virtualised resources of the
selected placement to determine the allocation of required virtualised resources on a host with adequate network
infrastructure. The VIM, or another Network Controller, can also use this information to establish the appropriate paths
and VLANs.
The VLD describes the basic topology of the connectivity (e.g. E-LAN, E-Line, E-Tree) between one or more VNFs
connected to this VL and other required parameters (e.g. bandwidth and QoS class). The VLD connection parameters
are expected to have similar attributes to those used on the ports on VNFs in ETSI GS NFV-SWA 001 [i.8]. Therefore a
set of VLs in a Network Service can be mapped to a Network Connectivity Topology (NCT) as defined in
ETSI GS NFV-SWA 001 [i.8]. Figure 6.3 shows an example of a Network Connection Topology (NCT) described by
the use of VLs referencing Connection Points of VNFs and the NS.
Identifier
id
vendor
descriptor_version
number_of_endpoints
root_requirement
Type
Leaf
Leaf
Leaf
Leaf
Leaf
leaf_requirement
Leaf
0...1
qos
test_access
Leaf
Leaf
0...N
1
connection
Reference
2...N
connectivity_type
vld_security
Leaf
Leaf
1
0...1
6.4.2
Cardinality
1
1
1
1
1
Description
ID of the VLD
Vendor generating this VLD
Version of this VLD
Number of endpoints available on this VL (e.g. E-Line=2)
Throughput of the link (e.g. bandwidth of E-Line, root
bandwidth of E-Tree, and aggregate capacity of E-LAN)
Throughput of leaf connections to the link (for E-Tree and
E-LAN branches)
QoS options available on the VL, e.g. latency, jitter, etc.
Test access facilities available on the VL (e.g. none,
passive monitoring, or active (intrusive) loopbacks at
endpoints
A reference to an attached Connection Point (nsd/
vnfd/pnfd:connection_point:id)
Connectivity types, e.g. E-Line, E-LAN, or E-Tree.
This is a signature of vld to prevent tampering. The
particular hash algorithm used to compute the signature,
together with the corresponding cryptographic certificate
to validate the signature should also be included
Virtual Link Record (vlr)
Records need to be kept of the resources used to implement a virtual network link, including bandwidth.
It is assumed that the VLR information elements contain the same elements as the VLD. If elements are added or
removed this is highlighted in the respective clause.
ETSI
59
NOTE:
In the E-Line case where there are two end points, the record could be attached to the VNFR Connection
Point, but this would also need to include an identifier of the "other" end of the link for service assurance
and fault management purposes. In the E-Tree case there are trunk and branch end points. The VNFR
could still be used at the branch augmented with an identifier of the "trunk" end. At the Trunk end,
multiple "other" (e.g. branch) end point identifiers would be needed, which gets extended every time a
new branch is added. In the E-LAN case there are multiple end points and mesh connectivity. This could
mean that each VNFR would need to have a list of all the other end points in the mesh, and that every
VNFR would need to be updated each time a new end point is added to the LAN. In this scenario it gets
hard to identify what is the "master" version of the configuration to recover in the event of a database
corruption. Again an alternative is that all the end points could have a 1:1 relationship with the VLR
holding the definition of the entire mesh in one place. Noting that every VNF needs at least one port
connection, it is proposed that the simplest solution is to have a consistent approach using a VLR
including the bandwidth and QoS class. The endpoint VNFR would then only need a pointer to the VLR
associated with each SWA port interface. Consequently it is suggested that a VLR record will represent
each deployed Virtual Link instance.
Identifier
id
vendor
version
number_of_enpoints
root_requirement
leaf_requirement
qos
test_access
connection
Type
Leaf
Leaf
Leaf
Leaf
Leaf
Leaf
Leaf
Leaf
Reference
Cardinality
1
1
1
1
1
0...1
0...N
1
1...N
parent_ns
Reference
1
vnffgr_reference
Reference
0...N
descriptor_reference
vim_id
allocated_capacity
status
Reference
Leaf
Leaf
Leaf
1
1
1...N
1
notification
Leaf
1...N
lifecycle_event_history Leaf
0...N
audit_log
Leaf
0...N
connectivity_type
Leaf
1
6.4.3
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Description
ID if the Virtual Link instance
See clause 6.4.1
See clause 6.4.1
See clause 6.4.1
See clause 6.4.1
See clause 6.4.1
See clause 6.4.1
See clause 6.4.1
References to Connection Point instances
(nsr/vnfr/pnfr:connection_point:id), see
clause 6.4.1
The reference for the Network Service instance (nsr:id)
that this VL instance is part of
References to the records of the VNFFG instances in which
this VL instance participates
Reference to the id of VLD used to instantiate this VL
The reference to the system managing this VL instance
Bandwidth allocated for each of the QoS options on this link
Flag to report status of the VL (e.g. 0=Link down, 1= normal
operation, 2= degraded operation, 3= Offline through
management action)
System that has registered to received notifications of status
changes
Record of significant VL lifecycle event (e.g. Creation,
Configuration changes)
Record of detailed operational events (e.g. link up/down,
Operator logins, Alarms sent)
Connectivity types, e.g. E-Line, E-LAN, or E-Tree
Relation between internal Virtual Links and Virtual Links
Figure 6.4 illustrates the combined use of internal Virtual Links as described in clause 6.3 and (external) Virtual Links
as described in clause 6.4.1. In this example VNF2 comprises 3 VNFCs connected to an internal Virtual Link. The first
VNFC has two Connection Points: one to the external Virtual Link, another one to the internal Virtual Link. The Virtual
Links are those depicted in figure 6.3.
ETSI
60
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Figure 6.4: Relationships between internal and external Virtual Links
6.5
Virtualised Network Function Forwarding Graph information
elements
A VNFFG information element contains metadata about the VNF Forwarding Graph itself, references to VLs, VNFs
and PNFs (as shown for instance in use cases described in clause F.1), and Network Forwarding Path elements, which
in turn include policies (e.g. MAC forwarding rules, routing entries, etc.) and references to Connection Points
(e.g. virtual ports, virtual NIC addresses, etc.).
NOTE:
When no Network Forwarding Paths element is included in a VNF Forwarding Graph descriptor, the
forwarding path applicable to a traffic flow is decided at runtime (e.g. routeing of signalling messages
across call servers).
Figure 6.5 shows an example of two VNFFGs established on top of a network topology. VNFFG1 has two NFPs
(VNFFG1:NFP1 and VNFFG1:NFP2) whereas VNFFG2 only has a single NFP (VNFFG2:NFP1).
ETSI
61
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Figure 6.5: Network Service with two VNFFGs with different NFPs
The NFV Architectural Framework (ETSI GS NFV 002 [i.2]) defines a Network Service as the subset of the end to end
service formed by Virtualised Network Functions and associated Virtual Links instantiated on the NFVI, as shown in
figure 6.6.
End-to-End Service
Network Service
VNFFG
VNF-4
VNF-2
VNF-5
VNF-1
NS
End
Point
VNF-3
Virtualisation Layer
End
Point
NFVI
Physical Link
Logical Link
Virtualisation
Compute/Storage
Infrastructure
Network
Figure 6.6: Example of a Network Service as part of an end-to-end service
with VNFs and a forwarding graph
ETSI
62
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
The Network Service Descriptor, used as a deployment template by NFVO, will need to include references to three
subcomponents:
•
Zero or more VNF Forwarding Graphs identifying the required VNFDs, and their required VLDs.
•
A definition of the interface to the physical devices, enabling the NFVO to select the appropriate VLDs as
defined in the PNFD.
VNF Forwarding Graphs (VNFFGs) are expected to be developed by service providers or their systems integration
partners, and could be customised from industry specified templates that encapsulate the common telecommunication
service patterns.
After instantiation of the VNFFG, the VNF Forwarding Graph Record (VNFFGR) has references to the VNF, PNF, and
Virtual Link instances used to instantiate the VNF Forwarding Graph. The VNFFG can have a requirement for adding a
service ID/tag to the VNFFGR and all VNF and PNF records used to fulfil the described VNFFG. This can help
fingerprint/track everything related to that service in a NFVO world (that could span multiple CMSs), and will be
needed to ensure that effective test access can be achieved for fault localisation.
Further study is needed to determine how the VNFFG will track changes to the deployed instances, such as additional
VNF instances being brought into service to handle scalability required for a traffic peak, and migration of VNF
workloads to an alternative infrastructure to enable operational maintenance or provide business continuity in the event
of major link or site failure. VNF instance migration is not addressed in the present document.
6.5.1
6.5.1.1
VNF Forwarding Graph Descriptor (vnffgd)
vnffgd base element
A top level descriptor referencing the information elements used by this VNFFG.
Identifier
id
vendor
version
Type
Leaf
Leaf
Leaf
Cardinality
1
1
1
number_of_endpoints
Leaf
1
number_of_virtual_
links
dependent_virtual_
link
network_forwarding_
path
connection_point
Leaf
1
Reference
1...N
Element
0...N
Reference
2...N
descriptor_version
constituent_vnfs
Leaf
Reference
1
1...N
vnffgd_security
Leaf
0...1
Description
ID of the VNFFG Descriptor
Specify the vendor generating this VNFFG
Specify the identifier (e.g. name), version, and description
of service this VNFFG is describing
Count of the external endpoints (connection_point
elements) included in this VNFFG, to form an index
Count of the VLs (dependent_virtual_link elements)
used by this VNFFG, to form an index
Reference to a VLD (vld:id) used to instantiate this
Forwarding Graph
This element describes a Network Forwarding Path within
the VNFFG, see clause 6.5.1.2
Reference to Connection Points
(nsd/vnfd/pnfd:connection_point:id) forming the
VNFFG including Connection Points attached to PNFs, see
clause 6.5.1.3
Version of this VNFFGD
Reference to a VNFD
(nsd:deployment_flavours:constituent_vnf:id)
used to instantiate this VNF Forwarding Graph
This is a signature of vnffgd to prevent tampering. The
particular hash algorithm used to compute the signature,
together with the corresponding cryptographic certificate to
validate the signature should also be included
ETSI
63
6.5.1.2
Network Forwarding Path (vnffgd:network_forwarding_path)
Name
id
policy
connection
6.5.2
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Type
Leaf
Leaf
Leaf
Cardinality
Description
1
Specify the identifier (e.g. name) of the Network Forwarding Path
0...1
A policy or rule to apply to the NFP
1...N
A tuple containing a reference to a Connection Point in the NFP and the
position in the path
VNF Forwarding Graph Record (vnffgr)
A top level record of the specific component resource instances used to form this VNFFG instance, and its operational
status.
6.5.2.1
vnffgr base element
Identifier
id
descriptor_reference
Type
Leaf
Reference
Cardinality
1
1
parent_ns
Reference
1
dependent_virtual_link
Reference
1...N
status
Leaf
1
notification
Leaf
1...N
lifecycle_event_history
Leaf
0...N
audit_log
Leaf
0...N
network_forwarding_path
Element
0...N
connection_point
Reference
2...N
member_vnfs
vendor
version
number_of_endpoints
number_of_vnfs
number_of_pnfs
number_of_virtual_links
Reference
Leaf
Leaf
Leaf
Leaf
Leaf
Leaf
1...N
1
1
1
1
1
1
6.5.2.2
Network Forwarding Path (vnffgr:network_forwarding_path)
Name
id
policy
connection
6.6
Description
A unique identifier for the instance of the VNFFG
Record of the VNFFGD (vnffgd:id) used to
instantiate this VNFFG
Reference to the record (nsr:id) for Network Service
instance that this VNFFG instance is part of
Reference to record for Virtual Link instance (vlr:id)
used to instantiate this VNFFG
Flag to report status of the VNFFG (e.g. 0=Failed,
1= normal operation, 2= degraded operation, 3=
Offline through management action)
Listing of systems that have registered to received
notifications of status changes
Record of significant VNFFG lifecycle events
(e.g. creation, configuration changes)
Record of detailed operational events, (e.g. graph
up/down, alarms sent)
List of Connection Points which form a Network
Forwarding Path and description of policies to
establish and rules to choose the path
Reference to Connection Points
(nsr/vnfr/pnfr:connection_point:id) forming
the VNFFG
VNF instance used to instantiate this VNFFG
See clause 6.5.1.1
See clause 6.5.1.1
See clause 6.5.1.1
See clause 6.5.1.1
See clause 6.5.1.1
See clause 6.5.1.1
Type
leaf
leaf
leaf
Cardinality
1
0...1
1...N
Description
A unique identifier for the instance of the NFP
A policy or rule to apply to the NFP
A tuple containing a reference to a Connection Point
instance in the NFP instance and the position in the path
Physical Network Function information elements
The PNF Descriptor (PNFD) will be used by the NFVO to create links between VNFs and the Physical Network
Functions, probably by communication with an appropriate specialised Virtual Infrastructure Manager (e.g. WAN
Infrastructure Manager) using a pre-NFV existing interface comprised in the Or-Vi reference point.
ETSI
64
6.6.1
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
PNF Descriptor (pnfd)
The Information contained within the PNFD is limited to the description of the connectivity requirements to integrate
PNFs. It contains:
•
The Connection Points exposed by the PNF.
•
The Virtual Link/links the Connection Point attaches to.
6.6.1.1
pnfd base element
Identifier
id
vendor
version
description
connection_point
Type
Leaf
Leaf
Leaf
Leaf
Element
descriptor_version
pnfd_security
Leaf
Leaf
6.6.1.2
Description
The ID (e.g. name) of this PNFD.
The vendor generating this PNFD.
The version of PNF this PNFD is describing.
Description of physical device functionality.
This element describes an external interface exposed by
this PNF enabling connection with a VL.
Version of the PNF descriptor.
This is a signature of pnfd to prevent tampering. The
particular hash algorithm used to compute the signature,
together with the corresponding cryptographic certificate to
validate the signature should also be included.
1
0...1
Connection Point (pnfd:connection_point)
Identifier
id
type
6.6.2
Cardinality
1
1
1
1
1...N
Type
Leaf
Leaf
Cardinality
1
1
Description
ID of the Connection Point.
This may be for example a virtual port, a virtual NIC
address, a physical port, a physical NIC address or the
endpoint of an IP VPN enabling network connectivity.
PNF Record (pnfr)
Following instantiation of a Network Service, PNF records would need to be created in an inventory indexed to the
Network Service Record to facilitate service management activities.
6.6.2.1
pnfr base element
Name
id
vendor
version
description
connection_point
parent_ns
Type
Leaf
Leaf
Leaf
Leaf
Element
Reference
Cardinality
1
1
1
1
1...N
1
descriptor_reference
Reference
1
vnffgr
Reference
0...N
oam_reference
connected_virtual_link
Leaf
Reference
1
1.,.N
pnf_address
Leaf
1...N
ETSI
Description
A unique identifier for the instance of the PNF
See clause 6.6.1.1
See clause 6.6.1.1
See clause 6.6.1.1
See clause 6.6.1.1
The reference for the record of the NS instance
(nsr:id) that this PNF instance is part of
The reference to the version of PNFD
(pnfd:version) used to instantiate this PNF
References to the records of VNFFG instances to
which this PNF is participating
The reference to the system managing this PNF
References to the VLRs (vlr:id) used to for the
management access path and all other external
connection interfaces configured for use by this
PNF instance
The network addresses (e.g. VLAN, IP) configured
for the management access and all other external
connection interfaces on this PNF
65
6.6.2.2
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Connection Point (pnfr:connection_point)
id
type
Identifier
Type
Leaf
Leaf
Cardinality
1
1
Description
ID of the Connection Point instance
See clause 6.6.1.2
6.7
VNF Instantiation input parameter
VNF Instantiation Input Parameters are exchanged in the messages between functional blocks in the NFV-MANO
architectural framework. In the present clause, only input parameters considered for the VNF instantiation operation are
covered. Other input parameters potentially needed for other VNF lifecycle operations may be covered later.
The following is not a complete list of parameters but a minimum subset considered necessary to instantiate a VNF and
identify the required deployment flavour for VNF. Parameters for operations on VNF component level (per VDU) have
not been explored and needed to be detailed; in the absence of those, the default in the VNFD/VDU would apply.
Information element
VNFD identification
Scaling methodology
Flavour ID
Threshold descriptor
Cardinality
Description
1
Information that uniquely identifies the VNFD based upon which this VNF is
being instantiated.
0...1
On-demand, auto-scaling, or scaling based on a management request.
1
The ID of the deployment flavour of the VNFD based on which this VNF
needs instantiation.
0...N
Please note that these parameters should be from amongst the list specified
in the VNFD, see clause 6.3.1.
A threshold descriptor has the info including but not restricted to monitoring
parameter, threshold values, action, etc.
Please note that these monitoring parameters should be from amongst the
list specified in the VNFD, see clause 6.3.1.2.1.
Auto-scale policy value
0...N
Represents the values for the policy meta data, which may include the
values for criteria parameter & action-type. The available criteria parameter
and action types are defined in vnfd:auto_scale_policy.
Example of values are:
•
Calls-per-second: > 100 k.
•
Action-type: scale-out to flavour ID 3.
Constraints
0...N
Represents a placeholder for any specific constraints.
Example: geographical location for each of the VNF components, based on
respective VDUs.
NS instance identification
0...N
Represents information that uniquely identifies NS instance(s) to which this
VNF instance may participate.
VNF address
0...N
Represents an address assigned to the VNF instance, bound to one of its
connection points (as described in clause 6.3.2.1. vnfr:base element).
See note.
NOTE:
A cardinality of 0 corresponds to the case where all addresses used by a VNF instance are dynamically
assigned by Management & Orchestration functions.
6.8
Network Service Instantiation Input Parameters
Network Service (NS) Instantiation Input Parameters are exchanged in the messages functional blocks in the NFVMANO architectural framework. In the present document, only input parameters considered for the NS instantiation
operation are covered. Other input parameters potentially needed for other NS lifecycle operations may be covered
later. The following is not a complete list of parameters but a minimum subset considered necessary to instantiate a
Network Service.
ETSI
66
Information element
NSD identification
Reference to an existing
VNF instance
Scaling methodology
Flavour ID
Threshold descriptor
Auto-scale policy value
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Cardinality
Description
1
Information that uniquely identifies the NSD based upon which this Network
Service is being instantiated.
1...N
List of references to existing VNF instances. This is required, for example,
when the Network Service is being instantiated bottom-up wherein the
members are already existing and need to be chained together.
0...1
On-demand, auto-scaling, or scaling based on a management request.
1
The ID of the deployment flavour of the NSD based on which this NS needs
instantiation.
0...N
0...N
Please note that these parameters should be from amongst the list specified
in the NSD, see clause 6.2.1.3).
A threshold descriptor has the info including but not restricted to monitoring
parameter, threshold values, action, etc.
Please note that these monitoring parameters should be from amongst the
list specified in the NSD (Refer clause 6.2.1.1).
Represents the values for the policy meta data, which may include the
values for criteria parameter & action-type.
The available criteria parameter and action types are defined in
nsd:auto_scale_policy.
The NS is scaled out/in according to the increment/decrement corresponding
to the difference in the resources corresponding to the current flavour ID and
scaling flavour ID.
Constraints
7
0...N
Example values are:
•
Calls-per-second: > 100 k.
•
Action-type: scale-out to flavour ID 3.
Represents a placeholder for any specific constraints.
EXAMPLE: Geographical location, requirement to first create an NFVI
resource reservation, etc.
NFV-MANO interfaces
Interfaces are defined focusing on the function they expose. Recommendation are made with respect to which
functional blocks can expose a particular interface and which functional blocks can consume such an interface; this
supports mapping of an interface to a reference point, while identifying the producer and consumer of such interface in
that reference point.
7.1
Interfaces concerning Network Services
7.1.1
Network Service Descriptor management
7.1.1.1
Description
Interface Name
Description
Notes
Produced By
Consumed By
Applicable
Reference Point(s)
Network Service Descriptor Management
This interface allows an authorized consumer functional block to manage the
Network Service Descriptor (NSD), including any related VNFFGD and VLD.
While not shown explicitly, interfaces may be consumed by authenticated and
authorized other parties.
NFVO.
OSS.
Os-Ma-nfvo.
ETSI
67
7.1.1.2
Operations
Operations
On-board Network
Service Descriptor
Disable Network
Service Descriptor
Enable Network
Service Descriptor
Update Network
Service Descriptor
Query Network
Service Descriptor
Delete Network
Service Descriptor
7.1.2
7.1.2.1
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Description
This operation allows submitting and
validating a Network Service
Descriptor (NSD), including any
related VNFFGD and VLD.
This operation allows disabling a
Network Service Descriptor, so that it
is not possible to instantiate it any
further.
This operation allows enabling a
Network Service Descriptor.
This operation allows updating a
Network Service Descriptor (NSD),
including any related VNFFGD and
VLD.
This update might include
creating/deleting new VNFFGDs
and/or new VLDs.
This operation is used to query the
information of the Network Service
Descriptor (NSD), including any
related VNFFGD and VLD.
This operation is used to remove a
disabled Network Service Descriptor.
Notes
Upon successful completion the Network Service
Descriptor is stored in the NS catalogue, and can
be used for Network Service lifecycle management.
Upon successful completion, that Network Service
Descriptor is disabled in the NS catalogue.
This operation has no effect on the NS instances
previously created, using the NSD.
Upon successful completion, that Network Service
Descriptor is enabled in the NS catalogue.
This operation has no effect on the NS instances
previously created, using the NSD.
Upon successful completion, the Network Service
Descriptor is updated in the NS catalogue (it may
be a new Network Service Descriptor, depending
on a case-by-case). This operation has no effect on
NS instances previously created, using the original
NSD.
The operation allows retrieving information from the
NSD, VNFFGDs, and VLDs. Examples include:
NSD version, list of participating VNFs,
service_deployment flavour, auto_scale_policy, etc.
Network Service lifecycle management
Description
Interface Name
Description
Notes
Produced By
Consumed By
Applicable Reference
Point(s)
Network Service Lifecycle Management.
This interface allows an authorized consumer functional block to perform lifecycle
operations on Network Service instances, such as instantiate, terminate, query, etc.
This includes (not limited to) managing the associations between different VNFs,
and of VNFs when connected to PNFs, the topology of the Network Service, and
the VNF Forwarding Graphs associated with the service.
While not shown explicitly, interfaces may be consumed by authenticated and
authorized other parties.
NFVO.
OSS.
Os-Ma-nfvo.
ETSI
68
7.1.2.2
Operations
Operations
Instantiate Network
Service
Description
This operation allows instantiating a
Network Service.
Terminate Network
Service
This operation allows terminating a
Network Service instance.
Graceful or forceful termination might be
possible based on input parameter.
This operation allows retrieving Network
Service instance attributes.
This operation allows scaling a Network
Service instance.
This operation allows updating a Network
Service instance.
Query Network Service
Scale Network Service
Update Network Service
Create VNFFG
Delete VNFFG
Query VNFFG
Update VNFFG
Create VL
Delete VL
Update VL
Query VL
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
This operation allows creating a new VNF
Forwarding Graph instance for a given
Network Service instance.
This operation allows deleting an existing
VNF Forwarding Graph instance within a
Network Service instance.
This operation allows retrieving VNFFG
instance attributes.
This operation allows updating an existing
VNF Forwarding Graph instance for a given
Network Service instance.
This operation allows creating a new VL for
a given Network Service instance.
This operation allows deleting an existing
VL within a Network Service instance.
This operation allows updating an existing
VL for a given Network Service instance.
This operation allows retrieving VL instance
attributes.
ETSI
Notes
As part of the instantiation of NS, some kind of
resource reservation information may be
exchanged.
Different options may be covered under this
operation. For example, an update may perform
instantiation of a new VNF instance and
updating an existing VNF Forwarding Graph.
This can be considered as a sub-operation of
the Update Network Service operation.
This can be considered as a sub-operation of
the Update Network Service operation.
This can be considered as a sub-operation of
the Update Network Service operation.
This can be considered as a sub-operation of
the Update Network Service operation.
69
7.1.3
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Network Service lifecycle change notification
7.1.3.1
Description
Interface Name
Description
Notes
Produced By
Consumed By
Applicable Reference
Point(s)
7.1.3.2
Operations
Notify
Network Service Lifecycle Change Notification
This interface is used to provide runtime notifications related with the changes made to
Network Service instances including (not limited to) instantiating/ terminating/ modifying
Network Service, adding/deleting VNF to a NS, adding/deleting/changing VNF Forwarding
Graphs and VLs in a NS.
These notifications are triggered after completion of the corresponding lifecycle operation.
(1) These notifications facilitate updating consuming functional blocks regarding completion
of operations that may have been triggered earlier (e.g. for keeping the OSS service
management updated).
(2) This interface is complementary to the Network Service Lifecycle Management interface.
(3) Guaranteeing delivery of notifications is considered as an implementation issue and
deliberately left out of scope.
(4) As the information carried in notifications is the most interesting for notification interfaces,
the Notify operation is described, while details of the mechanism for registering for
notifications (e.g. subscribe) and all other possible operations related to the actual
notification delivery mechanism are deliberately left out of scope.
(5) The event notification type and format is not in-scope for the present document.
(6) While not shown explicitly, interfaces may be consumed by authenticated and authorized
other parties.
NFVO.
OSS.
Os-Ma-nfvo.
Operations
Description
This operation allows providing lifecycle change notifications on Network Services.
ETSI
Notes
70
7.1.4
7.1.4.1
Network Service performance management
Description
Interface Name
Description
Notes
Produced By
Consumed By
Applicable
Reference Point(s)
7.1.4.2
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Network Service Performance Management
This interface allows performance management (measurement results collection and
notifications) on Network Services.
Network Service metrics might be calculated from measurement results coming from the
underlying layers.
Examples of Network Service metrics might be related to network performance, e.g.
latency and usage of the Virtual Links, e.g. average/maximum bandwidth used for a
certain time interval, or resource consumption, e.g. number of VMs or vCPUs used by this
Network Service.
(1) Guaranteeing delivery of notifications is considered as an implementation issue and
deliberately left out of scope.
(2) As the information carried in notifications is the most interesting for notification
interfaces, the Notify operation is described, while details of the mechanism for
registering for notifications (e.g. subscribe) and all other possible operations related
to the actual notification delivery mechanism are deliberately left out of scope.
(3) NFVO may forward performance information to the consumer functional block
received from other functional blocks, mapped to VNF, NS or some combination of
those.
(4) The event notification type and format is not in-scope for the present document.
(5) While not shown explicitly, interfaces may be consumed by authenticated and
authorized other parties.
NFVO.
OSS.
Os-Ma-nfvo.
Operations
Operations
Get performance
measurement results
Notify
Description
This operation allows collecting performance measurement results
generated on Network Services.
This operation allows providing performance notifications on
Network Services.
ETSI
Notes
71
7.1.5
7.1.5.1
Network Service fault management
Description
Interface Name
Description
Network Service Fault Management.
This interface is used to provide fault information on Network Services. These include
(not limited to) fault information resulting from the processing of information received
from other functional blocks, as well as forwarding of fault information received from
other functional blocks after correlation to the Network Service instance affected.
These notifications facilitate fault management operation on Network Services
performed by OSS.
(1)
Guaranteeing delivery of notifications is considered as an implementation issue
and deliberately left out of scope.
(2)
As the information carried in notifications is the most interesting for notification
interfaces, the Notify operation is described, while details of the mechanism for
registering for notifications (e.g. subscribe) and all other possible operations
related to the actual notification delivery mechanism are deliberately left out of
scope.
(3)
NFVO may forward fault information received from other functional blocks,
mapped to VNF, NS or some combination of those.
(4)
The event notification type and format is not in-scope for the present document.
(5)
While not shown explicitly, interfaces may be consumed by authenticated and
authorized other parties.
NFVO.
OSS.
Os-Ma-nfvo.
Notes
Produced By
Consumed By
Applicable
Reference Point(s)
7.1.5.2
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Operations
Operations
Notify
Get NS fault information
Description
This operation allows providing fault notifications on Network Services.
This operation allows collecting Network Service fault information.
7.2
Interfaces concerning Virtualised Network Functions
7.2.1
VNF Package management
7.2.1.1
Notes
Description
Interface Name
Description
Notes
Produced By
Consumed By
Applicable
Reference Point(s)
VNF Package Management.
This interface allows an authorized consumer functional block to manage the VNF
Package, including the VNFD and the software image(s).
While not shown explicitly, interfaces may be consumed by authenticated and
authorized other parties.
NFVO.
NFVO.
OSS.
VNFM.
Os-Ma-nfvo.
Or-Vnfm.
ETSI
72
7.2.1.2
Operations
Operations
On-board VNF
Package
Description
This operation allows submitting and
validating the VNF Package.
Disable VNF
Package
This operation allows disabling the
VNF Package, so that it is not
possible to instantiate any further.
Enable VNF
Package
This operation allows enabling the
VNF Package.
Update VNF
Package
This operation allows updating the
VNF Package.
Query VNF
Packages
Delete VNF
Package
This operation is used to query
information on VNF Packages.
This operation is used to remove a
disabled VNF Package.
7.2.2
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Notes
Upon successful completion, the VNF Package is stored in
the VNF catalogue, and can be used for VNF lifecycle
management.
Upon successful completion, that VNF Package is
disabled in the VNF catalogue.
This operation has no effect on the VNF instances
previously created, using the VNF Package.
Upon successful completion, that VNF Package is enabled
in the VNF catalogue.
This operation has no effect on the VNF instances
previously created, using the VNF Package.
Upon successful completion the VNF Package is updated
in the VNF catalogue (it may be a new VNF Package,
depending on a case-by-case). This operation has no
effect on VNF instances previously created, using the
original VNF Package.
An example of information retrieved regarding VNF
Package is VNFD.
VNF software image management
7.2.2.1
Description
Interface Name
Description
Notes
Produced By
Consumed By
Applicable
Reference Point(s)
7.2.2.2
Operations
Operations
Add Image
Delete Image
Update Image
Query Image
Copy Image
VNF Software Image Management.
This interface allows an authorized consumer functional block to manage the VNF
software images in VIM.
(1) While not shown explicitly, interfaces may be consumed by authenticated and
authorized other parties.
(2) The interface exposure assumes (but does not restrict) VNF software images
are stored in repositories managed by the VIM(s) in order to minimize delays
incurred on transferring and installing such software images as part of the
VNF instance lifecycle.
(3) The consumption of the interface over Os-Ma-nfvo reference point may only
apply to a subset of the operations described in clause 7.2.2.2, in particular to
"Copy Image" and "Query Image". The rest of operations can be performed
by using the proper "VNF Package management" interface as described in
clause 7.2.1.
VIM.
VIM.
NFVO.
NFVO.
VNFM.
OSS.
Or-Vi.
Vnfm-Vi.
Os-Ma-nfvo.
Description
This operation allows adding new VNF software
images to the image repository.
This operation allows deleting in the VNF
software images from the image repository.
This operation allows updating the VNF
software images in the image repository.
This operation allows querying the information
of the VNF software images in the image
repository.
This operation allows copying images from a
VIM to another.
ETSI
Notes
The image repository to be used for VNF
instances is assumed to be in the VIM.
For example, this would allow retrieving a
selection of images previously provisioned,
based on filtering criteria (e.g. all images
available to a particular VIM).
73
7.2.3
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
VNF lifecycle operation granting
7.2.3.1
Description
Interface Name
Description
Notes
Produced By
Consumed By
Applicable Reference
Point(s)
7.2.3.2
Operations
Grant Lifecycle
Operation
VNF Lifecycle Operation Granting.
This interface allows an authorized consumer functional block to request a
producer functional block permission and relevant information for performing
certain VNF lifecycle operations (e.g. instantiation, scaling, etc.), so that it is
ensured that such operations are consistent with operator's policies as to avoid
raising conflicts in terms of resources usage or Network Service planning.
(1) It is assumed that policies used in the granting of the permission are in effect
at the producer of the interface (or a third party functional block from which
the producer of the interface can check), in order to avoid potential conflicts
with regards to managing Network Service lifecycle, other VNF lifecycle
operation permission requests, and resources.
(2) The permission procedure will result in granting or denying the execution of
the lifecycle operation. In case of granting the permission, the permission
can include as well specific information needed to execute the operation (e.g.
Infrastructure Domain where instantiation needs to be performed). Therefore,
the procedure can have an impact on the sub-sequent interaction of the
consumer functional block with other functional blocks, and in particular, with
entities responsible for resource management (e.g. the permission obtained
via this interface may have to be presented in sub-sequent interactions to
other producer functional bocks).
(3) While not shown explicitly, interfaces may be consumed by authenticated
and authorized other parties.
NFVO.
VNFM.
Or-Vnfm.
Operations
Description
This operation allows
requesting the
permission to
perform a certain
VNF lifecycle
operation on a new
or existing VNF. The
sub-type of lifecycle
operation is
parameterized in the
operation.
Notes
The grant lifecycle operation request should contain information
representative of necessary components to execute the requested VNF
lifecycle operation. In particular, the request should have sufficient
information for the producer functional block to make an impact
assessment and grant or not grant permission in accordance with
available policies. For instance, in some cases the granting of a VNF
scale-out operation may require knowing how many more virtualised
resources are needed, as well as some form of reservation (e.g. internal to
NFVO or to VIM) may be required. The response to the grant request may
return information about such reservation to the consumer functional
block. In other cases it may only require knowing the type of lifecycle
operation. In addition, the granted operations need to provide enough
information back to the consumer functional block in order to further
execute the lifecycle operation, e.g. Infrastructure Domain endpoint.
ETSI
74
7.2.4
7.2.4.1
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
VNF lifecycle management
Description
Interface Name
Description
Notes
Produced By
Consumed By
Applicable Reference
Point(s)
VNF Lifecycle Management.
This interface allows an authorized consumer functional block to perform lifecycle
operations on VNF(s), i.e. all operations needed to request and manage associations
of NFVI Resources to a VNF, and maintain such associations in conformance with
the VNF Descriptor and authorized run-time changes, throughout the lifecycle of the
VNF.
(1) The specific operations authorized by this interfaces need to be continued by
the operations facilitated by the "Virtualised Resources Management" interface.
(2) The NFVO produced interface may be extended or derived from the one
produced by VNFM.
(3) While not shown explicitly, interfaces may be consumed by authenticated and
authorized other parties.
VNFM.
NFVO.
NFVO.
OSS.
Or-Vnfm.
Os-Ma-nfvo.
ETSI
75
7.2.4.2
Operations
Operations
Instantiate VNF
Query VNF
Scale VNF
Check VNF
instantiation
feasibility
Heal VNF
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Description
This operation allows creating a
VNF instance.
This operation allows retrieving
VNF instance state and attributes.
This operation allows scaling
(out/in, up/down) a VNF instance.
This operation allows verifying if
the VNF instantiation is possible.
This operation is used to request
appropriate correction actions in
reaction to a failure.
Update VNF
software
This operation allows applying a
minor/limited software update
(e.g. patch) to a VNF instance.
Modify VNF
This operation allows making
structural changes
(e.g. configuration, topology,
behavior, redundancy model) to a
VNF instance.
Upgrade VNF
software
This operation allows deploying a
new software release to a VNF
instance.
Terminate VNF
This operation allows terminating
gracefully or forcefully a previously
created VNF instance.
Notes
Attributes returned may include for example number and
location of VMs allocated to the VNF instance.
No VNF instance is created as a result of the operation.
This assumes operational behaviour for healing actions by
VNFM has been described in the VNFD.
An example may be switching between active and standby
mode.
The software update implies no structural changes (e.g. in
configuration, topology, behavior, redundancy model).
The goal is to not require termination/re-instantiation, or at
least not for the entire VNF.
This operation is only required if the software update
procedure requires a change in the VNF infrastructure
e.g. changing an image. Software updates that can be
performed using solely pre-existing functional blocks
(e.g. EM) may not need to invoke this operation.
This may have to be designed as a transaction, rather
than an atomic operation.
Depending on the "modification plan", it may involve
multiple catalogue Management and VNF lifecycle
management atomic operations, and may involve multiple
VNFM managers.
It is possible that this transaction may not be exposed as
an operation by the VNFM, and only offered as an
extension when exposed by the NFVO.
This may have to be designed as a transaction, rather
than an atomic operation.
Depending on the "software upgrade plan", it may involve
multiple catalogue Management and VNF lifecycle
management atomic operations, and may involve multiple
VNFM managers.
It is possible that this transaction may not be exposed as
an operation by the VNFM, and only offered as an
extension when exposed by the NFVO.
This operation is only required if the software upgrade
procedure requires a change in the VNF infrastructure
e.g. changing an image. Software upgrades that can be
performed using solely pre-existing functional blocks
(e.g. EM) may not need to invoke this operation.
ETSI
76
7.2.5
VNF lifecycle change notification
7.2.5.1
Description
Interface Name
Description
Notes
Produced By
Consumed By
Applicable Reference
Point(s)
7.2.5.2
VNF Lifecycle Change Notification
This interface is used to provide runtime notifications related to the state of the VNF
instance, as a result of changes made to VNF instance including (not limited to)
changes in the number of VDUs, changing VNF configuration and/or topology due to
auto-scaling/update/upgrade/termination, switching to/from standby etc.
(1) These notifications facilitate updating consuming functional blocks regarding
completion of operations that may have been triggered earlier (e.g. for keeping
the NFV Instances catalogue updated and/or the EM updated).
(2) This interface is complementary to the VNF Lifecycle Management interface;
when the VNF Lifecycle Management interface is produced by the VNFM, then
VNFM also produces the VNF Lifecycle Changes Notification interface; when the
VNF Lifecycle Management interface is produced by the NFVO, then the NFVO
also produces the VNF Lifecycle Changes Notification interface.
(3) When VNFM is the producer, and NFVO the consumer, there is no expectation
that the NFVO automatically forwards such notifications to the OSS. However,
NFVO may process such notifications to assess the impact on a Network
Service instance, and issue at its turn notifications to the OSS via the Network
Service Lifecycle Changes Notification interface.
(4) Guaranteeing delivery of notifications is considered as an implementation issue
and deliberately left out of scope.
(5) As the information carried in notifications is the most interesting for notification
interfaces, the Notify operation is described, while details of the mechanism for
registering for notifications (e.g. subscribe) and all other possible operations
related to the actual notification delivery mechanism are deliberately left out of
scope.
(6) The event notification type and format is not in-scope for the present document.
(7) While not shown explicitly, interfaces may be consumed by authenticated and
authorized other parties.
VNFM.
NFVO.
VNFM.
NFVO.
OSS.
EM.
Or-Vnfm.
Os-Ma-nfvo.
Ve-Vnfm-em.
Operations
Operations
Notify
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Description
This operation allows providing notifications on state changes
of a VNF instance, related to the VNF Lifecycle.
ETSI
Notes
77
7.2.6
7.2.6.1
VNF configuration
Description
Interface Name
Description
Notes
Produced By
Consumed By
Applicable Reference
Point(s)
7.2.6.2
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
VNF Configuration.
This interface allows configuring a VNF after successful instantiation.
(1) This interface needs to be exposed by every VNF.
(2) The operations of this interface work on "configuration objects", which are
structured data entities that represent a collection of configuration
parameters.
(3) If consumed by the VNFM, this interface can be used to pass configuration
parameters to the VNF without the need to be aware of the semantics of
these configuration parameters.
(4) Guaranteeing delivery of notifications is considered as an implementation
issue and deliberately left out of scope.
(5) As the information carried in notifications is the most interesting for
notification interfaces, the Notify operation is described, while details of the
mechanism for registering for notifications (e.g. subscribe) and all other
possible operations related to the actual notification delivery mechanism
are deliberately left out of scope.
(6) The event notification type and format is not in-scope for the present
document.
(7) While not shown explicitly, interfaces may be consumed by authenticated
and authorized other parties.
VNF
VNF
VNFM
EM
Ve-Vnfm-vnf
Out of scope
Operations
Operations
Get Config Object
Create Config Object
Delete Config Object
Set Config Object Attributes
Notify Config Change
Description
Retrieve configuration data for a specific configuration object
Create a configuration object
Delete a configuration object
Modify a configuration object
This operation allows providing notifications about
configuration changes for a configuration object
ETSI
Notes
78
7.2.7
VNF performance management
7.2.7.1
Description
Interface Name
Description
VNF Performance Management.
This interface is used to provide VNF performance management (measurement
results collection and notifications) related to the behavior of the VNF applicationlayer.
(1) VNF application-layer refers to "VNF as an application".
(2) VNF application-layer performance-related information may enable triggering
run-time operations like auto-scaling based on (not limited to) exceeding set
thresholds for calls-per-second, number-of-subscribers.
(3) The VNFM produced performance-related information may just be the
replication of VNF produced performance-related information for forwarding the
same information further (e.g. to understand implications on Network Services).
No reference point exists between VNF and NFVO; hence any communication
between VNF and NFVO flows through the VNFM.
(4) Guaranteeing delivery of notifications is considered as an implementation issue
and deliberately left out of scope.
(5) As the information carried in notifications is the most interesting for notification
interfaces, the Notify operation is described, while details of the mechanism for
registering for notifications (e.g. subscribe) and all other possible operations
related to the actual notification delivery mechanism are deliberately left out of
scope.
(6) The event notification type and format is not in-scope for the present document
(e.g. events indicating a need for scaling).
(7) While not shown explicitly, interfaces may be consumed by authenticated and
authorized other parties.
VNF.
VNFM.
VNF.
VNFM.
NFVO.
EM.
Ve-Vnfm-vnf.
Or-Vnfm.
Out of scope.
Notes
Produced By
Consumed By
Applicable Reference
Point(s)
7.2.7.2
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Operations
Operations
Get performance
measurement results
Notify
Description
This operation allows collecting performance
measurement results generated on resources.
This operation allows providing notifications with
application-layer performance measurement results.
ETSI
Notes
79
7.2.8
7.2.8.1
VNF fault management
Description
Interface Name
Description
Notes
Produced By
Consumed By
Applicable Reference
Point(s)
7.2.8.2
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
VNF Fault Management.
This interface allows providing VNF application-layer fault information (e.g. network
function configuration failures, communication failures between software modules).
(1) VNF application-layer refers to "VNF as an application".
(2) VNF application-layer fault information may facilitate fault management operations
performed by other functional blocks (e.g. fault correlation, root-cause analysis)
and/or may be used to trigger VNF instance healing operations.
(3) The VNFM produced fault information may just be the replication of VNF
produced fault information for forwarding the same information further. No
reference point exists between VNF and NFVO; hence any communication
between VNF and NFVO flows through the VNFM.
(4) This interface can be used for VNF health-check by an external functional block
(e.g. VNFM); the mechanisms to enable, trigger and/or use this interface for VNF
health-check are not covered in the present document.
(5) Guaranteeing delivery of notifications is considered as an implementation issue
and deliberately left out of scope.
(6) As the information carried in notifications is the most interesting for notification
interfaces, the Notify operation is described, while details of the mechanism for
registering for notifications (e.g. subscribe) and all other possible operations
related to the actual notification delivery mechanism are deliberately left out of
scope.
(7) The event notification type and format is not in-scope for the present document.
(8) While not shown explicitly, interfaces may be consumed by authenticated and
authorized other parties.
VNF.
VNFM.
VNF.
VNFM.
NFVO.
EM.
Ve-Vnfm-vnf.
Or-Vnfm.
Out of scope.
Operations
Operations
Get VNF fault information
Notify
Description
This operation allows collecting VNF application-layer fault information.
This operation allows providing application-layer fault notifications.
ETSI
Notes
80
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
7.3
Interfaces concerning virtualised resources
7.3.1
Virtualised resources catalogue management
7.3.1.1
Description
Interface Name
Description
Notes
Produced By
Consumed By
Applicable Reference
Point(s)
7.3.1.2
Virtualised Resources Catalogue Management.
This interface allows an authorized consumer functional block to query the
catalogues of virtualised resources and get notifications about their changes.
(1) While not shown explicitly, interfaces may be consumed by authenticated and
authorized other parties.
(2) Guaranteeing delivery of notifications is considered as an implementation issue
and deliberately left out of scope.
(3) As the information carried in notifications is the most interesting for notification
interfaces, the Notify operation is described, while details of the mechanism for
registering for notifications (e.g. subscribe) and all other possible operations
related to the actual notification delivery mechanism are deliberately left out of
scope.
(4) The event notification type and format is not in-scope for the present document.
VIM.
VIM.
NFVO.
VNFM.
Or-Vi.
Vi-Vnfm.
Operations
Operations
Query Resource
Catalogue
Notify Resources
Catalogue Changes
Description
This operation allows retrieving the list of catalogued virtualised
resources, and/or a specific catalogued resource on which the
consumer is allowed to perform subsequent operations.
This operation provides change notifications on virtualised resources
catalogues managed by the producer functional block.
ETSI
Notes
81
7.3.2
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Virtualised resources capacity management
7.3.2.1
Description
Interface Name
Description
Notes
Produced By
Consumed By
Applicable
Reference Point(s)
7.3.2.2
Operations
Query Capacity
Notify Capacity
Changes
Virtualised Resources Capacity Management.
This interface allows an authorized consumer functional block to perform operations related to
NFVI-PoP capacity and usage reporting (see note (1)). The interface allows retrieving information
about:
•
NFVI-PoP total resources capacity over which virtualised resources are provisioned.
•
Virtualised resources capacity and density, e.g. how many virtualised resources can be
created from existing NFVI-PoP resources.
•
Statistics and mapping of NFVI-PoP resources usage to virtualised resources usage
(see note (2)), global at NFVI-PoP level, and per deployed virtualised partition (see
note (3)).
The interface enables the capture of information for reporting NFVI-PoP total resources usage and
executing analytics for capacity planning, capacity changes, and consequently for Network
Service planning, etc.
(1) The management of the NFVI-PoP capacity (e.g. adding, removing physical hosts, adding
network switches, etc.), the topology, and its configuration, as well as the management of
specific physical equipment are out of the scope of this interface.
(2) Virtualised resources usage is facilitated by the "Virtualised Resources Management"
interface.
(3) A virtualised partition refers to a group of allocated virtualised resources that the producer of
the interface can correlate (e.g. as managed under the same deployment project).
(4) Guaranteeing delivery of notifications is considered as an implementation issue and
deliberately left out of scope.
(5) As the information carried in notifications is the most interesting for notification interfaces,
the Notify operation is described, while details of the mechanism for registering for
notifications (e.g. subscribe) and all other possible operations related to the actual
notification delivery mechanism are deliberately left out of scope.
(6) The event notification type and format is not in-scope for the present document.
(7) While not shown explicitly, interfaces may be consumed by authenticated and authorized
other parties.
VIM.
NFVO.
Or-Vi.
Operations
Description
This operation allows querying the capacity usage of an NFVI-PoP. The operation
can be used to gather information at different levels, from specific virtualised
partition capacity usage, to total capacity availability in the NFVI-PoP.
This operation allows notifying about capacity changes in the NFVI-PoP.
ETSI
Notes
82
7.3.3
7.3.3.1
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Virtualised resources management
Description
In the following, the types of virtualised resources that can be consumed from an NFVI-PoP comprise computing,
storage and networking; they have been described in clause 4.2.
Interface Name
Description
Notes
Produced By
Consumed By
Applicable Reference
Point(s)
Virtualised Resources Management.
This interface allows an authorized consumer functional block to perform
operations on virtualised resources available to the consumer functional block.
The interface includes common operations for creating, querying, updating and
terminating compute, storage and network isolated virtualised resources, or a
composition of different types in a resource grouping (see note (1)), as well as
managing virtualised resource reservations.
(1) A virtualised resource grouping consists of one or more computing, storage
and network virtualised resources associated with each other and that can be
handled as an atomic object.
(2) While not shown explicitly, interfaces may be consumed by authenticated
and authorized other parties.
VIM.
VIM.
NFVO.
NFVO.
VNFM.
VNFM.
Or-Vi.
Vi-Vnfm.
Or-Vnfm.
ETSI
83
7.3.3.2
Operations
Operations
Allocate
Resource
Description
This operation allows requesting the instantiation and
assignment of a virtualised resource to the VNF, as
indicated by the consumer functional block.
This operation allows querying a virtualised resource,
i.e. retrieve information about an instantiated virtualised
resource.
This operation allows updating the configuration and/or
parameterization of an instantiated virtualised resource.
Query
Resource
Update
Resource
Scale
Resource
Update
Resource
Reservation
This operation allows scaling a virtualised resource by
adding or removing capacity, e.g. adding vCPUs to a
virtual machine.
This operation allows moving virtualised resources
between locations. For instance, the operation performs
the migration of a computing resource from one host to
another host; while for a storage resource, it migrates
the resource from one storage location to another.
This operation allows executing specific commands on
certain allocated virtualised resources. Examples on
compute resources can include (but not limited to): start,
stop, pause, suspend, capture snapshot, etc.
This operation allows de-allocating and terminating an
instantiated virtualised resource. This operation frees
resources and returns them to the NFVI resource pool.
This operation allows requesting the reservation of a set
of virtualised resources to a consumer functional block
without performing the steps of "Allocate Resource".
This operation allows querying an issued resources
reservation, e.g. to discover the virtualised resources
included in a specific reserved resources pool, or the
amount of free resources in such a pool.
This operation allows updating an issued resources
reservation to increase or decrease the amount of
virtualised resources in the reserved resources pool.
Release
Resource
Reservation
This operation allows releasing an issued resources
reservation, hence freeing the reserved virtualised
resources.
Migrate
Resource
Operate
Resource
Release
Resource
Create
Resource
Reservation
Query
Resource
Reservation
ETSI
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Notes
Virtualised resources can have associated
elements like templates expressing the set
of configurations that can be used when
creating, reading, updating, and deleting
resources. At the discretion of the
interface producer and Service Provider's
policy, a given authorized consumer
functional block may be allowed the
creation of new templates and/or
configurations.
The migration operation relies and
executes its commands based on the
abstraction of hardware resources
performed by the producer of the interface
(or the implementer of such abstractions).
Depending on Service Provider policy,
deployments may need to be able to
prevent the update of virtualised resource
reservations when resources in the
reserved set are still allocated.
Depending on Service Provider policy,
deployments may need to be able to
prevent the release of virtualised resource
reservations when resources in the
reserved set are still allocated.
84
7.3.4
7.3.4.1
Virtualised resources performance management
Description
Interface Name
Description
Notes
Produced By
Consumed By
Applicable Reference
Point(s)
7.3.4.2
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Virtualised Resources Performance Management
This interface allows providing performance management (measurement results
collection and notifications) related to virtualised resources (see note (1)) including
(not limited to) resource consumption level, e.g. vCPU power consumption, VM
memory usage oversubscription, VM disk latency, etc.
(1) Only types of resources that have been catalogued and offered through right
abstractions to consumer functional blocks are in scope (refer to clause 7.1.3.3
"Virtualised Resources Management").
(2) The VNFM produced notifications may be a replication of VIM produced
notifications for forwarding the same information further.
(3) Guaranteeing delivery of notifications is considered as an implementation issue
and deliberately left out of scope.
(4) As the information carried in notifications is the most interesting for notification
interfaces, the Notify operation is described, while details of the mechanism for
registering for notifications (e.g. subscribe) and all other possible operations
related to the actual notification delivery mechanism are deliberately left out of
scope.
(5) The event notification type and format is not in-scope for the present document.
(6) While not shown explicitly, interfaces may be consumed by authenticated and
authorized other parties.
VIM.
VIM.
VNFM.
NFVO.
VNFM.
EM.
Or-Vi.
Vnfm-Vi.
Ve-Vnfm-em.
Operations
Operations
Get performance
measurement results
Notify
Description
This operation allows collecting performance measurement results
generated on virtualised resources.
This operation allows providing notifications with performance
measurement results on virtualised resources.
ETSI
Notes
85
7.3.5
7.3.5.1
Virtualised resources fault management
Description
Interface Name
Description
Notes
Produced By
Consumed By
Applicable Reference
Point(s)
7.3.5.2
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Virtualised Resources Fault Management
This interface allows providing fault information related to the resources (see
note (1)) visible to the consumer functional block, including virtual containers
(VMs) crashes, virtual network ports errors, virtual container's to storage
disconnection, etc. The interface also provides information about faults
related to the pools of resources, for instance, reserved resources
unavailable, resource exhaustion, etc.).
(1) Only types of resources that have been catalogued and offered through
right abstractions to consumer functional blocks are in scope (refer to
clause 7.1.3.3 "Virtualised Resources Management").
(2) The VNFM produced notifications may be a replication of VIM produced
notifications for forwarding the same information further.
(3) Guaranteeing delivery of notifications is considered as an
implementation issue and deliberately left out of scope.
(4) As the information carried in notifications is the most interesting for
notification interfaces, the Notify operation is described, while details of
the mechanism for registering for notifications (e.g. subscribe) and all
other possible operations related to the actual notification delivery
mechanism are deliberately left out of scope.
(5) The event notification type and format is not in-scope for the present
document.
(6) While not shown explicitly, interfaces may be consumed by
authenticated and authorized other parties.
VIM.
VIM.
VNFM.
NFVO.
VNFM.
EM.
Or-Vi.
Vnfm-Vi.
Ve-Vnfm-em.
Operations
Operations
Get resource fault
information
Notify
Description
This operation allows collecting virtualised resource fault information.
This operation allows providing fault notifications on virtualised resources.
7.4
Policy administration interface
7.4.1
Description
Interface Name
Description
Notes
Produced By
Consumed By
Applicable
Reference
Point(s)
Notes
Policy Management
This interface allows an authorized consumer functional block to manage NFV policies.
(1) While not shown explicitly, interfaces may be consumed by authenticated and
authorized other parties.
(2) This interface is exposed by VNFM and VIM to provide a way to manage policies
related with VNF and NFVI Resources respectively on request from NFVO.
NFVO.
VNFM.
VIM.
OSS.
NFVO.
NFVO.
Os-Ma-nfvo.
Or-Vnfm.
Or-Vi.
ETSI
86
7.4.2
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Operations
Operations
Create policy
Update policy
Delete policy
Query policy
Activate policy
De-activate policy
Description
This operation allows defining policy rules include
conditions and actions
This operation allows updating an existing policy
Notes
This involves modifying policy
metadata including conditions, actions
This operation allows delete policy after being created
This operation allows querying about a particular policy
or a querying the list of available policies
This operation enables activating an available policy
This operation enables de- activating an active policy
7.5
Network Forwarding Path management interface
7.5.1
Description
Interface Name
Description
Notes
Produced By
Consumed By
Applicable Reference
Point(s)
7.5.2
Network Forwarding Path Management.
This interface allows an authorized consumer functional block to perform Network
Forwarding Path management and notification operations. This interface provides the
facility to have policy based linkages on a VNF Forwarding Graph as expressed by a
Network Forwarding Path.
(1) The Network Forwarding Path containing an ordered list of Connection Points
along with rules/policies associated to the list.
(2) Network Forwarding Path rule related information may enable Network
Controllers to configure accordingly forwarding tables in NFVI network
resources.
(3) As the information carried in notifications is the most interesting for notification
interfaces, the Notify operation is described, while details of the mechanism for
registering for notifications (e.g. subscribe) and all other possible operations
related to the actual notification delivery mechanism are deliberately left out of
scope.
(4) While not shown explicitly, interfaces may be consumed by authenticated and
authorized other parties.
(5) The event notification type and format is not in-scope for the present document.
(6) The Network Forwarding Path is limited to the resource controlled by VIM
exposing the interface.
VIM.
NFVO.
Or-Vi.
Operations
Operations
Create Network
Forwarding Path
Update Network
Forwarding Path
Delete Network
Forwarding Path
Query Network
Forwarding Path
Notify
Description
This operation allows creating a Network Forwarding Path.
This operation allows updating the information associated with a Network
Forwarding Path.
This operation allows deleting a Network Forwarding Path.
This operation allows querying information about a specified Network Forwarding
Path instance.
This operation allows providing information about a Network Forwarding Path rule.
ETSI
Notes
87
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
7.6
NFVI hypervisor management interface
7.6.1
Description
The hypervisor exposes a wide array of functionalities, grouped by the following (non-exhaustive) categories of
information:
•
Host and VM Configuration/Life cycle.
•
Resources and VM inventory management.
•
Networking/connectivity.
•
CPU, Pools, Clusters management and metrics.
•
Memory and storage management and metrics.
•
Utilities for task and distributed scheduling.
•
Alarms and Events management.
•
Logging for SLA, debugging.
Interface Name
Description
Notes
Produced By
Consumed By
Applicable
Reference Point(s)
NFVI Hypervisor Management Interface.
This interface allows an authorized consumer functional block to request a producer
functional block to perform operations on hypervisor-accessed resources (e.g. compute,
storage and networking) in the NFVI.
(1) This interface needs to be exposed by NFV participating Hypervisors.
(2) This interface maps to the Nf-Vi/H interface described in ETSI GS NFV-INF 001 [i.4]
and ETSI GS NFV-INF 004 [i.6].
(3) Guaranteeing delivery of notifications is considered as an implementation issue and
deliberately left out of scope.
(4) As the information carried in notifications is the most interesting for notification
interfaces, the Notify operation is described, while details of the mechanism for
registering for notifications (e.g. subscribe) and all other possible operations related
to the actual notification delivery mechanism are deliberately left out of scope.
(5) The event notification type and format is not in-scope for the present document.
(6) While not shown explicitly, interfaces may be consumed by authenticated and
authorized other parties.
NFVI.
VIM.
Nf-Vi.
ETSI
88
7.6.2
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Operations
Operations
Description
Create virtual machine This operation is needed to create
and start a virtual machine
Notes
This and all subsequent virtual machine related
operations fall under the "Host & VM configuration
/Lifecycle" and "Resources and VM inventory
management" hypervisor information categories.
An example is the "create" command in virsh (KVM);
rules apply for different VM goals such as affinity
requirements.
Shutdown virtual
This operation is needed to
An example is the "shutdown" command in virsh
machine
shutdown a virtual machine
(KVM).
Destroy virtual
This operation is needed to destroy a This does an immediate ungraceful shutdown of the
machine
virtual machine
virtual machine.
An example is the "destroy" command in virsh (KVM).
Update virtual
This operation is needed to update a This can be used to update memory or vCPU of the
machine
created virtual machine
virtual machine or attach, update or detach a disk or
network interface.
Examples include the following commands in virsh
(KVM):
•
update-device, setmem, setvcpus;
•
attach-device, attach-disk, attach-interface;
•
detach-device, detach -disk, detach -interface.
List the virtual
This operation is needed to list the
An example is the "list" command in virsh (KVM).
machines
virtual machines
Query a virtual
This operation is needed to show the Examples are the "dumpxml, dominfo" commands in
machine by name id
details of a given virtual machine
virsh (KVM).
Reboot a virtual
This operation reboots a virtual
An example is the "reboot" command in virsh (KVM).
machine
machine
Suspend a virtual
This operation suspends a virtual
The suspended VM will not consume the processor
machine
machine
resources but the memory will be maintained.
An example is the "suspend" command in virsh (KVM).
Resume a virtual
This operation resumes a virtual
An example is the "resume" command in virsh (KVM).
machine
machine, which has been previously
suspended
Save a virtual
This operation stops the virtual
An example is the "save" command in virsh (KVM).
machine
machine and save the data to a file.
Restore a virtual
Recreate the virtual machine from a An example is the "restore" command in virsh (KVM).
machine
file created by the save operation.
Create Storage Pool
This operation creates and starts a
This, and all subsequent storage pool and virtual
storage pool
machine storage related operations fall under the
"Memory and storage management and metrics"
hypervisor information category.
Examples include pool-build, pool-create command in
virsh (KVM).
Modify Storage Pool
This operation modifies a specified
An example is pool-edit command in virsh (KVM).
storage pool
List Storage pools
This operations lists the storage
An example is pool-list command in virsh (KVM).
pools known to the hypervisor
Delete Storage Pool
This operation deletes a specified
Examples include pool-delete or pool-destroy
storage pool
command in virsh (KVM).
Query Storage Pool
This operation provides the
An example is pool-info command in virsh (KVM).
information about a storage pool
Create Virtual
This operation creates a storage for
Examples include vol-create, vol-create-as,
Machine storage
the virtual machine on a given pool
vol-create-from commands in virsh (KVM).
Delete Virtual Machine This operation deletes the specified
An example is vol-delete command in virsh (KVM).
Storage
virtual machine storage
List Virtual machine
This operations lists the virtual
An example is vol-list command in virsh (KVM).
storage
machine storage on a storage pool
Query Virtual Machine This operation provides the
Examples include vol-info, vol-dumpxml, vol-pool
Storage
information about the specified virtual command in virsh (KVM).
machine storage
Create Snapshot
This operation creates a snapshot for Snapshots take the disk, memory, and device state of
a virtual machine
a virtual machine at a specified point-in-time, and save
it for future use.
An example is the "snapshot create" command in virsh
(KVM).
ETSI
89
Operations
Create hypervisor
policies
Description
This operation creates a hypervisor
policy (a policy regarding hypervisoraccessed resources, e.g. compute,
storage, networking) for a NFV
tenant
Update hypervisor
This operation updates a hypervisor
policies
policy for a NFV tenant
List hypervisor policies This operation lists all the hypervisor
policies created for a given NFV
tenant
Query hypervisor
This operation shows the details of a
policy
given specific hypervisor policy
relative to different resources and
VMs in the database
Delete hypervisor
This operation deletes the hypervisor
policy
policy relative to specific resources
Change VM
These operations adds/removes
configuration
resources to/from the VM
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Notes
This and all subsequent hypervisor policy operations
fall under the "Resources and VM inventory
management" hypervisor information category.
This falls under the "Utilities" hypervisor information
category, covering tasks relative to the VM scaling,
porting, or consolidating workloads; rules for
placement may apply, such as affinity; power
management.
Migrate virtual
This operation migrates the specified This falls under the "Utilities" hypervisor information
machine
VM to another hypervisor
category. Example include migrate command in virsh
(KVM).
Get hypervisor
This operation allows retrieving
This falls under the "Logging for SLA and debugging",
measurement results
performance and reliability
"CPUs, clusters, pools" and "Memory and Storage"
measurement results. These are
hypervisor information categories.
multiple operations
Examples include metrics on: CPU processors,
Memory, Fans, Temperature, Voltage, Power,
Network, Battery, Storage, Cable/Interconnect,
Software components, affinity Watchdog, PCI devices;
VM stalls.
Create virtual network This operation creates a specified
This and all subsequent operations referring to "virtual
device
virtual network device
network device" fall under the "Network/connectivity"
hypervisor information category. The intent here is to
represent this as a broad category, rather than listing
in detail each virtual network device type and the
operations allowed on it.
Delete virtual network This operation deletes a specified
device
virtual network device
Update the virtual
This operation updates a virtual
network device
network device
List the virtual network This operation lists virtual network
devices
devices created by the requesting
entity
Query a virtual
This operation retrieves details of a
network device
specified virtual network device
Notify
This operation allows providing
This falls under the "Events and alarms management"
notifications about hypervisorhypervisor information category.
accessed resources changes
ETSI
90
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
7.7
NFVI compute management interface
7.7.1
Description
Interface Name
Description
NFVI Compute Management Interface.
This interface allows an authorized consumer functional block to request a
producer functional block to perform management operations on physical compute
and storage resources in the NFVI.
(1) The description and operations of the interface are not complete in the
present document. In the present document, it is assumed that NFV-MANO
supports the use of this interface in order to provide notifications from NFVI
to the VIM regarding changes in physical resources (e.g. fault information,
inventory information), without defining the specific interface operations.
(2) This interface maps to the Nf-Vi/C interface described in
ETSI GS NFV-INF 001 [i.4] and ETSI GS NFV-INF 003 [i.5].
NFVI.
VIM.
Nf-Vi.
Notes
Produced By
Consumed By
Applicable Reference
Point(s)
7.8
NFVI networking management interface
7.8.1
Description
Interface Name
Description
Notes
Produced By
Consumed By
Applicable
Reference Point(s)
NFVI Networking Management Interface.
This interface allows an authorized consumer functional block to request a
producer functional block to perform management operations on networking
resources in the NFVI.
(1) While some operations in this interface can also be exposed via the NFVI
Hypervisor Management Interface (handling the "Network/Connectivity"
hypervisor information category), this interface focuses on functionality
exposed by network (SDN) controllers comprised in NFVI, via appropriate
abstractions.
(2) This interface maps to the Nf-Vi/N interface described in
ETSI GS NFV-INF 001 [i.4] and ETSI GS NFV-INF 005 [i.7].
(3) The operations represent an example of a high level abstraction of
functionality that could be exposed, but it should be understood that in
alternative implementations, several finer level abstractions could be
exposed to collectively achieve each of the outlined operations.
(4) Guaranteeing delivery of notifications is considered as an implementation
issue and deliberately left out of scope.
(5) As the information carried in notifications is the most interesting for
notification interfaces, the Notify operation is described, while details of the
mechanism for registering for notifications (e.g. subscribe) and all other
possible operations related to the actual notification delivery mechanism are
deliberately left out of scope.
(6) The event notification type and format is not in-scope for the present
document.
(7) While not shown explicitly, interfaces may be consumed by authenticated
and authorized other parties.
NFVI.
VIM.
Nf-Vi.
ETSI
91
7.8.2
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Operations
Operations
Create virtual
network
Delete virtual
network
Update the virtual
network
List the virtual
networks
Query a virtual
network
Create a subnet
Update a subnet
List the subnets
Query subnet
Delete a subnet
Create port
Update port
List ports
Query port
Delete port
Notify
7.9
Description
Notes
This operation is needed to create virtual networks(L2/L3 overlay
or infrastructure) for inter VNF or inter VNFC interconnectivity
This operation is needed to delete the virtual networks which
have been created for inter VNF/inter VNFC connectivity
This operation is needed to update the attributes of a created
virtual network belonging to a tenant, e.g. updating the admin
status of a virtual network
This operation is needed to list the virtual networks belonging to a
NFV tenant
This operation is needed to show the details of a given virtual
network
This operation creates a subnet for a given virtual network
This operation updates the information associated to a subnet
This operation is needed to list the subnets belonging to a virtual
network
This operation is needed to show the details of a given subnet
This operation deletes a subnet associated with a virtual network Deletion fails if any of the IP
addresses in the subnet are
still allocated to any of the
VNF VMs
This operation creates a port on a given virtual network
This operation updates the information associated with a port
associated with a virtual network
This operation lists the ports associated with a virtual network
This operation shows the details of a given port
This operation deletes a port from the virtual network
This operation is needed to provide notifications regarding virtual
networks, subnets, ports
Interfaces exposed between different service providers
In order to support federation between different service providers' NFV Infrastructures and some of the NFV use cases
(e.g. NFVIaaS, VNFaaS, VNPaaS) exchanges between service providers should be supported via interfaces exposed by
a service provider A and consumed by another service provider B.
There are multiple scenarios to consider, all technically possible (including various combinations). In all cases,
authentication is necessary and authorization policies (e.g. role-based) shall control the access of service provider B to
functions exposed by service provider A:
1)
Functional blocks deployed by service A may allow access to any functional block belonging to service
provider B:
a)
2)
3)
In this case, no new interface may need to be specified; policies determine who has access to what and in
what conditions.
A designated NFV-MANO functional block deployed by service provider A is identified as a single point of
contact that exposes interfaces for interactions with functional blocks deployed by service provider B. This has
2 sub-cases:
a)
Interfaces that are already exposed by the designated functional block deployed by service provider A are
sufficient; the subset of the interfaces to be exposed for consumption by functional blocks of service
provider B, and the appropriate authorization policies shall be defined.
b)
New interfaces are needed to be defined for some reason; new authentication mechanism and
authorization policies shall be defined.
Some service provider A's OSS may expose interfaces to other service provider B functional, in which case
blocks; new interfaces may be needed to be exposed to satisfy the requirements; authentication mechanisms
and authorization policies shall be defined.
ETSI
92
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Among the options presented, the recommendation is for service provider A to expose interfaces to external parties
(e.g. service provider B) at a single point of contact in the NFV-MANO architecture, in order to minimize security risks
and integration and operational complexity.
The recommendation is that the functional block designated as the single point of contact for exposing service provider
A's interfaces is the NFV Orchestrator, for the following reasons:
•
It is the designated NFV-MANO entity that handles lifecycle management of Network Services, and most
interactions between different service providers are likely to be at the level of services produced by one service
provider, and consumed by another service provider.
•
It is already targeted to, and may re-expose, as needed, interfaces (extended or derived) that other NFVMANO functional blocks (VIM, VNFM) produce.
•
It requires less additional development/integration in order to fulfil this role, in comparison to other functional
blocks in the NFV-MANO Architectural Framework, because it is the only functional block in the NFVMANO Architectural Framework that either produces and/or consumes interfaces to/from all other functional
blocks in the NFV-MANO Architectural Framework: OSS/BSS, VNFM, and VIM.
NOTE:
Additional interfaces mentioned in 2.b above may be assessed and described in a later stage.
ETSI
93
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Annex A (informative):
VNF Instance management and orchestration case study
This annex provides use cases of NFV management and Orchestration. Each use case is addressed in a separate second
level section.
A.1
IMS MRF management and orchestration case study
This case study assumes NFVO and NFV Infrastructure (NFVI) as main actors.
This clause is aimed at describing an IMS MRF orchestration and management use case.
A brief summary of the identified VNFs is provided in figure A.1.
Figure A.1: IMS MRF VNF Forwarding Graph and VNFs
Two VNFs and one VNF Forwarding Graph have been identified for this use case:
•
MRB VNF.
•
MRF VNF with 2 VNF components :MRF-C+P and MRF storage.
•
IMS MRF VNF Forwarding Graph made of 2 VNFs: MRB and MRF.
The details of the IMS MRF VNF Forwarding Graph and the VNFs that compose it are described in
ETSI GS NFV-SWA 001 [i.8].
Note that as for the VNF mapping, depending on vendor choice, the deployment use case might be slightly different.
The VNF deployment use case proposed here can be considered as a typical use case. It is not the goal of the present
document to cover all possible IMS MRF deployment scenarios.
A possible target deployment of an IMS MRF VNF Forwarding Graph is shown in figure A.2.
ETSI
94
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Figure A.2: IMS MRF target deployment
This deployment example is showing the IMS MRF VNF Forwarding Graph deployed over 3 HW resources and 7 VMs
(grey boxes). It illustrates also the affinity and anti-affinity rules described in the previous clause.
HW resource #1 contains 3 VMs:
•
VM-1 for an instantiation of MRB VNF (MRB-1), used as active.
•
VM-2 for an instantiation of MRF-C+P (MRF-1).
•
VM-3 for an instantiation of MRF Storage (MRF Storage-1), used as master.
HW resource #2 contains also 3 VMs:
•
VM-4 for an instantiation of MRB (MRB-2), used as standby.
•
VM-5 for an instantiation of MRF-C+P (MRF-2).
•
VM-6 for an instantiation of MRF Storage (MRF Storage-2), used as slave.
HW resource #3 contains a single VM (VM-7) with an instantiation of MRF-C+P (MRF-3).
MRB, MRF-C+P and MRF-Storage have 3 different redundancy models that have direct implication for NFVO:
•
Active/Standby for MRB.
•
N+1 Active for MRF-C+P.
•
Master/Slaves for MRF-Storage.
As per affinity rule, MRB, MRF-C+P and MRF Storage are co-located on the same hardware resources.
As per anti-affinity rules, MRB primary and secondary are located on different hardware resources. Same is true for
MRF Storage primary and secondary.
Prerequisites are:
•
The code for both VNFs (MRB, MRF) and the corresponding VNFDs have been developed and tested.
ETSI
95
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
•
The IMS MRF VNF Forwarding Graph and the corresponding VNF Forwarding Graph Descriptor (VNFFGD)
has been developed and tested.
•
VM images for the various components (MRB, MRF-C+P, MRF Storage) are available.
•
This scenario assumes that any needed network setup required as prerequisite (e.g. VLAN with the appropriate
QoS, holes in firewall opened, multicast/IGMP enabled) has been done. This is not a requirement and might be
considered done dynamically as part of the deployment.
A.1.1
IMS MRF on-boarding
The overall process of on-boarding the IMS MRF VNF Forwarding Graph is shown on the figure A.3.
The overall sequencing is the following:
1)
NFVO receives a request to on-board the MRB VNF with the MRB VNFD attached.
2)
NFVO validates the content of the MRB VNFD and if valid, stores it in its VNF catalogue.
3)
NFVO receives a request to on-board the MRF VNF with the MRF VNFD attached.
4)
NFVO validates the content of the MRF VNFD and if valid, stores it in its VNF catalogue.
5)
NFVO receives a request to on-board the IMS MRF VNF Forwarding Graph with the IMS MRF VNFGD
attached.
6)
NFVO validates the content of the IMS MRF VNFGD. This would include checking that the MRB and MRF
VNFD are present in the VNF catalogue, as they are referenced by the IMS MRF VNFFGD. If the IMS MRF
VNFFGD is valid, NFVO stores it in its NS catalogue.
7)
If any error, return error to caller.
The 2 VNFDs (MRB, MRF) and the IMS MRF VNF Forwarding Graph are now on-boarded and available to the
NFVO.
ETSI
96
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Figure A.3: IMS MRF VNF Forwarding Graph on-boarding
A.1.2
IMS MRF instance provisioning and configuration
Figure A.4 summarizes the overall deployment use case.
Provision IMS MRF
Inst
Data
NFVO
Provide initial configuration for each
deployed artifact
MRB
MRF-C +
MRF-P
Create servers for each
VM needed
MRF
Storage
NFVI
Logical Environment
Figure A.4: IMS MRF VNF Forwarding Graph deployment use case
The assumption is that NFVO, when receiving the command to provision an IMS MRF VNF Forwarding Graph,
receives at the same time the instantiation definition data, containing all the instantiation information needed:
ETSI
97
•
The number of MRB to provision (2 in this use case).
•
The number of MRF-C+P pairs to deploy (3 in this use case).
•
The number of MRF storage to deploy (2 in this case).
•
IP address for S-CSCF external component.
•
IP address for app server external component.
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
NOTE 1: IP addresses for the various instances can be provided in the instantiation definition information or can be
constructed dynamically. The latter is assumed in this use case.
NOTE 2: Instantiation definition data can be provided in various possible forms: file, parameters of the request; the
exact format would depend on the selected request format.
It is also assumed that the 2 VNFs for MRB, MRF and the IMS MRF VNF Forwarding Graph have been on-boarded as
shown in clause A.1.1 IMS MRF instance on-boarding and are present in the VNF and NS catalogues of the NFVO.
The overall sequencing of this typical deployment is as follow:
1)
NFVO receives the ‘provision IMS MRF' request with the instantiation definition data defined above.
2)
NFVO checks that the IMS MRF VNF Forwarding Graph is on-boarded and that the IMS MRF VNFFGD is
present in the VNFG catalogue as well as the MRB and MRF VNFD in the VNF catalogue.
3)
NFVO checks the validity of the provided instantiation data against the catalogued VNFDs for MRB and MRF
and the IMS MRF VNFGD.
4)
NFVO applies the affinity and anti-affinity rules for each VNF and VDU to determine location of instances:
a)
The 2 MRB instances, MRB-1 and MRB-2 need to be on separate HW resources to guarantee
availability.
b)
The 3 instances of MRF-C+P need to be on separate HW resources to guarantee availability.
c)
The 2 MRF storage instances need to be on separate HW resources to guarantee availability.
d)
Each MRF storage instance should be co-located with an MRF instance as per affinity rule to get better
performance.
e)
Result of the validation is the need for 7 VMs on 3 different HW resources as described earlier. The
characteristics needed from each VM are provided by the information in the VNFD.
5)
NFVO validates the appropriate resources identified in the previous step are available for fulfilling this
request.
6)
If any error, return error to caller.
ETSI
98
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Figure A.5: IMS MRF deployment - initial steps
7)
NFVO sends a create server command to the NFV infrastructure (for instance a cloud management system
(CMS) to create VM-1, hosting instance MRB-1 with information coming from MRB VNF.
NOTE 3: The Cloud Management System might not always be present and if not, NFVO would directly create the
VM as shown in the next step.
8)
The NFVI in turn sends a request to the hypervisor to create the VM-1 using the provided image. NFVI returns
the information on the VM created to NFVO.
9)
NFVO sends a create server command to the NFVI to create VM-2, hosting instance MRF-1 with information
coming from MRB VNF.
10) NFVI in turn sends a request to the hypervisor to create the VM-2 using the provided image. NFVI returns the
information on the VM created to NFVO.
11) NFVO sends a create server command to the NFVI to create VM-3, hosting instance MRF Storage-1 with
information coming from MRB VNF.
12) NFVI in turn sends a request to the hypervisor to create the VM-3 using the provided image. NFVI returns the
information on the VM created to NFVO.
13) If any error, return error to caller.
ETSI
99
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Figure A.6: IMS MRF deployment steps for HW resource 1
14) Steps 7 to 13 are repeated for the second HW resources to create:
a)
VM-4 hosting MRB-2 (role: standby);
b)
VM-5 hosting MRF-2, (role: active);
c)
VM-6 hosting MRF Storage 2 (role: slave).
15) Steps 8 & 9 are repeated for the third HW resource to create VM-7 hosting MRF-3 (role: active).
NOTE 4: Steps 7 to 15 are presented as ordered for ease of reading, but creation of servers might be done in any
order and might be parallelized.
16) Once the VM is created, its starts and waits for its instantiation definition file.
17) NFVO augments the Instantiation Definition data with data from the VNFD and data obtained as result of the
VM creation (e.g. IP addresses, subnet mask, etc.). This global instantiation definition file contains
information on the type of VDU running on each VM (MRB, MRF, MRF Storage), HA role for each instance
(active/ standby, master/ slave), IP addresses of the other instances of same type for redundancy, IP addresses
of external components (S-CSCF, App Server).
NOTE 5: There might be multiple ways of provide configuration information to the application and this use case is
just using a simple one.
18) NFVO then pushes the augmented instantiation definition file to a pre-defined storage location that will be
mounted by the corresponding VM. This file will be used to provide the initial configuration of the
application,
19) Once all instantiation definition files have been pushed, NFVO provides a successful response to the provision
IMS MRF.
20) Each VM reads the instantiation definition file, discovers its VDU type and start the application.
NOTE 6: This case study is using an Instantiation File read by the VM for configuring the VDU. There might be a
lot of alternative ways to do the initial configuration of the IMS MRF components, but from a scenario
point of view, the end result should be the same.
ETSI
100
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
21) Based on its VDU type and the data from the Instantiation Definition file, the application instance (MRB-1,
MRB-2, MRF-1, MRF-2, MRF-3, MRB Storage-1, and MRB Storage-2) synchronizes with each other. For
instance, secondary registers with primary.
NOTE 7: Start command might be explicitly sent from NFVO or EM to MRF or MRB or both, or can be implicit,
i.e. as soon as it is configured, each component can check with its peer and start in the right order. In this
use case, application start is automatic and the application is started when the VM is configured and it
automatically configures itself once it got its Instantiation Definition file.
This is illustrated by the following sequence diagram where VM #X is one of the VM instantiated and VNF #X is one
of the application artefact instances.
Figure A.7: IMS MRF Deployment Sequence Diagram - Configuration phase
A.2
Network Service fault Management case study
Network Service is achieved through a set of network functions which are VNFs and/or existing network elements.
When a network element fails, it may affect performance anomaly of the others and even cause a series of alarms which
are almost fake alarms.
Network Service fault refers to Network Service performance anomaly, including service interruption, performance
degradation, etc. It can happen that the Network Service performance runs far from the expected capacity as configured
at the beginning of Network Service instantiation. At this time, the VNFs may still work, but probably at low
performance.
For example, Network Service faults in NFV are shown in Figure A.8. This example assumes that the Network Service
is composed of the following network functions which are NE1, NE2, NE3, VNF-4, VNF-5 and VNF-6. Moreover, the
example assumes the NMS/OSS has the topology of the Network Service. NE2 and NE3 are the same type of devices,
e.g. routers in an administration domain. NE1 has the routing knowledge of the Network Service. When NE2 fails to
offer any service, the traffic of Network Service which is used to pass through NE2 has to change to pass through NE3.
At the same time NE2 will send an alarm (e.g. outage alarm) to NMS through EM-2. VNF-4/5 will also send alarm
(e.g. VNF-4 out of service alarm, VNF-5 overload alarm) to NMS/OSS via EM-4/5 respectively. The Network Service
fault management also requires NMS/OSS and NFVO to support subscription to Network Service fault information.
For Network Service fault management, NMS/OSS should support correlate Network Service fault alarms and NFVI
fault events, for Root Cause Analysis (RCA) purpose, eventually, can enhance network availability in NFV
environment.
ETSI
101
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Regarding the enhancements to existing NMS/OSS, for Network Service fault management and relevant NFVI
information of VM, hardware and networking service, NMS/OSS should support to subscribe and receive relevant
NFVI fault event from NFVO. NFVO should support to subscribe and receive Network Service fault information from
network management system.
For Networks Service management and RCA reasons, NMS may trigger EM to query and receive events about
virtualised resources used by the VNF, e.g. failures on a VM.
Figure A.8: Network Service fault
ETSI
102
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Annex B (informative):
VNF lifecycle management
B.1
Introduction
This annex represents a collection (non-exhaustive) of management flows related to VNF Lifecycle that should be
possible to implement by means of the interfaces described in the main part of the present document.
The following principles are used for all the flows in this clause:
•
The NFVO is the single point of access for all requests from the OSS to simplify the interfacing.
•
The NFVO handles lifecycles of Network Service and VNF Forwarding Graph.
•
The VNF Manager handles VNF lifecycle from an application point of view.
•
The NFVO has the end-to-end view of the resources being allocated across Network Services and VNFs by
VNF Managers: all requests for resource allocation transit through, or are verified and granted by the NFVO.
All the flows in this clause are informative, representing best practices for each of the lifecycle operation and have the
main goal of identifying needed interfaces as well as information needed on those interfaces.
At each step, in case of failure, an immediate return might happen. All the failure cases have not been shown on the
sequence diagrams or on the text for sake of simplicity.
The dotted-line arrows represent the return path of the request or in some specific cases, a notification. The return path
of a request can be a true reply in case of a request/reply exchange or a notification otherwise.
For VNF lifecycle management flows, allocation of resource can be done by the NFVO or by the VNF Manager. For
each lifecycle operation, the 2 options, when applicable, are presented in separate clauses.
ETSI
103
B.2
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
VNF Package on-boarding flows
Figure B.1: Use case diagram for VNF Package on-boarding
The use case diagram above provides the following use cases related to VNF Package on-boarding:
•
On-board VNF Package.
•
Disable VNF Package.
•
Enable VNF Package.
•
Update VNF Package.
•
Query VNF Package.
•
Delete VNF Package.
ETSI
104
B.2.1
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
On-board VNF Package flow
VNF Package on-boarding refers to the process of submitting VNF Package to the NFVO to be included in the
catalogue.
Sender
NVF Orchestrator
Catalog
VIM
(Image Repository)
1. On-board VNF Package
2. Validate VNFD
3. Notify Catalog
4. Upload image(s)
6. Ack VNF Package On-boarding
5. Ack image(s) upload
Figure B.2: VNF Package on-boarding message flow
The main steps for VNF Package on-boarding are:
1.
VNF Package is submitted to NFVO for on-boarding VNFD using the operation On-board VNF Package of
the VNF Package Management interface.
2.
NFVO processes the VNFD including (not limited to):
a)
Checking for the existence of mandatory elements.
b)
Validating integrity and authenticity of the VNFD using manifest file and manifest file security received
in VNFD.
3.
NFVO notifies the catalogue.
4.
NFVO makes VM images available to each applicable VIM using the operation Add Image of the VNF
Software Image Management interface. It is expected that the VIM validates the integrity of VNF software
images(s) as part of this operation.
5.
VIMs acknowledge the successful uploading of the image.
6.
NFVO acknowledges the VNF Package on-boarding to the sender.
ETSI
105
B.2.2
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Disable VNF Package flow
Disabling a VNF package refers to the process of marking a VNF Package as disabled in the catalogue, so that it is not
possible to instantiate VNFs with it any further.
Figure B.3: Disable VNF Package message flow
The main steps for disabling a VNF Package are:
1.
A request to disable a VNF Package is submitted to the NFVO using the operation Disable VNF package of
the VNF package management interface.
2.
The NFVO processes the request and checks if the VNF package exists, is enabled and the contained VNFD is
referenced by any NSD.
3.
If the VNF Package is enabled and the contained VNFD is not referenced by any NSD, the NFVO notifies the
catalogue to disable the VNF Package in the catalogue.
4.
The NFVO acknowledges the VNF Package disable request.
B.2.3
Enable VNF Package flow
Enabling a VNF Package refers to the process of marking a VNF Package as enabled in the catalogue, so that it can be
used to instantiate VNFs again.
Figure B.4: Enable VNF Package message flow
ETSI
106
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
The main steps for enabling a VNF Package are:
1.
A request to enable a VNF Package is submitted to the NFVO using the operation Enable VNF Package of the
VNF Package management interface.
2.
The NFVO processes the request and checks if the VNF Package exists, is disabled and is not marked as
deletion pending. Optionally, the NFVO may validate the stored VNFD before enabling it.
3.
If the VNF Package is disabled and is not marked as deletion pending, the NFVO notifies the catalogue to
enable the VNF Package in the catalogue.
4.
The NFVO acknowledges the VNF Package enable request.
B.2.4
Update VNF Package flow
Updating VNF Package refers to the process of submitting a modified VNF Package to the NFVO to be included in the
catalogue.
Figure B.5: Update VNF Package message flow
The main steps for VNF Package update are:
1.
Modified VNF Package is submitted to the NFVO using the operation Update VNF Package of the VNF
Package management interface.
2.
The NFVO checks if a VNF Package exists already for this VNF and is not marked as deletion pending and if
so, processes the updated VNF Package including (not limited to).
3.
a)
Checking for the existence of mandatory elements.
b)
Validating integrity and authenticity of the VNFD using manifest file and manifest file security received
in VNFD.
The NFVO notifies the catalogue for insertion of a new version of the VNF Package in the catalogue.
NOTE 1: Some existing VNFs might still use the previous version of the VNF Package, so modifying it will create
a new version in the catalogue.
4.
Optionally, The NFVO makes new VM images available to each applicable VIM using the operation Update
Image of the VNF Image Management interface. It is expected that the VIM validates the integrity of VNF
software images(s) as part of this operation.
5.
VIMs acknowledge the successful uploading of the image if step 4 has been done.
NOTE 2: Steps 4 and 5 need to be performed if VNF images should be updated. As an alternative, the images may
come from EM or VNF and be uploaded to VIM by VNFM.
ETSI
107
6.
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
The NFVO acknowledges the VNF Package update.
B.2.5
Query VNF Packages flow
Querying VNF Packages allows returning from the catalogue the information of the VNF Packages.
Figure B.6: VNF Package query message flow
The main steps for VNF Package query are:
1.
The NFVO receives a query request for the VNF Package(s) using the operation Query VNF Packages of the
VNF Package management interface. One or more filter parameters may be included in the Query operation to
filter the VNF Packages.
2.
The NFVO gets from the catalogue the VNF Package(s) that satisfies the specified filter conditions.
3.
The NFVO acknowledges the VNF Packages query.
B.2.6
Delete VNF Package flow
Note that there might be multiple versions of the VNF Package, and the assumption is that the delete request removes
all versions.
ETSI
108
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Figure B.7: Delete VNF Package message flow
The main steps for VNF Package deletion are:
1.
Request to delete VNF Package is submitted to the NFVO using the operation Delete VNF Package of the
VNF Package Management interface.
2.
The NFVO checks if the VNF Package is disabled and not in use. If VNF Package is not disabled, the request
is rejected.
3.
If VNFD is disabled and not in use, then ask the catalogue to remove all the versions of the VNF Package. The
catalogue will then remove it/them.
4.
The NFVO deletes image(s) from VIM(s) stored during VNF Package on-boarding.
5.
VIM acknowledges the successful deleting of the image(s).
6.
If VNF Package is disabled and still in use, the NFVO set the VNF Package in deletion pending.
7.
The NFVO acknowledges VNF Package deletion request.
NOTE:
If the VNF Package is in deletion pending, the NFVO will later check if there is any VNF instance using
the VNF Package during the VNF instance termination process. If there is no VNF instance using the
VNF Package any more, the NFVO asks the catalogue to remove the corresponding version(s) of the
VNF Package.
B.3
VNF instantiation flows
B.3.1
VNF instantiation flows with resource allocation done by
NFVO
ETSI
109
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
B.3.1.1 VNF Check Feasibility
The Check Feasibility runs a feasibility check of the VNF instantiation or scaling request to reserve resources before doing the actual instantiation/scaling.
Figure B.8: VNF check feasibility message flow
ETSI
110
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
NFVO receives a request to check feasibility of VNF instantiation/scaling. This request might come from an OSS,
commissioning of a new VNF or VNF scaling, or part of an order for a Network Service instantiation/scaling. In most
of the case, it will be an option of a VNF instantiation or scaling request.
The main steps for VNF check feasibility are:
1.
NFVO receives a request to check feasibility of VNF instantiation/scaling.
2.
NFVO calls VNF Manager to check feasibility of instantiation/scaling request with the input parameters
provided using the operation Instantiate/Scale VNF of the VNF Lifecycle Management interface.
3.
VNF Manager validates the request and processes the VNFD and the input parameters. This might include
modifying/complementing the input parameters with VNFD data and VNF lifecycle specific constraints. See
details in clause B.3.1.5 VNF Manager: Request validation and processing.
4.
VNF Manager returns in response to step 2 the (possibly) updated list of input parameters to NFVO as
response of the Check VNF instantiation/scaling feasibility.
5.
NFVO executes any needed pre-allocation processing work. See details in clause B.3.1.6 NFVO:
Pre-allocation processing.
6.
Optionally, NFVO requests to VIM for availability of additional resources (compute, storage and network)
needed for the various VDUs of the VNF instance and reservation of those resources using the operation
Create Resource Reservation of the Virtualised Resources Management interface.
7.
VIM checks the availability of additional resources (compute, storage and network) needed for the various
VDUs of the VNF instance and reserves them.
8.
VIM returns result of reservation back to NFVO as response to step 6.
9.
The NFVO acknowledges the completion of the check feasibility.
NOTE:
The feasibility check can be considered from 2 perspectives:
1)
Feasibility from the perspective of the overall availability of resources. This is the role of the
NFVO initiating the feasibility check/ reservation.
2)
Feasibility from the perspective whether the VNF instantiation/scaling with its input parameters
and the VNFD make sense. This is the VNF Manager role initiating the feasibility
check/reservation. This particular operation has been specified on the VNF Lifecycle management
interface where the interface is produced by the VNFM and consumed by the NFVO.
This flow considers the first perspective. Another variant, if this flow is not executed, would be to consider the
feasibility check as an option done by the VNF Manager.
ETSI
111
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
B.3.1.2 VNF instantiation flow
VNF instantiation refers to the process of identifying and reserving the virtualised resources required for a VNF, instantiating the VNF and starting the VDU associated with
each VNF.
Figure B.9: VNF instantiation message flow
ETSI
112
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
NFVO receives a request to instantiate a new VNF. This request might come from an OSS, commissioning of a new
VNF or part of an order for a Network Service instantiation, or might come from the VNF Manager when the need to
instantiate a new VNF is detected by the VNF Manager itself or by the EM. Thus the Sender in the above diagram can
be the OSS or a VNF Manager.
The main steps for VNF instantiation are:
1.
NFVO receives a request to instantiate a new VNF using the operation Instantiate VNF of the VNF Lifecycle
Management interface along with instantiation data required to deploy the VNF.
2.
NFVO validates the request. See details in clause B.3.1.3 NFVO: Validation.
3.
Optionally, NFVO runs a feasibility check of the VNF instantiation request to reserve resources before doing
the actual instantiation as described in clause B.3.1.1.
4.
NFVO calls VNF Manager to instantiate the VNF, with the instantiation data and, if step 3 has been done, the
reservation information using the operation Instantiate VNF of the VNF Lifecycle Management interface. See
details in clause B.3.1.4 NFVO: Request to VNF Manager to instantiate the VNF.
5.
VNF Manager validates the request and processes it. This might include modifying/complementing the input
instantiation data with VNFD data and VNF lifecycle specific constraints.
6.
VNF Manager then calls the NFVO for resource allocation using the operation Allocate Resource of the
Virtualised Resources Management interface. See details in clause B.3.1.5 VNF Manager: Request validation
and processing.
7.
NFVO executes any needed pre-allocation processing work. See details in clause B.3.1.6 NFVO:
Pre-allocation processing.
8.
NFVO requests allocation of resources to the VIM (compute, storage and network) needed for the various
VDUs of the VNF instance using the operation Allocate Resource of the Virtualised Resources Management
interface. See details in clause B.3.1.7 Orchestration: Resource allocation (compute, storage and network) and
interconnection setup.
9.
VIM allocates the internal connectivity network.
10.
VIM allocates the needed compute (VMs) and storage resources and attaches instantiated VMs to internal
connectivity network.
NOTE:
The VIM utilises southbound interfaces towards the NFVI to allocate resources. Allocation includes
starting up services such as a VM.
11.
Acknowledgement of completion of resource allocation back to NFVO.
12.
NFVO acknowledges the completion of the resource allocation back to VNF Manager, returning appropriate
configuration information.
13.
VNF Manager configures the VNF with any VNF specific lifecycle parameters and using the get/create/set
config object operations over the VNF configuration interface. Additionally, the VNFM notifies the EM (if
present) of the new VNF.
14.
The EM configures the VNF using the get/create/set config object operations over the VNF configuration
interface.
15.
VNF Manager acknowledges the completion of the VNF instantiation back to the NFVO.
16.
The NFVO acknowledges the completion of the VNF instantiation.
B.3.1.3 NFVO: validation
When receiving a request to instantiate a VNF, the NFVO executes the following steps:
1)
Verify validity of request to instantiate VNF. It includes validating that the sender is authorized to issue this
request.
ETSI
113
2)
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
If needed, validate parameters passed for VNF instantiation for technical correctness and policy conformance,
e.g. mandatory parameters present, instantiation parameters within policies.
As an option, the NFVO can run a feasibility check of whether the VNF instantiation request is feasible or not up to the
reservation of resources before doing the actual instantiation. This complete step is optional, but when executed, all
sub-steps should be done. It would involve additional exchange with the VNF Manager to get VDU details and with
VIM to get resource availability.
In case of validation error, the NFVO would return the call after this step.
This assumes that the NFVO knows how to verify that the VNF Manager is up and running.
B.3.1.4 NFVO: request to VNF Manager to instantiate the VNF
The NFVO requests the VNF Manager to instantiate the VNF.
If the resource reservation (step 3) has been done beforehand, then the NFVO will use as input parameters the
instantiation data and the reservation information provided by VIM.
If the resource reservation has not been done, then the NFVO will use as input parameter the instantiation data
provided.
B.3.1.5 VNF Manager: request validation and processing
The VNF Manager executes the following steps:
1)
Verify validity of request to instantiate VNF.
2)
If needed, validate parameters passed for VNF instantiation for technical correctness.
3)
Any other VNF lifecycle specific validation, e.g. license check.
4)
Based on VNF lifecycle constraints, modify/complement the input parameters coming from the NFVO with
information from VNFD data and application specific constraints.
NOTE:
Those changes should be compatible with both the VNFD and the policies defined at NFVO level.
As an option, VNF Manager can ask for resource reservation either globally as a single request, concatenating the
information related to all VDUs in the same instantiation data structure or can issue to the NFVO multiple requests (one
per VDU). In all cases, VNF Manager will send the reservation request(s) to the NFVO that will process it (them) with
the VIM.
The VNF Manager then calls the NFVO for allocating the resources needed for this VNF instance. For sake of
simplicity, a single call is assumed, but one call per VDU might be possible.
B.3.1.6 NFVO: pre-allocation processing
When receiving a request to allocate resources for a VNF, the NFVO executes the following steps:
1)
If needed, validate parameters passed for VNF resource allocation for technical correctness (e.g. against
VNFD) and policy conformance.
2)
Location selection: The selection of where to locate a VNF instance could be based on the request, available
resources, the nature of the VNF, the Network Service(s) in which the VNF instance is participating in as well
as defined policies. Note that not all VDU instances will need to be placed in the selected location-it only
serves to provide the primary affinity/non-affinity point for the VNF instance.
3)
Dependency checking: Availability of all required external dependencies from the required location need to be
checked. If the VNF instance has any QoS requirements, it also needs to be verified that they can be met in the
selected location. Note that the QoS requirements could be on compute or network resources or on external
services on which the VNF instance is dependent.
ETSI
114
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
B.3.1.7 Orchestration: resource allocation (compute, storage and network)
and interconnection setup
The NFVO executes the following steps:
1)
Resource pool selection: The resource pool to be used needs to be selected. Note that this is not the same as the
VNF location. Multiple resource pools could exist in the same location or some VDU instances that are part of
a VNF instance may need to be located remotely from the rest.
2)
Request instantiation of the internal connectivity network: For a VNF that requires dedicated virtual networks
to interconnect VDU instances (networks that are only used as internal to the VNF instance) these virtual
networks need to be created.
3)
Requesting instantiation of the needed compute and storage resources from the infrastructure (Virtualised
Infrastructure Manager). Note that there might be multiple distributed VIMs.
4)
Attach instantiated VMs to internal connectivity network.
Each virtual NIC on the VMs hosting the VDUs that constitute the VNF instance need to be appropriately connected. If
there are particular QoS requirements on a network connection, the network may need to be properly configured to
support these.
At the end of this step, the NFVO acknowledges the completion of the resource and network allocation back to VNF
Manager and returns back information on the allocated resources and network connections.
Note that resource allocation request is likely to be asynchronous and depending on the implementation of the VIM, the
NFVO might have to poll for completion of resource allocation or wait for a notification from VIM.
ETSI
115
B.3.2
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
VNF instantiation flows with resource allocation done by
VNF Manager
B.3.2.1 VNF instantiation from EM
Figure B.10: VNF Instantiation from EM flow
NOTE:
For this flow to be valid, the EM has to consume the VNF Lifecycle Management interface exposed by
the VNFM over the VeNm-Vnfm reference point.
The main steps for VNF instantiation from EM are:
1.
EM requests to the VNF Manager instantiation of a new VNF in the infrastructure using the operation
Instantiate VNF of the VNF Lifecycle Management interface. It sends information on what VNF type shall be
instantiated.
2.
The VNF Manager requests granting to the NFVO to instantiate the VNF according to the information in the
VNFD (CPU, Memory, IP, etc.) using the operation Grant Lifecycle Operation of the VNF Lifecycle
Operation Granting interface.
3.
The NFVO checks resource request from VNFM against its capacity database for free resource availability.
4-5. The NFVO may otherwise optionally do resource reservation for the requested resources by using the Create
Resource Reservation operation over the Virtualised Resources Management interface.
6.
The NFVO responds to the VNF Manager request by sending the VIM Identifier, as well as the needed
reservation information if performed (step 4), to inform the VNF Manager where to instantiate the VNF as a
response to step 2.
ETSI
116
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
7.
The VNF Manager sends the request to create and start the VMs as appropriate and as instructed by the
NFVO, sending VIM Identifier and VM parameters using the operation Allocate Resource of the Virtualised
Resources Management interface.
8.
The VIM creates and starts the VMs and relevant networking resources, then acknowledges successful
operation to the VNF Manager.
9.
VNF Manager configures VNF data being specific for VNF instantiation using the add/create/set config object
operations of the VNF Configuration interface.
10.
VNF Manager notifies successful VNF instantiation back to the EM as response to request made in step 1.
11.
EM and VNF Manager add the new VNF as managed device.
12.
EM configures the VNF with application specific parameters.
13.
VNF Manager reports successful VNF instantiation to the NFVO using the VNF Lifecycle Change
Notification interface. The NFVO now is aware that the new VNF is instantiated in the infrastructure.
14.
VNF NFVO maps the VNF to the proper NFVI-PoP and Resource Pool.
B.3.2.2 VNF instantiation from NFVO
EM
VNF Manager
VIM
NFV Orchestrator
1. Instantiate VNF (VNF)
2. Instantiate VNF (VNF)
3. Grant Lifecycle Operation (VNF, Instantiation..)
4. Check free resources
are available (e.g. Pool,
Resource type, CPU,
Memory, etc.), optionally
reserve towards VIM.
5. (opt) Create Resource Reservation
7. ACK (VIM Id,)
6. Result of Reservation
8. Allocate Resource
9. ACK
VNF
10. Configure VNF (deployment Specific Parameters)
11. Notify VNF Instantiated
12. Add VNF as managed device
13. Configure VNF (App. Specific Parameters)
14. Notify VNF Instantiated
15. Map VNF to VIM and
Resource Pool
Figure B.11: VNF instantiation from NFVO flow
ETSI
117
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
The main steps for VNF instantiation from NFVO are:
1.
The NFVO receives a trigger to instantiate a VNF in the infrastructure (this can be a manual trigger or an
automatic service creation trigger request e.g. from the OSS/BSS) using the operation Instantiate VNF of the
VNF Lifecycle Management interface.
2.
NFVO requests to the VNF Manager instantiation of a new VNF in the infrastructure using the operation
Instantiate VNF of the VNF Lifecycle Management interface. VNF instantiation information is included.
3.
The VNF Manager requests granting to the NFVO to instantiate the VNF according to the information in the
VNFD (CPU, Memory, IP, etc.) using the operation Grant Lifecycle Operation of the VNF Lifecycle
Operation Granting interface.
4.
The NFVO checks resource request according to the information received in the granting request against its
capacity database for free resource availability.
5-6. The NFVO may otherwise optionally do resource reservation for the requested resources by using the Create
Resource Reservation operation over the Virtualised Resources Management interface.
7.
The NFVO responds to the VNF Manager request by sending the VIM Identifier, as well as other needed
reservation information if performed (step 5), to inform the VNF Manager where to instantiate the VNF as a
response to step 4.
8.
The VNF Manager sends the request to create and start the VMs as appropriate and as instructed by the
NFVO, sending VIM Identifier and VMs parameters using the operation Allocate Resource of the Virtualised
Resources Management interface.
9.
The VIM creates and starts the VMs and the networking resources then acknowledges successful operation to
the VNF Manager.
10.
VNF Manager configures VNF data being specific for VNF instantiation using the add/create/set config object
operations of the VNF Configuration interface.
11.
VNF Manager notifies the EM that a new VNF is created using the VNF Lifecycle Change Notification
interface.
12.
EM and VNF Manager add the new VNF as managed device.
13.
EM configures the VNF with application specific parameters.
14.
VNF Manager reports successful VNF instantiation to the NFVO as response to request made in step 3. The
NFVO now is aware that the new VNF is instantiated in the infrastructure.
15.
The NFVO maps the VNF to the proper VIM and resource pool.
B.4
VNF instance scaling flows
B.4.1
Detecting need to scale
VNF instance scaling is often the result of a service quality threshold being crossed - whether because service quality is
no longer acceptable, requiring expanding capacity or because service quality and utilization is such that capacity can
be contracted without affecting quality delivered.
About the source that will initiate the decision process, it may come from:
1)
VNF, when it embeds a monitoring function/threshold crossing detection and event notification. VNF may
send that event to EM and decision about actions may be implemented in EM and forwarded to VNF Manager.
VNF may send that event directly to the VNF Manager as well.
2)
VNF Manager, when reception of a single VNF or infrastructure event might be sufficient to detect the need to
scale, then information on the event to monitor and the corresponding scaling action might be provided in the
VNF Descriptor.
ETSI
118
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
3)
VIM, when VIM implements a monitoring function/threshold crossing detection and event notification. Event
would be about network congestion, number of sessions, etc. VNF Manager will listen to those events and
implement the decision about actions.
4)
EM when the monitoring function/threshold crossing detection and event notification is not in the VNF.
Decision about actions may be implemented in EM and forwarded to VNF Manager.
5)
OSS/BSS when the monitoring function/threshold crossing detection & event notification is not in the EM or
crosses several EM. OSS/BSS would be both the detector and decision point.
6)
OSS/BSS when this is a change management process/capacity planning process based on traffic projections for
example.
7)
Manual change triggered by an operator.
From the list above, the scaling use cases can be grouped in 3 categories:
1)
Auto-scaling, in which the VNF Manager monitors the state of a VNF instance and triggers the scaling
operation when certain conditions are met. For monitoring a VNF instance's state, it can for instance track
infrastructure-level and/or VNF-level events. Infrastructure-level events are generated by the VIM. VNF-Level
events may be generated by the VNF instance or its EM.
2)
On-demand scaling, in which a VNF instance or its EM monitor the state of a VNF instance and trigger a
scaling operation through explicit request to the VNF Manager.
3)
Scaling based on management request, where the scaling request is triggered by some sender (OSS/BSS or
operator) towards VNFM via the NFVO.
For sake of simplicity, the same flow will be used for both cases even if some simplification might be possible.
B.4.2
Determining scaling action
The nature of the VNF and the measurement results that cross their thresholds will generally indicate the type of change
required. These can be:
•
configuration changes to the VM (scale up, e.g. add CPU or memory);
•
add a new VDU instance (scale out);
•
shut down and remove instances (scale in);
•
release resources from existing instances (scale down);
•
increase available network capacity;
•
provide increased bandwidth (or other network changes).
Determining the scaling action may need to look beyond the VNF instance being examined. For example, solving a
quality issue with VNF instance-1 may require changes to VNF instance-2, which might be of a different VNF type. For
example:
•
The reason VNF instance-1 is performing poorly is because of congestion caused by VNF instance-2. If it is
costly to move VNF instance-1 but not costly to move VNF instance-2, the right solution is to move VNF
instance-2, even if the service quality it is delivering acceptable.
•
VNF instance-1 could have an external dependence on VNF instance-2. If VNF instance-2 can only adequately
support a given number of components, VNF instance-1 creation or scaling may require that VNF instance-2 is
proactively scaled.
Note that such use cases would most likely be detected by an OSS or an EM if both VNF types are managed by the
same EM.
ETSI
119
B.4.3
Scaling flow with resource allocation done by NFVO
Figure B.12: VNF instance scaling message flow
ETSI
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
120
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
In accordance with clause B.4.1 the sender can be the VNF manager, or OSS/BSS, or else be manually triggered by an
operator. The possible originators of the scaling request are listed in clause B.4.1.
In case the VNF Manager is the one issuing the scaling request, some of these flows might be simplified.
The main steps for the VNF instance scaling are:
1)
The NFVO receives the scaling request from the sender, e.g. OSS using the operation Scale VNF of the VNF
Lifecycle Management interface.
2)
The NFVO validates the request for policy conformance.
3)
NFVO finds the VNF Manager relevant for this VNF type.
4)
Optionally, NFVO runs a feasibility check of the VNF scaling request to reserve resources before doing the
actual scaling as described in clause B.3.1.1.
5)
The NFVO sends the scaling request to the VNF Manager, with the scaling data and, if step 4 has been done,
the reservation information using the operation Scale VNF of the VNF Lifecycle Management interface.
6)
The VNF Manager executes any needed preparation work: request validation, parameter validation. This might
include modifying/complementing the input scaling data with VNF lifecycle specific constraints. If step 4 was
done then the VNFM will skip step 6.
7)
The VNF Manager calls the NFVO for resource change using the operation Allocate Resource or Update
Resource or Scale Resource of the Virtualised Resources Management interface.
8)
NFVO requests from VIM allocation of changed resources (compute, storage and network) needed for the
scaling request using the operations Allocate Resource or Update Resource or Scale Resource of the
Virtualised Resources Management interface.
9)
VIM modifies as needed the internal connectivity network.
10) VIM creates and starts the needed new compute (VMs) and storage resources and attaches new instantiated
VMs to internal connectivity network.
11) Acknowledgement of completion of resource change back to NFVO.
12) NFVO acknowledges the completion of the resource change back to VNF Manager.
13) The VNF Manager configures the scaled VNF as necessary using the add/create/set config object operations of
the VNF configuration interface.
14) VNF Manager acknowledges the end of the scaling request back to the NFVO.
15) The NFVO acknowledges the end of the scaling request back to the requester.
In case the VNF Manager is issuing the scaling request, steps 1 to 3 of this flow and steps 1 to 3 of the check feasibility
flow will be skipped and step 4 of the feasibility flow will be replaced by an (optional) request from the VNF Manager
to the NFVO to grant the scaling request and optionally return the reserved resources information, using the VNF
lifecycle operation granting interface. Steps 5 and 6 of this flow will also be skipped in this case.
ETSI
121
B.4.4
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Scaling flows with resource allocation done by VNF
Manager
B.4.4.1 Automatic VNF expansion triggered by VNF performance
measurement results
Figure B.13: Automatic VNF expansion flow triggered by VNF performance measurement results
VNF expansion refers to the addition of capacity that is deployed for a VNF. Expansion may result in a scale out of a
VNF by adding VNFCs to support more capacity or may result in a scale-up of virtualised resources in existing
VNF/VNFCs. VNF expansion may be controlled by an automatic process or may be a manually triggered operation.
The following flow achieves automatic VNF expansion with the scale out example.
The main steps for the automatic VNF expansion are:
1.
The VNF Manager collects measurement results from the VNF (application specific) using the operation
Notify or Get performance measurement results of the VNF Performance Management interface.
2.
VNF Manager detects a capacity shortage that requires expansion (more resources).
3.
The VNF Manager requests granting to the NFVO for the VNF expansion based on the specifications listed in
the VNFD (CPU, Memory, IP, etc.) using the operation Grant Lifecycle Operation of the VNF Lifecycle
Operation Granting interface.
ETSI
122
4.
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
The NFVO takes scaling decision and checks resource request (CPU, Memory, IP, etc.) against its capacity
database for free resource availability.
5-6. The NFVO may otherwise optionally do resource reservation for the requested resources by using the Create
Resource Reservation operation over the Virtualised Resources Management interface.
7.
The NFVO grants the scale-out operation of the VNF to the VNF Manager and sends back sufficient
information to further execute the scaling operation.
8.
The VNF Manager sends the request to create and start the VMs as appropriate and as instructed by the
NFVO, sending VIM Identifier and VMs parameters using the operations Allocate Resource or Update
Resource or Scale Resource of the Virtualised Resources Management interface.
9.
The VIM creates and starts the VMs and the relevant networking resources, then acknowledges successful
operation to the VNF Manager.
10.
VNF Manager configures VNF data specific for VNF instantiation using the add/create/set config object
operations of the VNF Configuration interface.
11.
VNF Manager notifies the EM that an existing VNF is updated with additional capacity using the VNF
Lifecycle Change Notification interface.
12.
EM and VNF Manager update the VNF as managed device.
13.
EM configures the VNF with application specific parameters.
14.
VNF Manager reports successful VNF expansion to the NFVO using the VNF Lifecycle Change Notification
interface. The NFVO now is aware that the new VNF configuration is instantiated in the infrastructure.
15.
The NFVO maps the VNF to the proper VIM and resource pool.
ETSI
123
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
B.4.4.2 EM initiated VNF expansion
Figure B.14: EM initiated VNF expansion flow
NOTE:
For this flow to be valid, the EM has to consume the VNF Lifecycle Management interface exposed by
the VNFM over the VeNm-Vnfm reference point.
The main steps for the EM initiated VNF expansion are:
1.
Manual Operator's request or automatic event to expand the capacity of a virtual node (VNF).
2.
EM requests capacity expansion to the VNF Manager using the operation Scale VNF of the VNF Lifecycle
Management interface.
3.
The VNF Manager requests granting to the NFVO for the VNF expansion based on the specifications listed in
the VNFD (CPU, Memory, IP, etc.) using the operation Grant Lifecycle Operation of the VNF Lifecycle
Operation Granting interface.
4.
The NFVO takes scaling decision and checks resource request according to the VNF template (CPU, Memory,
IP, etc.) against its capacity database for free resource availability.
5-6. The NFVO may otherwise optionally do resource reservation for the requested resources by using the Create
Resource Reservation operation over the Virtualised Resources Management interface.
7.
The NFVO grants the scale-out operation of the VNF to the VNF Manager and sends back sufficient
information to further execute the scaling operations.
8.
The VNF Manager sends the request to create and start the VMs as appropriate and as instructed by the
NFVO, sending VIM Identifier and VMs parameters using the operations Allocate Resource or Update
Resource or Scale Resource of the Virtualised Resources Management interface.
9.
The VIM creates and starts the VMs and the relevant networking resources, then acknowledges successful
operation to the VNF Manager.
ETSI
124
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
10.
VNF Manager configures VNF data being specific for VNF instantiation using the add/create/set config object
operations of the VNF Configuration interface.
11.
VNF Manager notifies to the EM that an existing VNF has been updated as requested with additional capacity.
12.
EM and VNF Manager update the VNF as a managed device.
13.
EM configures the VNF with application specific parameters.
14.
VNF Manager reports successful VNF expansion to the NFVO using the VNF Lifecycle Change Notification
interface. The NFVO now is aware that the new VNF configuration is instantiated in the infrastructure.
15.
The NFVO maps the VNF to the proper VIM and resource pool.
B.4.4.3 Automatic VNF contraction triggered by VNF performance
measurement results
EM
VNF
NFV Orchestrator
VNF Manager
VIM
1. Notify performance measurement results
2. Contraction required
3. Grant Lifecycle Operation (VNF, Scale in..)
4. Scaling decision
5. ACK (Scale VNF)
6. Graceful termination of VNF component
7 . Release resource
8. ACK
10. Notify VNF update
9. Notify VNF update
11. Update VIM and
resource pool map
12. Update managed device
Figure B.15: Automatic VNF contraction flow triggered by VNF performance measurement results
VNF contraction refers to the removal of capacity from a deployed VNF. Contraction may result in scale-in of a VNF
by removing VNFCs to free resources in the NFVI, or may result in scale-down of computing and memory resource
from existing VNF/VNFCs. VNF contraction may be controlled by an automatic process or may be manually triggered
by operator intervention, it typically occurs upon load conditions relaxes below a configured threshold. The following
flow achieves automatic VNF contraction with the scale in example.
The main steps for the automatic VNF contraction are:
1.
The VNF Manager collects measurement results from the VNF (Application Specific) using the operation
Notify or Get performance measurement resultsof the VNF Performance Management interface.
2.
VNF Manager detects a capacity release opportunity and triggers contraction (release of resources).
ETSI
125
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
3.
The VNF Manager requests validation to the NFVO for the VNF contraction based on proper template (CPU,
Memory, IP, etc.) using the operation Grant Lifecycle Operation of the VNF Lifecycle Operation Granting
interface.
4.
The NFVO takes a scaling decision (e.g. based on configured policies).
5.
The NFVO grants the scale-in operation of the VNF to the VNF Manager.
6.
VNF gracefully terminates a VNF component (i.e. without affecting the ongoing service).
7.
Once the application is shut down (no more traffic is handled), the VNF Manager requests deletion of the
VM(s) to the VIM using the operation Release Resource of the Virtualised Resources Management interface.
8.
The VIM releases resources and acknowledges to the VNF Manager.
9.
VNF Manager reports successful VNF contraction to the NFVO using the VNF Lifecycle Change notification
interface. The NFVO now is aware that resources have been released by the VIM.
10.
VNF Manager notifies the EM that an existing VNF is updated with capacity release using the VNF Lifecycle
Change Notification interface.
11.
The NFVO updates the proper VIM and resource pool map.
12.
EM and VNF Manager update the VNF as a managed device.
B.4.4.4 EM initiated VNF contraction
Figure B.16: EM initiated VNF contraction flow
NOTE:
For this flow to be valid, the EM has to consume the VNF Lifecycle Management interface exposed by
the VNFM over the VeNm-Vnfm reference point.
The main steps for the EM initiated VNF contraction are:
1.
Manual Operator's request or automatic event to contract the capacity of a virtual node (VNF).
ETSI
126
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
2.
EM requests capacity release for the VNF to the VNF Manager using the operation Scale VNF of the VNF
Lifecycle Management interface.
3.
The VNF Manager requests validation to the NFVO for the VNF Contraction based on the specifications listed
in the VNFD (CPU, Memory, IP, etc.) using the operation Grant Lifecycle Operation of the VNF Lifecycle
Operation Granting interface.
4.
The NFVO takes a scaling decision (e.g. based on configured policies).
5.
The NFVO grants the scale-in operation of the VNF to the VNF Manager.
6.
VNF gracefully terminates a VNF component (i.e. without affecting the ongoing service).
7.
Once the application is shut down (no more traffic is handled), the VNF Manager requests deletion of the
VM(s) to the VIM using the operation Release Resource of the Virtualised Resources Management interface.
8.
The VIM releases resources and acknowledges to the VNF Manager.
9.
VNF Manager reports successful VNF contraction to the NFVO using the VNF Lifecycle Change Notification
interface. The NFVO now is aware that resources have been released by the VIM.
10.
VNF Manager acknowledges to the EM that an existing VNF has been updated as requested with capacity
release.
11.
The NFVO updates the proper VIM and resource pool map.
12.
EM and VNF Manager update the VNF managed device.
ETSI
127
B.5
VNF instance termination flows
Figure B.17: VNF instance termination message flow
ETSI
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
128
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
NFVO receives a request to terminate an existing VNF instance. This request might be sent from an OSS,
decommissioning of a VNF or part of an order, or might come from the VNF Manager when the need to terminate a
VNF is detected by VNF Manager itself or by the EM.
The main steps for VNF instance termination are:
1.
NFVO receives a request to terminate an existing VNF instance using the operation Terminate VNF of the
VNF Lifecycle Management interface.
2.
NFVO validates the request. It verifies the validity of the request (including sender's authorization) and
verifies that the VNF instance exists.
3.
NFVO calls VNF Manager to terminate of the VNF using the operation Terminate VNF of the VNF Lifecycle
Management interface.
4.
VNF Manager terminates the VNF. This step may include a graceful shutdown of the VNF possibly in
coordination with other management entities or the VNF itself.
5.
Once the VNF is terminated, VNF Manager acknowledges the completion of the VNF termination back to the
NFVO.
6.
Using resource information kept for this VNF instance, NFVO requests deletion of resources (compute,
storage and network) used by the various VDUs of the VNF instance using the operation Release Resource of
the Virtualised Resources Management interface.
7.
VIM deletes the internal connectivity network.
8.
VIM deletes the compute (VMs) and storage resources of the various VDUs of the VNF instance.
9.
Acknowledgement of completion of resource release back to NFVO.
10.
NFVO acknowledges the completion of the VNF instance termination using the VNF Lifecycle Change
Notification interface.
If the termination request is issued by the VNF Manager, then steps 1 to 3 are optional. If the termination request is
issued by the EM, Step 1 Terminate VNF Instance will be sent to VNF Manager, then, forwarded to NFV Orchestrator.
Steps 1 to 3 are optional.
B.6
NFV fault management
This clause provides examples of operational flows for fault management. While such flows apply to each VNF
lifecycle, their scope is broader than the lifecycle of a single VNF instance, because certain faults may impact multiple
VNFs and multiple NSs, and fault information may be further processed for purposes different than the lifecycle of the
VNF (e.g. data analytics, capacity/inventory management, SLA management, etc.).
The concept of fault and performance management is described in clause 4.5.
Fault information may be the result of several different sources of faults: physical infrastructure (i.e. physical NFVI
compute, storage, and networking related faults); virtualised infrastructure (e.g. VM-related faults), and application
logic (i.e. VNF instance related faults).
When fault information related to the same primary cause is issued by some or all of those multiple sources, it needs to
be correlated. In the NFV-MANO architecture, such correlation could happen in multiple places: the NFVO, the VNF
Manager, the EM and/or some OSS; the NFV architectural framework has the flexibility to support any of these
alternatives, as well as combinations of these alternatives. Once correlation point(s) are chosen, other functional blocks
have to forward the fault information to the targeted correlation point.
While multiple alternatives may be possible, it is not the intent of this clause to document all possible fault management
operational flows. The definition of the interfaces and policies controlling such interfaces allows for a large variety of
alternative flows to be implemented, as per Service Provider and vendor implementation agreements.
ETSI
129
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
In this clause, a single fault management flow alternative is presented. The flow illustrates collection of fault
information from multiple sources, different fault correlation points, and different fault resolution/correction points. The
flow is to be understood as a snapshot in the continuous process of fault management, which is not initiated by an ondemand trigger, but rather is a continuous cycle of monitoring functional blocks for fault information and reacting by
processing that information.
The sequence diagram below shows an example of NFV fault management with multiple options, depending on the
fault correlation point, and the fault resolution point. For simplification, VIM represents collectively the source of fault
information for all virtualised infrastructure resources.
Figure B.18: NFV fault management flow
The main steps of the NFV fault management flow are:
1.
Virtualised infrastructure faults (1.1) related to a specific VNF instance are received by the VNF Manager for
that VNF instance. Infrastructure faults not related to a specific VNF can also be received directly by the
NFVO (1.2 for virtualised infrastructure faults).
2.
VNF Manager may perform its own fault correlation for certain selected events.
3.
VNF Manager may forward correlated fault information regarding the VNF instance to different other fault
correlation points, primarily EM (3.1) and NFVO (3.2).
4.
VNF instance may send application layer faults to different correlation points, primarily the EM (4.1) and
VNFM (4.2). VNFM may further forward the fault information to NFVO (4.3).
5.
Fault correlation may happen at any of the fault correlation points identified. EM may perform fault
correlation to determine the root cause and the impact on the VNF (5.1). NFVO may perform fault correlation
to determine the root cause and the impact on the Network Service (5.2). To support some scenarios, the
NFVO may just map infrastructure faults to VNF and/or NS, without correlating with VNF application layer
faults, and forward the results to an OSS (see 6.3).
ETSI
130
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
6.
At the different fault correlation points, a fault resolution action may ensue. EM may forward correlated fault
information regarding the VNF instance to an OSS (6.1). EM may trigger a corrective action request towards
the VNF Manager (6.2). NFVO may forward correlated fault information to an OSS (6.3). NFVO may trigger
a corrective action towards the VNF Manager (6.4). VNF Manager may itself trigger a corrective action if
responsible for correlating certain events (6.5).
7.
Additional fault correlation and processing may happen in the OSS.
NOTE 1: As described in this clause, there is not necessarily a single event correlation and/or resolution point in the
NFV Framework. Event Correlation/Resolution can, and should be executed as soon as the event is fully
understood, and what action needs to be taken is known. The danger is that an action may be taken
prematurely and/or that conflicting actions may be taken in different places in the event management
architecture.
EXAMPLE:
In NFV the expectation is that certain event correlation and resolution will be performed in the
VIM, other in the NFVO, other in the VNF Manager, other in the EMS, other in an NMS/OSS depending on the nature of the event.
The corrective actions will be different, depending on event. Some corrective actions may in fact involve executing
some of the VNF instance lifecycle management flows (e.g. VNF scaling, VNF termination) or some of the Network
Service instance lifecycle management flows (e.g. Network Service termination).
NOTE 2: Some of the flows require the presence of a mechanism to set thresholds on particular metrics, and a
mechanism to allow a functional block to selectively subscribe to event notifications sent by the VIM to
the subscribed functional block, when the threshold regarding a specific metric related a virtualised
resource is reached. It is also required by some flows that a VIM may also forward events triggered by
physical infrastructure faults (e.g. NFVI-POP down).
ETSI
131
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Annex C (informative):
Network Service lifecycle management flows
C.1
Introduction
This annex represents a collection (non-exhaustive) of management flows related to the Network Service Lifecycle that
should be possible to implement by means of the interfaces described in the main part of the present document.
The following principles are used for all the flows in this clause:
•
The NFVO is the single point of access for all requests from the OSS to simplify the interfacing.
•
The NFVO handles lifecycles of Network Service and VNF Forwarding Graph.
•
The VNF Manager handles VNF lifecycle from an application point of view.
•
The NFVO has the end-to-end view of the resources being allocated across Network Services and VNFs by
VNF Managers, so all requests for resource allocation transit through the NFVO.
All the flows in this clause are informative, representing best practices for each of the lifecycle operation and have the
main goal of identifying needed interfaces as well as information needed on those interfaces.
At each step, in case of failure, an immediate return might happen. All the failure cases have not been shown on the
sequence diagrams or on the text for sake of simplicity.
The dotted-line arrows represent the return path of the request or in some specific cases, a notification. The return path
of a request can be a true reply in case of a request/reply exchange or a notification otherwise.
C.2
Network Service on-boarding flows
The use case diagram shown by figure C.1 provides the following use cases related to Network Services on-boarding:
•
On-board Network Service Descriptor.
•
Disable Network Service Descriptor.
•
Enable Network Service Descriptor.
•
Update Network Service Descriptor.
•
Query Network Service Descriptor.
•
Delete Network Service Descriptor.
ETSI
132
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
On-Boarding Network
Service Descriptor
Disable Network Service
Descriptor
Enable Network Service
Descriptor
Update Network Service
Descriptor
Top Package::External Actor
Query Network Service
Descriptor
Delete Network Service
Descriptor
Figure C.1: Use case diagram for Network Service on-boarding
C.2.1
On-board Network Service Descriptor flow
Network Service on-boarding refers to the process of submitting a Network Service Descriptor (NSD) to the NFVO to
be included in the catalogue.
Figure C.2: Network Service Descriptor on-boarding message flow
Note that the entity Sender above could be any entity that sends the Network Service Descriptor to the NFVO on behalf
of the operator; it can be the operator itself or a service design application or a vendor or an entity in the SP domain.
The main steps for Network Service on-boarding are:
1.
Network Service Descriptor is submitted to the NFVO for on-boarding the Network Service using the
operation On-board Network Service Descriptor of the Network Service Descriptor interface.
ETSI
133
2.
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
The NFVO processes the Network Service Descriptor including (not limited to):
a.
Validate the integrity and authenticity of NSD, The security information needed for validation may be
provided as part of NSD.
b.
Checking presence of VNF Package for the VNFs that are part of the Network Service.
c.
Checking for the existence of mandatory elements.
d.
Checking the presence of needed external interfaces required to provide the Network Service, in the VNF
Descriptors of the VNFs that are part of the Network Service.
3.
The NFVO notifies the catalogue for insertion of the Network Service Descriptor in the catalogue. The
catalogue will insert the NSD with its version number.
4.
The NFVO acknowledges the Network Service on-boarding.
C.2.2
Disable Network Service Descriptor flow
Disabling a Network Service Descriptor refers to the process of marking a Network Service Descriptor (NSD) as
disabled in the catalogue, so that it is not possible to instantiate Network Services with it any further.
Figure C.3: Network Service Descriptor disable message flow
The main steps for disabling a Network Service Descriptor are:
1.
A request to disable a Network Service Descriptor is submitted to the NFVO using the operation Disable
Network Service Descriptor of the Network Service Descriptor interface.
2.
The NFVO processes the request and checks if the NSD exists and is enabled.
3.
If the NSD is enabled, The NFVO notifies the catalogue to disable the Network Service Descriptor in the
catalogue.
4.
The NFVO acknowledges the Network Service Descriptor disable request.
C.2.3
Enable Network Service Descriptor flow
Enabling a Network Service Descriptor refers to the process of marking a Network Service Descriptor (NSD) as
enabled in the catalogue, so that it can be used to instantiate Network Services again.
ETSI
134
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Figure C4: Network Service Descriptor enable message flow
The main steps for enabling a Network Service Descriptor are:
1.
A request to enable a disabled Network Service Descriptor is submitted to the NFVO using the operation
Enable Network Service Descriptor of the Network Service Descriptor interface.
2.
The NFVO processes the request and checks if the NSD exists, is disabled, and is not marked as deletion
pending. Optionally, the NFVO may validate the integrity of the NSD.
3.
If the NSD is disabled and is not marked as deletion pending, The NFVO notifies the catalogue to enable the
Network Service Descriptor in the catalogue.
4.
The NFVO acknowledges the Network Service Descriptor enable request.
C.2.4
Update Network Service Descriptor flow
Network Service Descriptor update refers to the process of submitting a modified Network Service Descriptor (NSD) to
the NFVO to be included in the catalogue. This update might include creating/deleting new VNFFGs and/or new VLDs.
Figure C.5: Network Service Descriptor update message flow
The main steps for Network Service Descriptor update are:
1.
Modified Network Service Descriptor is submitted to the NFVO for the Network Service using the operation
Update Network Service Descriptor of the Network Service Descriptor interface.
ETSI
135
2.
3.
The NFVO checks if a NSD exists already for this NS and is not marked as deletion pending and if so,
processes the Network Service Descriptor including (not limited to):
a.
Validate the integrity and authenticity of NSD, The security information needed for validation may be
provided as part of NSD.
b.
Checking presence of VNF Package for the VNFs that are part of the Network Service.
c.
Checking for the existence of mandatory elements.
d.
Checking the presence of needed external interfaces required to provide the Network Service, in the VNF
Descriptor of the VNFs that are part of the Network Service.
The NFVO notifies the catalogue for insertion of a new version of the Network Service Descriptor in the
catalogue.
NOTE:
4.
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Some existing Network Services might still use the previous version of the NSD, so updating it will
create a new version in the catalogue.
The NFVO acknowledges the Network Service Descriptor update.
C.2.5
Query Network Service Descriptor flow
Querying Network Service Descriptor allows returning from the catalogue the information of the Network Service
Descriptors, including any related VNFFGD and VLD.
Figure C.6: Network Service Descriptor query message flow
The main steps for Network Service Descriptor query are:
1.
The NFVO receives a query request for the Network Service Descriptor using the operation Query Network
Service Descriptor of the Network Service Descriptor interface. One or more filter parameters may be included
in the Query operation to filter the Network Service Descriptors.
2.
The NFVO gets from the catalogue the Network Service Descriptors satisfying the filter conditions in details,
including any related VNFFGD and VLD.
3.
The NFVO returns the Network Service Descriptor details.
C.2.6
Delete Network Service Descriptor flow
Network Service Descriptor deletion refers to the process of asking the NFVO to delete a Network Service Descriptor
(NSD) from the catalogue.
Note that there might be multiple versions of the NSD, but the assumption is that the delete request remove all versions,
otherwise, it fails.
ETSI
136
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Figure C.7: Network Service Descriptor deletion message flow
The main steps for Network Service deletion are:
1.
Request to delete Network Service Descriptor is submitted to the NFVO using the operation Delete Network
Service Descriptor of the Network Service Descriptor interface.
2.
The NFVO checks if the Network Service Descriptor is disabled and in use. If NSD is not disabled, the request
is rejected.
3.
If NSD is disables and not in use, then ask the catalogue to remove all versions of the NSD. The catalogue will
then remove them.
4.
If NSD is disabled and still in use, the NFVO set the NSD in deletion pending.
5.
The NFVO acknowledges Network Service Descriptor deletion request.
NOTE:
If the NSD is in deletion pending, the NFVO will later check if there is any NS instance using the NSD
during the NS instance termination process. If there is no NS instance using the NSD any more, the
NFVO asks the catalogue to remove the corresponding version(s) of the NSD.
ETSI
137
C.3
Network Service instantiation flows
Figure C.8: Network Service instantiation message flow
ETSI
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
138
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
NFVO receives a request to instantiate a new Network Service. This request might come from an OSS, receiving an
order for Network Service instantiation.
The NFVO is the single point of access for Network Service instantiation simplifying interfacing with OSS.
When creating a Network Service, several variants might be possible:
a)
No VNF instance needed for this Network Service exists, so the Network Service instantiation will include the
instantiation of the needed VNF instances.
b)
All needed VNF instances might already be instantiated: in this case, Network Service instantiation would
only deal with the interconnection of the VNF instances.
c)
A combination of the above where some VNF instances might exist and some might need to be created and
some of the network connectivity between the VNFs may already exist, and it only may need to be extended.
The instantiation flows below covers variants (c), thus checking if the needed VNF instances exist.
No assumption is made either on whether VNFs can be shared or not between Network Services.
The flow below assumes a single VNF Forwarding Graph for the Network Service. If multiple VNF Forwarding Graphs
are defined for a Network Service, then steps 3 to 13 would need to be repeated for each VNF Forwarding Graph.
The main steps for Network Service instantiation are:
1.
NFVO receives a request to instantiate a new Network Service using the operation Instantiate Network Service
of the Network Service Lifecycle Management interface.
2.
NFVO validates the request, both validity of request (including validating that the sender is authorized to issue
this request) and validation of the parameters passed for technical correctness and policy conformance. In case
the Network Service contains multiple VNF Forwarding Graphs, policy rules might result in only a subset
being valid for a given Network Service instance.
3.
For each VNF instance needed in the Network Service, the NFVO checks with the VNF Manager if an
instance matching the requirements exists already using the operation Query VNF of the VNF Lifecycle
Management interface. If such a VNF instance exists, it will be used as part of the Network Service.
NOTE 1: It requires NFVO, if needed, to find the corresponding VNF Manager and to instantiate it if it does not
exist. For sake of simplicity, those steps have been collapsed.
4.
Optionally, NFVO runs a feasibility check of the VNF interconnection setup. Steps 3 to 5 constitute the
feasibility check of the request. Step 3 consists of the following sub-steps:
a.
NFVO requests to VIM availability of network resources needed for the VNF Interconnection and
reservation of those resources using the operation Create Resource Reservation of the Virtualised
Resources Management interface. Note that some of the network connectivity between the VNFs might
already exist.
b.
VIM checks the availability of network resources needed for the VNF Interconnection and reserves them.
c.
VIM returns result of reservation back to NFVO.
5.
Optionally, once the list of VNF instances to be provisioned is known and assuming it is not empty, the NFVO
validates if resources are available to honour the VNF instantiation requests and if so, reserves them using the
operation Create Resource Reservation of the Virtualised Resources Management interface.
6.
NFVO requests from VIM instantiation of the network connectivity using the operations Allocate Resource or
Update Resource from the Virtualised Resources Management interface. Note that some of the network
connectivity between the VNFs might already exist and might only need to be extended.
7.
VIM instantiates the connectivity network needed for the Network Service.
8.
VIM acknowledges completion.
ETSI
139
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
9.
Assuming the list of VNF instances to be provisioned is not empty, the NFVO instantiate the new VNF
instances needed. This is done by calling the "Instantiate VNF" request using the operation Instantiate VNF of
the VNF Lifecycle Management interface as illustrated in clause 7.2, VNF Instantiation.
10.
Once all VNF instances are available and for the VNFs not already connected, NFVO requests VIM to connect
them together using the operations Allocate Resource or Update Resource from the Virtualised Resources
Management interface. It includes:
a.
Requesting VIM to connect external interfaces of each VNFs.
b.
Requesting VIM to attach needed VDUs (VMs) to the Network Service's connectivity network.
11.
VIM connects needed VDUs (VMs) to the connectivity network.
12.
VIM acknowledges completion.
13.
If needed, NFVO requests Network Manager to connect VNF external interfaces to physical network function
interfaces.
NOTE 2: The Network Manager can be an OSS, an NMS, an EM, or a WIM.
14.
NFVO acknowledges the completion of the Network Service instantiation.
NOTE 3: Connection to physical network functions is managed by the Resource Orchestration function of the
NFVO.
C.4
Network Service instance scaling
The NSD can be on boarded with 1 or more deployment flavours each having their resource requirements in terms of
number of VNF instances, interconnectivity, links, etc.
C.4.1
Network Service instance scale-out
The following example illustrates how Network Service instance scale-out works. The assumption is that 2. The
Network Service has been instantiated based on a flavour-A(1* VNF-A +2*VNF-B). At that point:
a)
The NS can be scaled based on auto-scaling policy (in the NSD) to a different NS deployment flavour-B
(2*VNF-A + 2*VNF-B, which is described in the same on boarded NSD).
b)
The Sender can manually initiate the NS scaling from the instantiated flavour -A to the deployment flavour-B;
For VNF-A it need to instantiate a new instance, and for VNF-B it need to scale those existing instances.
ETSI
140
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Figure C.9: Network Service scaling out message flow
NFVO receives a request to scale a Network Service instance. This request might come from an OSS, receiving an
order for Network Service instance scaling.
The NFVO is the single point of access for Network Service lifecycle simplifying interfacing with OSS.
1.
The Sender requests the NS to be scaled out to a new deployment flavour which is already present in the
preloaded NSD.
2.
NFVO validates the request, both validity of request (including validating that the sender is authorized to issue
this request) and validation of the parameters passed for technical correctness and policy conformance. NFVO
correlates the request with the on boarded NSD in the NS catalogue.
3.
(Optional) Check feasibility:
4.
a)
NFVO would check feasibility of scaling the relevant member VNFs to reflect the new flavour.
b)
Each VNF Manager would process this request and determine if the VNF can be scaled out to the new
flavour.
c)
VNF Manager(s) returns the result of the feasibility check on the VNF scaling out.
Scaling-out a NS involves scaling-out its constituent VNFs. It is foreseen that a VNF can be scale-out in two
different ways either by allocating more resources to VNF instance or by instantiating a new VNF instance. It
is foreseen to have these preferences documented as scaling mechanism in related VNFDs. On receiving the
scale-out request NFVO will identify, based on related NSD, the VNF instance(s) needed to be scale-out and
the related VNF scaling mechanism. Depending on which scaling mechanism to perform, NFVO will either
execute step 5 or step 6 or even both. The new VNF instance will be created (Step 6) in case the existing VNF
Instance has already scaled-out to its maximum capabilities, e.g. due to their its ability of automatic scaling.
ETSI
141
5.
VNF instance scaling flow as provided in clause B.4.
6.
VNF instantiation flow as provided in clause B.3.
7.
The NFVO would request VIM to allocate the changed resources:
8.
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
a)
NFVO would request VIM to allocate the changed resources (such as interconnectivity between VNFs
required by the new deployment flavour as mandated by the VNFFGDs and VLDs).
b)
The VIM would allocate the interconnectivity accordingly.
c)
VIM would return the result of the operation to the NFVO.
The NFVO acknowledges the end of the scaling request back to the requester.
C.4.2
Network Service instance scale-in
The following example illustrates how Network Service instance scale-in works. The assumption is that the Network
Service has been instantiated based on a flavour-B (2* VNF-A +2*VNF-B). At that point:
a)
The NS can be scaled based on auto-scaling policy (in the NSD) to a different NS deployment flavour-A
(1*VNF-A + 2*VNF-B, which is described in the same on boarded NSD).
b)
The Sender can manually initiate the NS scaling from the instantiated flavour -B to the deployment flavour-A;
For VNF-A it need to terminate an existing instance, and for VNF-B it need to scale in those existing
instances.
Figure C.10: Network Service scaling in message flow
ETSI
142
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
NFVO receives a request to scale in a Network Service instance. This request might come from an OSS, receiving an
order for Network Service instance scaling in.
The main steps for Network Service scaling in are:
1.
The Sender requests the NS to be scaled in to a new deployment flavour which is already present in the
preloaded NSD.
2.
NFVO validates the request, both validity of request (including validating that the sender is authorized to issue
this request) and validation of the parameters passed for technical correctness and policy conformance.
3.
(Optional) Check feasibility:
a)
NFVO would check feasibility of scaling the relevant member VNFs to reflect the new flavour.
b)
Each VNF Manager would process this request and determine if the VNF can be scaled in to the new
flavour.
c)
VNF Manager(s) returns the result of the feasibility check on the VNF scaling in.
4.
Scaling-in a NS involves scaling-in its constituent VNFs. It is foreseen that a VNF can be scale-in in two
different ways either by revoking allocated resources to VNF instance or by terminating an entire VNF
instance. It is foreseen to have these preferences documented as scaling mechanism in related VNFDs. On
receiving the scale-in request NFVO will identify, based on related NSD, the VNF instances needed to be
scale-in and the related VNF scaling mechanism. Depending on which scaling mechanism to perform, NFVO
will either execute step 5 or step 6 or even both. The existing VNF instance will be terminated (Step 6) in case
it has already scaled-in to its minimum capabilities, e.g. due to its ability of automatic scaling.
5.
VNF instance scaling flow as provided in clause B.4.
6.
VNF instance termination flow as provided in clause B.5.
7.
The NFVO would request VIM to modify or delete the changed resources:
8.
C.5
a)
NFVO would request VIM to modify or delete the changed resources (such as interconnectivity between
VNFs required by the new deployment flavour as mandated by the VNFFGDs and VLDs).
b)
The VIM would modify or delete the interconnectivity accordingly.
c)
VIM would return the result of the operation to the NFVO.
The NFVO acknowledges the end of the scaling request back to the requester.
Network Service instance update flows due to VNF
instance modification
Network Service instance update due to VNF Instance modification is about replacing existing VNF Instances with new
Instances by on boarding a new VNFD instantiating a VNF Instance out of it and then updating an existing VNF
Forwarding Graph with the instantiated VNF for the Network Service.. As a new VNFD is required, the new Instance
may differ in terms of deployment (e.g. NFVI Resource requirements, external interfaces) and operational (e.g. lifecycle
management) characteristics of the VNF, basically all what can be present in VNFD may change. Since a VNF Instance
modification may change the requirements/dependencies on other VNF, the process of modifications may require
changing dependent VNFs prior to the target VNF.
The following figure shows the before and after picture of this kind of Network Service instance update. It shows that
VNF X, in NS A, is identified to be modified in terms of its storage requirements and the external interfaces. In order to
honour the changed storage requirement the resources assigned to VNF Z are upgraded.
ETSI
143
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Figure C.11: Before and after a VNF instance update
Sender
NFVO
VIM
Catalog
1. Decide to update NS
2. Update Request
3. Validation
4. Resources and dependencies check
5. VNF Instantiate flow
6. Update VNFFG
7. ACK Update
8. Terminate old VNFs
9. Update VNF, Service catalogue
Figure C.12: NS instance update due to VNF instance modification flow
The main steps for the NS Update due to VNF instance modification are:
1.
Sender makes a decision to update a NS by modifying one more constituent VNF Instances.
2.
Sender sends an update request to the NFVO to update a particular Network Service instance using Update
Network Service operations of Network Service lifecycle management interface. The request will include:
a.
Identification of the existing Network Service instance that needs to be updated.
b.
Identification of the existing VNFD whose Instances (comprising in the Network Service) need to be
updated.
c.
A reference to the on-boarded new VNFD for the VNF to be updated.
NOTE 1: In case a new VNFD is provided it should be on-boarded before executing modify.
d.
3.
Optionally, dependent VNFDs identification and requirements (e.g. version, resources, etc.).
The NFVO validates the request including validation that the sender is authorized to issue this request and
validation of the parameters passed for technical correctness and policy conformance.
ETSI
144
4.
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
NFVO identifies the instances of the dependent VNFs according the dependent VNFDs identification provided
and analyses or validates the impact of the Network Service update including checking dependencies (e.g.
based on version, existence, resources, etc.) with other VNF Instances (as stated in related NSD) and querying
the resources assigned to them (as specified in the old VNFD), to check if the capabilities (e.g. version,
resources)assigned to the dependent VNF Instances are enough to honour the dependencies. NFVO may find
that the capabilities of dependent VNF Instances cannot support the requirements of the target VNF Instance,
e.g. the version of the dependent VNF is lower then what is required in the new VNFD. The NFVO will then
create a list of VNFs to be modified considering the dependencies with dependent VNFs. This may include
modifying dependent VNFs Instances prior to modifying the target VNF Instance (identified according to the
new VNFD identification provided in the request), e.g. in terms of assigned resources, versions, etc.
NOTE 2: VNF Package will be considered on-boarded beforehand. That may require related NSD to be updated
accordingly.
5.
To modify the VNFs identified above, the NFVO initiates VNF instantiation flow for each VNF to be
modified providing instantiation data using the operation "Instantiate VNF" of the VNF Lifecycle
Management interface. This includes identifying the VNFD for the dependent VNF instances to be used to
update the instance of dependent VNF. In case several VNF instances need to be modified and errors occur
during single VNF instance modification, roll-back mechanisms are required in order to revert back the VNF
instances which have been already modified, i.e. revert to the old VNF instance. These procedures require
interaction with corresponding VNF Managers while instantiating and terminating the VNF instances.
NOTE 3: This flow example only shows the steps corresponding to the instantiation and termination of modified
VNF instances. The additional steps for modifications on other dependent VNF instances are not shown.
6.
Once all new VNF instances are available, NFVO updates the VNF forwarding graphs which includes
applying new VNF instances to the forwarding graph in place of the old VNF instances. The flows of the VNF
Forwarding Graph update process are described in clause C.7.
7.
NFVO acknowledges the update.
8.
NFVO monitors old VNF and terminates them when appropriate by calling VNF Termination flow.
9.
NFVO updates the NFV Instances repository to reflect the existence new VNF Instances just created.
ETSI
145
C.6
Network Service instance termination flows
Figure C.13: Network Service termination message flow
ETSI
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
146
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
NFVO receives a request to terminate a Network Service instance. This request might come from an OSS, receiving an
order for Network Service instance termination.
The NFVO is the single point of access for Network Service termination simplifying interfacing with OSS.
When terminating a Network Service instance, several variants might be possible:
a)
Some VNF instances were already instantiated before the instantiation of the Network Service. In this case,
those VNFs should not be removed.
b)
All affected VNF instances contribute to the Network Service that needs to be terminated and were created
when instantiating the Network Service. In this case, all those VNF instances need to be terminated, and the
interconnectivity between these VNF instances need to be removed.
c)
Some VNF instances are contributing to other Network Service instances. In this case, only those VNF
instances that do not contribute to other Network Services instances need to be terminated, and the
interconnectivity between them needs to be removed, leaving the other VNF instances in place and the
interconnectivity between them intact.
The flow below assumes a single VNF Forwarding Graph for the Network Service instance. If multiple VNF
Forwarding Graphs are defined for a Network Service instance, then steps 6 to 8 would need to be repeated for each
VNF Forwarding Graph.
Note that it is assumed that VNF Forwarding Graphs are not shared between Network Service instances.
The main steps for termination of a Network Service instance are:
1.
NFVO receives a request to terminate a Network Service instance using the operation Terminate Network
Service of the Network Service Lifecycle Management interface.
2.
NFVO validates the request. It verifies the validity of the request (including sender's authorization) and
verifies that the Network Service instance exists.
3.
NFVO requests VNF Manager to terminate any VNF instances that were instantiated along with the Network
Service instantiation provided they are not used by another Network Service. This is done by calling the
"Terminate VNF instance" request using the operation Terminate VNF of the VNF Lifecycle Management
interface as illustrated in clause 7.2.4, VNF instance termination.
4.
VNF Manager terminates the required VNF. This step might include some graceful termination of the VNFs
involved possibly in coordination with other management entities or the VNFs themselves.
5.
Once the VNFs are terminated, the VNF Manager acknowledges completion of the termination request back to
the NFV Orchestrator.
6.
Using information kept for this Network Service instance, NFVO requests deletion of network connectivity for
this Network Service instance using the operations Release Resource or Update Resource of the Virtualised
Resources Management interface. Note that some network connectivity might have been present before the
instantiation of the Network Service. This connectivity will not be deleted.
7.
VIM deletes the connectivity network for this Network Service instance.
8.
VIM acknowledges the completion of resource deletion back to NFVO.
9.
If needed, NFVO requests Network Manager the removal of connections to physical network function
interfaces.
NOTE:
The Network Manager can be an OSS, an NMS or an EM.
10.
Network Manager acknowledges completion of the removal of connections.
11.
NFVO acknowledges the completion of the Network Service instance termination. If the NSD is in deletion
pending and there is no Network Service instance using it any more, the NFVO asks the catalogue to remove
the corresponding version(s) of the NSD. The catalogue will then remove it.
ETSI
147
C.7
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
VNF Forwarding Graph lifecycle management flows
The interface "Network Service lifecycle management" defined in clause 7.1.2 describes the operations for the VNF FG
lifecycle management.
This clause provides the flows of these operations for the VNF FG lifecycle management:
•
Create VNF Forwarding Graph;
•
Update VNF Forwarding Graph;
•
Query VNF Forwarding Graph; and
•
Delete VNF Forwarding Graph.
The Network Service instance needs to be updated when any members (e.g. VNF or VL) composing this Network
Service is changed. Therefore, the instantiation of a new VNF and modification of the information on VLR are needed
for updating the Network Service. Each VNF FG is formed by multiple Connection Points and possibly a Network
Forwarding Path (NFP) which indicates a traffic flow of the Network Service. The VNF FG needs to be updated when
any VNF FG members (e.g. Connection Points or NFP) is changed. Therefore, the modification of VNF FG Record and
update of NFP are needed for updating the VNF FG.
C.7.1
Create VNF Forwarding Graph
This flow refers to the process of creating a new VNF Forwarding Graph for an existing Network Service. Note that the
corresponding Network Service has been instantiated already with the needed VNFs and VLs.
Figure C.14: VNFFG creation message flow
The main steps for creating a VNF Forwarding Graph are:
1.
The NFVO receives a request to create a new VNF Forwarding Graph for an existing Network Service using
the operation Create VNFFG of the Network Service Lifecycle Management interface.
2.
The NFVO validates the request including validation that the sender is authorized to issue this request and
validation of the parameters passed for technical correctness and policy conformance.
3.
The NFVO requests the allocation of the necessary network resources to set up the VNF Forwarding Graph
using the operation Allocation Resource of the Virtualised Resources Management interface.
4.
The VIM assigns the needed network resources for the VNF Forwarding Graph.
5.
The VIM returns result of resource allocation back to the NFVO.
6.
The NFVO returns result of VNF Forwarding Graph creation back to the Sender.
ETSI
148
C.7.2
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Update VNF Forwarding Graph
This flow refers to the process of updating an existing VNF Forwarding Graph for a Network Service. For this, the
information of VNF Forwarding Graph Record and NFP associated to the VNF Forwarding Graph are updated. Note
that this update process assumes that a new VNF or new VNFs might already be instantiated to be incorporated into an
existing VNF Forwarding Graph.
Figure C.15: VNFFG update message flow
The main steps for updating a VNF Forwarding Graph are:
1.
The NFVO receives a request to update a VNF Forwarding Graph for an existing Network Service using the
operation Update VNFFG of the Network Service lifecycle management interface.
2.
The NFVO validates the request including validation that the sender is authorized to issue this request and
validation of the parameters passed for technical correctness and policy conformance.
3.
The NFVO requests to allocate necessary resources and network connectivity to update the VNF Forwarding
Graph using the operation Allocation Resource of the Virtualised Resources Management interface.
4.
The VIM assigns the needed network resources for the VNF Forwarding Graph.
5.
The VIM returns result of resource allocation back to the NFVO.
6.
The NFVO requests to the VIM to update NFP using the operation Update Network Forwarding Path of the
Network Forwarding Path rule management interface.
7.
The VIM processes the NFP update and performs the required configuration.
8.
The VIM acknowledges the NFVO the NFP update.
9.
The NFVO updates the VNF Forwarding Graph Record to reflect the information of new VNFs.
10.
The NFVO returns result of VNF Forwarding Graph update back to the Sender.
ETSI
149
C.7.3
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Query VNF Forwarding Graph
This flow refers to the process of retrieving the information of an existing VNF Forwarding Graph for a Network
Service.
Figure C.16: VNFFG query message flow
The main steps for querying a VNF Forwarding Graph are:
1.
The NFVO receives a request to query a VNF Forwarding Graph for an existing Network Service using the
operation Query VNFFG of the Network Service Lifecycle Management interface.
2.
The NFVO validates the request including validation that the sender is authorized to issue this request and
validation of the parameters passed for technical correctness and policy conformance.
3.
The NFVO returns the detailed information of VNF Forwarding Graph from the Network Service/ VNF
Forwarding Graph repository to the Sender.
C.7.4
Delete VNF Forwarding Graph
This flow refers to the process of deleting an existing VNF Forwarding Graph which belongs to an existing Network
Service.
Figure C.17: VNFFG delete message flow
ETSI
150
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
The main steps for deleting a VNF Forwarding Graph are:
1.
The NFVO receives a request to delete a VNF Forwarding Graph for an existing Network Service using the
operation Delete VNFFG of the Network Service Lifecycle Management interface.
2.
The NFVO validates the request including validation that the sender is authorized to issue this request and
validation of the parameters passed for technical correctness and policy conformance.
3.
The NFVO requests the release of the network resources assigned to this VNF Forwarding Graph using the
operation Release Resource of the Virtualised Resources Management interface.
4.
The VIM de-allocates the network resources used by the VNF Forwarding Graph.
5.
The VIM returns result of resource de- allocation back to the NFVO.
6.
The NFVO returns result of VNF Forwarding Graph deletion back to the Sender.
ETSI
151
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Annex D (informative):
Orchestration flows
D.1
Introduction
This annex represents a collection (non-exhaustive) of Orchestration flows that should be possible to implement by
means of the interfaces described in the main part of the present document.
D.2
NFVI-PoP setup and configuration
VIM 1
VIM 2
VIM n
1. Resources setup and
configuration in the related
NFVI
4. Resources setup and
configuration in the related
NFVI
7. Resources setup and
configuration in the related
NFVI
NFV Orchestrator
2. Notify Resources Catalogue Changes
3. ACK
5. Notify Resources Catalogue Changes
6. ACK
NFVO builds up the
NFVI view, and
related resources
map
8. Notify Resources Catalogue Changes
9. ACK
Figure D.1: NFVI-PoP setup and configuration flow
1.
Operator setup the NFVI-PoP configuration in the VIM including processing and storage resources.
2.
The VIM notifies the NFVO that one or more NFVI-PoPs have been created including all the relevant
information (VIM Identifier, NFVI-PoP, location, resource list, capacity and specifications) using the
operation Notify Resources Catalogue Changes of the Resources Catalogue Management interface.
3.
The NFVO acknowledges the NFVI-PoP creation.
4 to 9.
Step 1 to 3 are repeated for each of the VIM (if more than one are deployed) and any time a NFVI-PoP is
setup.
The NFVO has all the required information to build the NFVI view including NFVI-PoPs and resources mapping.
ETSI
152
D.3
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Resources provisioning/de-provisioning
VIM n
NFV Orchestrator
1. Resources
Added/Resources
removed/NFVI unavailable
2. Notify Resources Catalogue Changes
3. Add/Remove Resources to
related NFVI/Remove NFVI
4. ACK
Figure D.2: Resources provisioning/de-provisioning flow
1.
HW resources are provisioned/de-provisioned to a NFVI-PoP in the VIM n, or a NFVI-PoP is deleted from the
VIM n.
2.
The VIM notifies the NFVO that more/less resources are now available including all the relevant information
(VIM Identifier, NFVI-PoP, Resource list, capacity and specifications) using the operation Notify Resources
Catalogue Changes of the Resources Catalogue Management interface.
3.
The NFVO adds resources to the proper NFVI-PoP and updates the mapping.
4.
The NFVO acknowledges resource provisioning and update.
This procedure is executed every time resources are provisioned/de-provisioned in any of the NFVI-PoP configured in
the infrastructure, or the NFVI-PoP is deleted. The NFVO has all the required information to update the NFVI view
including NFVI-PoPs and resources mapping.
ETSI
153
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Annex E (informative):
VNFD/NSD representations using TOSCA
E.1
Describing IMS MRF with TOSCA
This clause provides an illustrative description of IMS MRF using the Topology and Orchestration Specification for
Cloud Applications (TOSCA) Service Template.
E.1.1
TOSCA meta-model
TOSCA specification provides a language to describe service components and their relationships using a service
topology, and it provides for describing the management procedures that create or modify services using orchestration
processes. The combination of topology and orchestration in a Service Template describes what is needed to be
preserved across deployments in different environments to enable interoperable deployment of cloud services and their
management throughout the complete lifecycle (e.g. scaling, patching, monitoring, etc.) when the applications are
ported over alternative cloud environments.
Figure E.1 shows the relevance of TOSCA meta-model with NFV descriptors modelling requirements. It provides the
mapping between the IMS MRF VNFD/NSD requirements (in red) and the corresponding TOSCA element (in black) to
be used to fulfil that requirement. It also shows what all TOSCA CSAR Package may include which imply its relevance
to VNF Package.
Figure E.1: Mapping of IMS MRF VNFD/NSD to TOSCA elements
E.1.2
IMS MRF representation in TOSCA
This clause provides information about how TOSCA Service Template can be used to describe VNFD/NSD in a
standardized format.
E.1.2.1 IMS MRF NSD
01
02
03
04
05
06
07
<Definitions name="IMSMRFServiceTemplateDefinition"
targetNamespace="http://www.etsi.org/sample">
<Tags>
<Tag name="author" value="[email protected]"/>
</Tags>
<Types>
<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema"
ETSI
154
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
08
elementFormDefault="qualified"
09
attributeFormDefault="unqualified">
10 <!-- This element may be used to model general information elements related with VNF e.g
"vendor", "VNF_ID", etc.-->
11
<xs:element name="ApplicationProperties">
12
<xs:complexType>
13
<xs:sequence>
14
<xs:element name="Vendor" type="xs:string"/>
15
<xs:element name="VNFD" type="xs:string"/>
16
</xs:sequence>
17
</xs:complexType>
18
</xs:element>
19
20 <!-- Level-1 element describing IMS MRF NFV Service.-->
21
<ServiceTemplate id="IMSMRFServiceTemplate">
22 <!-- This element describe the Topology of MRF NFV Service including MRB and MRF C+P and the
relationship (dependencies) between them.-->
23
<TopologyTemplate id="IMSMRFService">
24 <!-- Element describing MRB VNF.-->
25
<NodeTemplate id="MyMRB"
26
name="My MRB"
27
nodeType="abc:MRBNode">
28 <!-- VNF related properties e.g VNFC_no,-->
29
<Properties>
30
<ApplicationProperties>
31
<Vendor>VendorX</Vendor>
32
<VNFD>IMS_MRB_x2</VNFD>
33
</ApplicationProperties>
34
</Properties>
35
<NodeTemplate/>
36 <!-- Element describing MRF VNF.-->
37
<NodeTemplate id="MyMRF"
38
name="My MRF"
39
nodeType="abc:MRFNode"/>
40 <!-- Dependencies between MRB and MRF i.e for a MRF to work an MRB is needed.-->
41
<RelationshipTemplate id="MyDeploymentRelationship"
42
relationshipType="dependOn">
43
<SourceElement id="MyMRF"/>
44
<TargetElement id="MyMRB"/>
45
</RelationshipTemplate>
46
47
</TopologyTemplate>
48
49
<Plans>
50
51
<Plan id="InstantiateMRB"
52
planType="http://docs.oasis53
open.org/tosca/ns/2011/12/PlanTypes/InstantiatePlan"
54
planLanguage="http://docs.oasis55
open.org/wsbpel/2.0/process/executable">
56
<PlanModelReference reference="prj:InitiateMRB"/>
57
</Plan>
58
59
<Plan id="InstantiateMRF "
60
planType="http://docs.oasis61
open.org/tosca/ns/2011/12/PlanTypes/TerminationPlan"
62
planLanguage="http://docs.oasis63
open.org/wsbpel/2.0/process/executable">
64
<PlanModelReference reference="prj:InitiateMRF"/>
65
</Plan>
66
</Plans>
67
68
</ServiceTemplate>
69
70
<NodeType name="MRBNode">
71
<documentation xml:lang="EN">
72
A reusable definition of a MRB node type representing an
73
MRB VNF that can be deployed.
74
</documentation>
75
<NodeTypeProperties element="ApplicationProperties"/>
76
<Interfaces>
77
<Interface name="InstantiationInterface">
78
<Operation name="InstantiateMRB">
79 <!-- These can be the runtime information provided at MRB Instantiation -->
80
<InputParameters>
81
<InputParamter name="InstanceName"
82
type="xs:string"/>
83
<InputParamter name="NumberOfInstance"
ETSI
155
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
84
type="xs:integer"/>
85
</InputParameters>
86
</Operation>
87
</Interface>
88
</Interfaces
89
</NodeType>
90
91
<NodeType name="MRFNode"
92
targetNamespace="http://www.example.com/sample">
93
<NodeTypeProperties element="ApplicationProperties"/>
94
<Interfaces>
95
<Interface name="InstantiationInterface">
96
<Operation name="InstantiateMRB"/>
97
<InputParameters>
98
<InputParamter name="InstanceName"
99
type="xs:string"/>
100
<InputParamter name="NumberOfInstance"
101
type="xs:integer"/>
102
</InputParameters>
103
</Interface>
104
</Interfaces>
105
</NodeType>
106
107
<RelationshipType name="dependOn">
108
<documentation xml:lang="EN">
109
A reusable definition of relation that expresses dependencies between two VNF in NFV
Service..
110
</documentation>
111
</RelationshipType>
112
113 </Definitions>
E.1.2.2 MRB VNFD
01 <Definitions name="MRBServiceTemplateDefinition"
02
targetNamespace="http://www.etsi.org/sample">
03
<Tags>
04
<Tag name="author" value="[email protected]"/>
05
</Tags>
06
<Types>
07
<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema"
08
elementFormDefault="qualified"
09
attributeFormDefault="unqualified">
10
<xs:element name="ApplicationProperties">
11
<xs:complexType>
12
<xs:sequence>
13
<xs:element name="Vendor" type="xs:string"/>
14
<xs:element name="ID" type="xs:string"/>
15
</xs:sequence>
16
</xs:complexType>
17
</xs:element>
18
19
<ServiceTemplate id="MRBServiceTemplate">
20
<TopologyTemplate id="MRBService">
21
22
<NodeTemplate id="MyMRB"
23
name="My MRB"
24
nodeType="abc:MRBNode">
25
<Properties>
26
<ApplicationProperties>
27
<Vendor>VendorX</Vendor>
28
<ID>MRB_x2</ID>
29
</ApplicationProperties>
30
</Properties>
31
<DeploymentArtifacts>
32
<DeploymentArtifact name="MRBVMImage"
33
type="http://www.example.com/
34
ns/tosca/2011/12/
35
DeploymentArtifactTypes/VMref">
36
VM-edf2cf99 <!-- This can be a reference to a VM Image (e.g AMI) included in
the CSAR.-->
37
</DeploymentArtifact>
38
</DeploymentArtifacts>
39
40
<NodeTemplate/>
41
42
</TopologyTemplate>
ETSI
156
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
<Plans>
<Plan id="InstantiateMRB"
planType="http://docs.oasisopen.org/tosca/ns/2011/12/PlanTypes/InstantiatePlan"
planLanguage="http://docs.oasisopen.org/wsbpel/2.0/process/executable">
<PlanModelReference reference="prj:InitiateMRB"/>
</Plan>
</ServiceTemplate>
<NodeType name="MRBNode">
<documentation xml:lang="EN">
A reusable definition of a MRB node type representing an
MRB VNF that can be deployed.
</documentation>
<NodeTypeProperties element="ApplicationProperties"/>
<Interfaces>
<Interface name="InstantiationInterface">
<Operation name="InstantiateMRB">
<InputParameters>
<InputParamter name="InstanceName"
type="xs:string"/>
<InputParamter name="NumberOfInstance"
type="xs:integer"/>
</InputParameters>
</Operation>
</Interface>
</Interfaces
</NodeType>
</Definitions>
E.1.2.3 MRF VNFD
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
<Definitions name="MRFServiceTemplateDefinition"
targetNamespace="http://www.etsi.org/sample">
<Tags>
<Tag name="author" value="[email protected]"/>
</Tags>
<Types>
<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema"
elementFormDefault="qualified"
attributeFormDefault="unqualified">
<xs:element name="ApplicationProperties">
<xs:complexType>
<xs:sequence>
<xs:element name="Vendor" type="xs:string"/>
<xs:element name="ID" type="xs:string"/>
</xs:sequence>
</xs:complexType>
</xs:element>
<ServiceTemplate id="MRFServiceTemplate">
<TopologyTemplate id="MRFService">
<NodeTemplate id="MyMRFC+P"
name="My MRFC+P"
nodeType="abc:MRFC+PNode">
<Properties>
<ApplicationProperties>
<Vendor>VendorX</Vendor>
<ID>IMS_MRFC+P_x2</ID>
</ApplicationProperties>
</Properties>
<DeploymentArtifacts>
<DeploymentArtifact name="MRFC+PVMImage"
type="http://www.example.com/
ns/tosca/2011/12/
DeploymentArtifactTypes/VMref">
VM-edf2cf89
</DeploymentArtifact>
<NodeTemplate/>
<NodeTemplate id="MyMRFStorage"
ETSI
157
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
42
name="My MRF Storage"
43
nodeType="abc:MRFStorageNode"/>
44
</TopologyTemplate>
45
46
<Plans>
47
48
<Plan id="InstantiateMRFC+P"
49
planType="http://docs.oasis50
open.org/tosca/ns/2011/12/PlanTypes/InstantiatePlan"
51
planLanguage="http://docs.oasis52
open.org/wsbpel/2.0/process/executable">
53
<PlanModelReference reference="prj:InitiateMRFC+P"/>
54
</Plan>
55
56
<Plan id="InstantiateMRFStorage "
57
planType="http://docs.oasis58
open.org/tosca/ns/2011/12/PlanTypes/TerminationPlan"
59
planLanguage="http://docs.oasis60
open.org/wsbpel/2.0/process/executable">
61
<PlanModelReference reference="prj:InitiateMRFStorage"/>
62
</Plan>
63
</Plans>
64
65
</ServiceTemplate>
66
67
<NodeType name="MRFC+PNode">
68
<documentation xml:lang="EN">
69
A reusable definition of a MRFC+P node type representing an
70
MRB VNF that can be deployed.
71
</documentation>
72
<NodeTypeProperties element="ApplicationProperties"/>
73
<Interfaces>
74
<Interface name="InstantiationInterface">
75
<Operation name="InstantiateMRFC+P">
76
<InputParameters>
77
<InputParamter name="InstanceName"
78
type="xs:string"/>
79
<InputParamter name="NumberOfInstance"
80
type="xs:integer"/>
81
</InputParameters>
82
</Operation>
83
</Interface>
84
</Interfaces
85
</NodeType>
86
87
<NodeType name="MRFStorageNode"
88
targetNamespace="http://www.example.com/sample">
89
<NodeTypeProperties element="ApplicationProperties"/>
90
<Interfaces>
91
<Interface name="InstantiationInterface">
92
<Operation name="InstantiateMRFStorage"/>
93
<InputParameters>
94
<InputParamter name="InstanceName"
95
type="xs:string"/>
96
<InputParamter name="NumberOfInstance"
97
type="xs:integer"/>
98
</InputParameters>
99
</Interface>
100
</Interfaces>
101
</NodeType>
102 </Definitions>
ETSI
158
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Annex F (informative):
YANG/XML VNFD & NSD model
F.1
Use Cases of Network Service with Physical Network
Functions
There are significant benefits for service providers if physical devices attached externally to the Infrastructure network
could be included in a Network Service VNF Forwarding Graph "to facilitate network evolution".
There are a number of use cases for this:
a)
Notably in some mobile Gi services LAN and Internet data centre front end architectures, application delivery
controller stages are used to provide traffic routing and resilience switching between "middlebox" processing
stages. This will also be useful in the NFV case where load balancing is needed for "stateful" appliances
where the go and return flows need to go through the same logical device. While software based Application
Delivery Controller (ADC) functions are available at the server/hypervisor level, providing ADC functionality
across multiple server instances is currently dominated by use of hardware appliances. This requirement is
expected to be removed in the long term as the mobile and fixed core evolves to a series of web services, but
that will take some time.
b)
Service providers may wish to have racks of appliances connected to the infrastructure network either because
the function cannot be easily virtualised, or they have a number of legacy devices that still have service life
and can be "cloudified". The end user may not care if the "firewall service" is delivered from software or
hardware appliances, as "it is all in the cloud".
c)
Putting a modern orchestration wrap round some hardware devices may get round challenges of updating
legacy OSS to introduce new services quickly.
F.2
Examples
vFW
vFWin CP
vNAT
vFW out CP
vNAT in CP
vNAT out CP
Internet
End point
Apn-web-access
End point
Extended
Virtual
Network
VLD
Private
Virtual
Network
VLD
Public
Virtual
Network
VLD
Apn-video-access
End point
vVideo
Cache
out CP
vVideo
Cache
in CP
vVideoCache
Gi-LAN Network Service
Figure F.1: Gi-LAN Network Service example used for modelling NSD and VNFD
ETSI
159
F.2.1
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
NSD for Gi LAN Network Service
<?xml version="1.0" encoding="UTF-8"?>
<nsd xmlns="urn:ietf:params:xml:ns:yang:nfvo"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="urn:ietf:params:xml:ns:yang:nfvo nsd.xsd ">
<name>Gi-Lan-Service</name>
<nsd-id/>
<provider>ETSI</provider>
<description>Gi-LAN services</description>
<version>1.0</version>
<monitoring-params>
<param-id>num-sessions</param-id>
<description>
Total number of sessions the Network Service can handle
</description>
</monitoring-params>
<end-points>
<apn-webaccess>
<end-point-id>apn-webaccess</end-point-id>
<description>APN for web access</description>
</apn-webaccess>
<apn-videoaccess>
<end-point-id>apn-videoaccess</end-point-id>
<description>APN for video access</description>
</apn-videoaccess>
<internet>
<end-point-id>internet</end-point-id>
<description>Internet Connection Point</description>
</internet>
</end-points>
<flavours>
<Gold>
<flavour-id>Gold</flaur-id>
<description>Gold Service</description>
<assurance-params>
<param-id>num-sessions</param-id>
<value>1000000</value>
</assurance-params>
<member-vnfs>
<vNAT>
<member-vnf-id>vNAT</member-vnf-id>
<vnf-flavour>Gold</vnf-flavour>
</vNAT>
<vFW>
<member-vnf-id>vFW</member-vnf-id>
<vnf-flavour>Gold</vnf-flavour>
</vFW>
<vVideoCache>
<member-vnf-id>vVideoCache</member-vnf-id>
<vnf-flavour>Gold</vnf-flavour>
</vVideoCache>
</member-vnfs>
<member-vlds>
<member-vld-id>PrivateVirtualNetwork</member-vld-id>
<member-vld-id>PublicVirtualNetwork</member-vld-id>
<member-vld-id>ExtendedVirtualNetwork</member-vld-id>
</member-vlds>
<forwarding-graphs>
<WebAccess>
<forwarding-graph-id>WebAccess</forwarding-graph-id>
<direction>bi-directional</direction>
<network-forwarding-path>
<end-point-src>apn-webaccess</end-point-src>
<forwarding-policy/>
<vld>ExtendedVirtualNetwork</vld>
<dst-connection-point>
ETSI
160
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
<vnf>vFW</vnf>
<vnf-connection-point>pkt-in</vnf-connection-point>
</dst-connection-point>
<src-connection-point>
<vnf>vFW</vnf>
<vnf-connection-point>pkt-out</vnf-connection-point>
</src-connection-point>
<forwarding-policy/>
<vld> PrivateVirtualNetwork </vld>
<dst-connection-point>
<vnf>vNAT</vnf>
<vnf-connection-point>pkt-in</vnf-connection-point>
</dst-connection-point>
<src-connection-point>
<vnf>vNAT</vnf>
<vnf-connection-point>pkt-out</vnf-connection-point>
</src-connection-point>
<forwarding-policy/>
<vld> PublicVirtualNetwork </vld>
<end-point-dst>internet</end-point-dst>
</network-forwarding-path>
</WebAccess>
<VideoAccess>
<forwarding-graph-id>VideoAccess</forwarding-graph-id>
<direction>bi-directional</direction>
<network-forwarding-path>
<end-point-src>apn-videoaccess</end-point-src>
<forwarding-policy/>
<vld>ExtendedVirtualNetwork</vld>
<dst-connection-point>
<vnf>vVideoCache</vnf>
<vnf-connection-point>pkt-in</vnf-connection-point>
</dst-connection-point>
<src-connection-point>
<vnf>vVideoCache</vnf>
<vnf-connection-point>pkt-out</vnf-connection-point>
</src-connection-point>
<forwarding-policy/>
<vld> PrivateVirtualNetwork </vld>
<dst-connection-point>
<vnf>vNAT</vnf>
<vnf-connection-point>pkt-in</vnf-connection-point>
</dst-connection-point>
<src-connection-point>
<vnf>vNAT</vnf>
<vnf-connection-point>pkt-out</vnf-connection-point>
</src-connection-point>
<forwarding-policy/>
<vld> PublicVirtualNetwork </vld>
<end-point-dst>internet</end-point-dst>
</network-forwarding-path>
</VideoAccess>
</forwarding-graphs>
</Gold>
</flavours>
</nsd>
F.2.2
VNFD for Virtual Firewall
<?xml version="1.0" encoding="UTF-8"?>
<vnfd xmlns="urn:ietf:params:xml:ns:yang:nfvo"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="urn:ietf:params:xml:ns:yang:nfvo vnfd-1.xsd ">
<name>Firewall</name>
<vnfd-id/>
<provider>ETSI</provider>
<description>Firewall</description>
<version>1.0</version>
<workflows>
<init/>
ETSI
161
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
<terminate/>
<graceful-shutdown/>
</workflows>
<connection-points>
<management-port>
<name>mgmt-interface</name>
<description>Management interface</description>
</management-port>
<pkt-in>
<name>pkt-in</name>
<description>Interface for packet in</description>
</pkt-in>
<pkt-out>
<name>packet-out</name>
<description>Packet out interface</description>
</pkt-out>
</connection-points>
<internal-connectivity/>
<monitoring-params>
<num-rules>
<para-id>num-rules</para-id>
<description>Total number of flow the VNF can handle</description>
</num-rules>
</monitoring-params>
<flavours>
<Gold>
<flavour-id>Gold</flavour-id>
<description>Gold Service flavour</description>
<assurance-params>
<param-id>num-rules</param-id>
<value>1000000</value>
</assurance-params>
<vdus>
<vFW>
<vdu-id>vFW</vdu-id>
<num-instances>10</num-instances>
<workflows/>
<vm-spec>
<uri>http://www.example.com/support/vFW1.qcow</uri>
</vm-spec>
<storage/>
<cpu>
<num-vpu>2</num-vpu>
</cpu>
<memory>
<total-memory-gb>8</total-memory-gb>
</memory>
<other-constraints/>
<network-interfaces>
<management-interface>
<name>management-interface</name>
<description>
Interface Used for management interface
</description>
<connection-point-ref>
connection-points/management-interface
</connection-point-ref>
<properties>
<driver>
<id>driver</id>
<value>e1000</value>
</driver>
</properties>
</management-interface>
<pkt-in>
<name>pkt-in</name>
<description>Packet in interface</description>
<connection-point-ref>
connection-points/pkt-in
</connection-point-ref>
<properties>
<driver>
<id>driver</id>
<value>DPDK-PMD</value>
</driver>
</properties>
</pkt-in>
ETSI
162
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
<pkt-out>
<name>pkt-out</name>
<description>Packet out interface</description>
<connection-point-ref>
connection-points/pkt-out
</connection-point-ref>
<properties>
<driver>
<id>driver</id>
<value>DPDK-PMD</value>
</driver>
</properties>
</pkt-out>
</network-interfaces>
</vFW>
</vdus>
</Gold>
<Silver>
<flavour-id>Silver</flavour-id>
<description>Silver Service flavour</description>
<assurance-params>
<param-id>num-rules</param-id>
<value>500000</value>
</assurance-params>
<vdus>
<vFW>
<vdu-id>vFW</vdu-id>
<num-instances>5</num-instances>
<workflows/>
<vm-spec>
<uri>http://www.example.com/support/vFW1.qcow</uri>
</vm-spec>
<storage/>
<cpu>
<num-vpu>2</num-vpu>
</cpu>
<memory>
<total-memory-gb>4</total-memory-gb>
</memory>
<other-constraints/>
<network-interfaces>
<management-interface>
<name>management-interface</name>
<description>
Interface Used for management interface
</description>
<connection-point-ref>
connection-points/management-interface
</connection-point-ref>
<properties>
<driver>
<id>driver</id>
<value>e1000</value>
</driver>
</properties>
</management-interface>
<pkt-in>
<name>pkt-in</name>
<description>Packet in interface</description>
<connection-point-ref>
connection-points/pkt-in
</connection-point-ref>
<properties>
<driver>
<id>driver</id>
<value>DPDK-PMD</value>
</driver>
</properties>
</pkt-in>
<pkt-out>
<name>pkt-out</name>
<description>Packet out interface</description>
<connection-point-ref>
connection-points/pkt-out
</connection-point-ref>
<properties>
<driver>
ETSI
163
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
<id>driver</id>
<value>DPDK-PMD</value>
</driver>
</properties>
</pkt-out>
</network-interfaces>
</vFW>
</vdus>
</Silver>
</flavours>
<dependencies/>
<autoscaling-policies/>
</vnfd>
F.2.3
VNFD for Virtual Carrier Grade NAT
<?xml version="1.0" encoding="UTF-8"?>
<vnfd xmlns="urn:ietf:params:xml:ns:yang:nfvo"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="urn:ietf:params:xml:ns:yang:nfvo vnfd-1.xsd ">
<name>CGNAT</name>
<vnfd-id/>
<provider>ETSI</provider>
<description>CGNAT</description>
<version>1.0</version>
<workflows>
<init/>
<terminate/>
<graceful-shutdown/>
</workflows>
<connection-points>
<management-port>
<name>mgmt-interface</name>
<description>Management interface</description>
</management-port>
<pkt-in>
<name>pkt-in</name>
<description>Interface for packet in</description>
</pkt-in>
<pkt-out>
<name>packet-out</name>
<description>Packet out interface</description>
</pkt-out>
</connection-points>
<internal-connectivity/>
<monitoring-params>
<para-id>num-sessions</para-id>
<description>
Total number of sessions the VNF can handle
</description>
</monitoring-params>
<flavours>
<Gold>
<flavour-id>Gold</flavour-id>
<description>Gold Service flavour</description>
<assurance-params>
<param-id>num-sessions</param-id>
<value>100000</value>
</assurance-params>
<vdus>
<vCGNAT>
<vdu-id>vCGNAT</vdu-id>
<num-instances>10</num-instances>
<workflows/>
<vm-spec>
<uri>http://www.example.com/support/vFW1.qcow</uri>
</vm-spec>
<storage/>
<cpu>
<num-vpu>2</num-vpu>
</cpu>
<memory>
<total-memory-gb>8</total-memory-gb>
</memory>
ETSI
164
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
<other-constraints/>
<network-interfaces>
<management-interface>
<name>management-interface</name>
<description>
Interface Used for management interface
</description>
<connection-point-ref>
connection-points/management-interface
</connection-point-ref>
<properties>
<driver>
<id>driver</id>
<value>e1000</value>
</driver>
</properties>
</management-interface>
<pkt-in>
<name>pkt-in</name>
<description>Packet in interface</description>
<connection-point-ref>
connection-points/pkt-in
</connection-point-ref>
<properties>
<driver>
<id>driver</id>
<value>DPDK-PMD</value>
</driver>
</properties>
</pkt-in>
<pkt-out>
<name>pkt-out</name>
<description>Packet out interface</description>
<connection-point-ref>
connection-points/pkt-out
</connection-point-ref>
<properties>
<driver>
<id>driver</id>
<value>DPDK-PMD</value>
</driver>
</properties>
</pkt-out>
</network-interfaces>
</vCGNAT>
</vdus>
</Gold>
<Silver>
<flavour-id>Silver</flavour-id>
<description>Silver Service flavour</description>
<assurance-params>
<param-id>num-sessions</param-id>
<value>50000</value>
</assurance-params>
<vdus>
<vCGNAT>
<vdu-id>vCGNAT</vdu-id>
<num-instances>5</num-instances>
<workflows/>
<vm-spec>
<uri>http://www.example.com/support/vFW1.qcow</uri>
</vm-spec>
<storage/>
<cpu>
<num-vpu>2</num-vpu>
</cpu>
<memory>
<total-memory-gb>4</total-memory-gb>
</memory>
<other-constraints/>
<network-interfaces>
<management-interface>
<name>management-interface</name>
<description>
Interface Used for management interface
</description>
<connection-point-ref>
ETSI
165
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
connection-points/management-interface
</connection-point-ref>
<properties>
<driver>
<id>driver</id>
<value>e1000</value>
</driver>
</properties>
</management-interface>
<pkt-in>
<name>pkt-in</name>
<description>
Packet in interface
</description>
<connection-point-ref>
connection-points/pkt-in
</connection-point-ref>
<properties>
<driver>
<id>driver</id>
<value>DPDK-PMD</value>
</driver>
</properties>
</pkt-in>
<pkt-out>
<name>pkt-out</name>
<description>
Packet out interface
</description>
<connection-point-ref>
connection-points/pkt-out
</connection-point-ref>
<properties>
<driver>
<id>driver</id>
<value>DPDK-PMD</value>
</driver>
</properties>
</pkt-out>
</network-interfaces>
</vCGNAT>
</vdus>
</Silver>
</flavours>
<dependencies/>
<autoscaling-policies/>
</vnfd>
F.2.4
VNFD for Virtual Video Cache
<?xml version="1.0" encoding="UTF-8"?>
<vnfd xmlns="urn:ietf:params:xml:ns:yang:nfvo"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="urn:ietf:params:xml:ns:yang:nfvo vnfd-1.xsd ">
<name>VideoCaching</name>
<vnfd-id/>
<provider>ETSI</provider>
<description>Video Caching</description>
<version>1.0</version>
<workflows>
<init/>
<terminate/>
<graceful-shutdown/>
</workflows>
<connection-points>
<management-port>
<name>mgmt-interface</name>
<description>Management interface</description>
</management-port>
<pkt-in>
<name>pkt-in</name>
<description>Interface for packet in</description>
</pkt-in>
<pkt-out>
<name>packet-out</name>
<description>Packet out interface</description>
ETSI
166
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
</pkt-out>
</connection-points>
<internal-connectivity/>
<monitoring-params>
<para-id>num-sessions</para-id>
<description>Total number of sessions the VNF can handle</description>
</monitoring-params>
<flavours>
<Gold>
<flavour-id>Gold</flavour-id>
<description>Gold Service flavour</description>
<assurance-params>
<param-id>num-sessions</param-id>
<value>100000</value>
</assurance-params>
<vdus>
<vVideoCache>
<vdu-id>vVideoCache</vdu-id>
<num-instances>10</num-instances>
<workflows/>
<vm-spec>
<uri>http://www.example.com/support/vFW1.qcow</uri>
</vm-spec>
<storage>
<cache>
<id>cache</id>
<size>100g</size>
</cache>
</storage>
<cpu>
<num-vpu>2</num-vpu>
</cpu>
<memory>
<total-memory-gb>16</total-memory-gb>
</memory>
<other-constraints/>
<network-interfaces>
<management-interface>
<name>management-interface</name>
<description>
Interface Used for management interface
</description>
<connection-point-ref>
connection-points/management-interface
</connection-point-ref>
<properties>
<driver>
<id>driver</id>
<value>e1000</value>
</driver>
</properties>
</management-interface>
<pkt-in>
<name>pkt-in</name>
<description>Packet in interface</description>
<connection-point-ref>
connection-points/pkt-in
</connection-point-ref>
<properties>
<driver>
<id>driver</id>
<value>DPDK-PMD</value>
</driver>
</properties>
</pkt-in>
<pkt-out>
<name>pkt-out</name>
<description>Packet out interface</description>
<connection-point-ref>
connection-points/pkt-out
</connection-point-ref>
<properties>
<driver>
<id>driver</id>
<value>DPDK-PMD</value>
</driver>
</properties>
ETSI
167
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
</pkt-out>
</network-interfaces>
</vVideoCache>
</vdus>
</Gold>
<Silver>
<flavour-id>Silver</flavour-id>
<description>Silver Service flavour</description>
<assurance-params>
<param-id>num-sessions</param-id>
<value>50000</value>
</assurance-params>
<vdus>
<vVideoCache>
<vdu-id>vVideoCache</vdu-id>
<num-instances>5</num-instances>
<workflows/>
<vm-spec>
<uri>http://www.example.com/support/vFW1.qcow</uri>
</vm-spec>
<storage>
<cache>
<id>cache</id>
<size>100g</size>
</cache>
</storage>
<cpu>
<num-vpu>2</num-vpu>
</cpu>
<memory>
<total-memory-gb>4</total-memory-gb>
</memory>
<other-constraints/>
<network-interfaces>
<management-interface>
<name>management-interface</name>
<description>
Interface Used for managment interface
</description>
<connection-point-ref>
connection-points/management-interface
</connection-point-ref>
<properties>
<driver>
<id>driver</id>
<value>e1000</value>
</driver>
</properties>
</management-interface>
<pkt-in>
<name>pkt-in</name>
<description>Packet in interface</description>
<connection-point-ref>
connection-points/pkt-in
</connection-point-ref>
<properties>
<driver>
<id>driver</id>
<value>DPDK-PMD</value>
</driver>
</properties>
</pkt-in>
<pkt-out>
<name>pkt-out</name>
<description>Packet out interface</description>
<connection-point-ref>
connection-points/pkt-out
</connection-point-ref>
<properties>
<driver>
<id>driver</id>
<value>DPDK-PMD</value>
</driver>
</properties>
</pkt-out>
</network-interfaces>
</vVideoCache>
ETSI
168
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
</vdus>
</Silver>
</flavours>
<dependencies/>
<autoscaling-policies/>
</vnfd>
F.2.5
VLDs
<?xml version="1.0" encoding="UTF-8"?>
<vld xmlns="urn:ietf:params:xml:ns:yang:nfvo" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="urn:ietf:params:xml:ns:yang:nfvo vld.xsd ">
<name>ExtendedVirtualNetwork</name>
<vld-id/>
<provider>ETSI</provider>
<description>Extended Virtual Network - Interworking with a Gateway Router </description>
<version>1.0</version>
<latency-assurance>10ms</latency-assurance>
<max-end-points>100</max-end-points>
</vld>
<?xml version="1.0" encoding="UTF-8"?>
<vld xmlns="urn:ietf:params:xml:ns:yang:nfvo"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="urn:ietf:params:xml:ns:yang:nfvo vld.xsd ">
<name>PrivateVirtualNetwork</name>
<vld-id/>
<provider>Generic</provider>
<description>Pure Virtual Network</description>
<version>1.0</version>
<latency-assurance>20ms</latency-assurance>
<max-end-points>100</max-end-points>
<properties/>
</vld>
<?xml version="1.0" encoding="UTF-8"?>
<vld xmlns="urn:ietf:params:xml:ns:yang:nfvo" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="urn:ietf:params:xml:ns:yang:nfvo vld.xsd ">
<name>PublicVirtualNetwork</name>
<vld-id/>
<provider>Generic</provider>
<description>Pure Virtual Network</description>
<version>1.0</version>
<latency-assurance>10ms</latency-assurance>
<max-end-points>50</max-end-points>
<properties/>
</vld>
F.3
YANG schema
This clauseis targeted as a preliminary example model towards using YANG as a candidate for modelling the ETSI
NFV information elements.
F.3.1
YANG schema for NSD
module nsd {
namespace "urn:ietf:params:xml:ns:yang:nfvo";
prefix "nsd";
import ietf-yang-types {
prefix yang;
}
import ietf-inet-types {
prefix inet;
}
organization
ETSI
169
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
"example.org";
contact
"Firstname Lastname - [email protected]";
description
"Network Service Descriptor";
revision 2013-12-23 {
description
"Initial revision";
}
container nsd {
leaf name {
description "Network Service Name";
type string;
}
leaf id {
description "Network Service Descriptor ID";
type yang:uuid;
}
leaf vendor {
description "Provider";
type string;
}
leaf description {
description "Description of the Network Service Descriptor";
type string;
}
leaf version {
description "Version";
type string;
}
list monitoring-params {
key param-id;
leaf param-id {
type string;
}
leaf description {
type string;
description "Description";
}
}
list connection-points {
description "End Points for the Network Service";
key connection-point-id;
leaf connection-point-id {
type string;
}
leaf description {
type string;
description "Description of the End point";
}
}
list flavours {
key flavour-id;
leaf flavour-id {
description "Flavour id";
type string;
}
leaf description {
description "Description of the flavour";
type string;
ETSI
170
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
}
list flavour-key {
description "Flavour key";
key param-id;
leaf param-id {
type leafref {
path "../../../monitoring-params/param-id";
}
}
leaf value {
type uint32;
}
}
list constituent-vnfs {
key member-vnf-id;
leaf member-vnf-id {
type string;
}
leaf vnfd {
type yang:uuid;
}
}
container vnf-dependencies {
list vvnf-dependency-list {
key vnf-id;
leaf vnf-id {
type leafref {
path "../../../member-vnfs/member-vnf-id";
}
}
leaf depends-on {
type leafref {
path "../../../member-vnfs/member-vnf-id";
}
}
}
}
list member-vlds {
key member-vld-id;
leaf member-vld-id {
type string;
}
leaf vld-id {
type yang:uuid;
}
}
}
}
}
F.3.2
YANG schema for VNFD
module vnfd {
namespace "urn:ietf:params:xml:ns:yang:nfvo";
prefix "vnfd";
import ietf-yang-types {
prefix yang;
}
import ietf-inet-types {
prefix inet;
}
organization
"example.org";
contact
"Firstname Lastname - [email protected]";
ETSI
171
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
description
"VNF Descriptor";
revision 2013-12-23 {
description
"Initial revision";
}
container vnfd {
leaf name {
description "VNFD Name";
type string;
}
leaf vnfd-id {
description "UUID of the VNF Descriptor";
type yang:uuid;
}
leaf provider {
description "Vendor";
type string;
}
leaf description {
description "Description of the VNF";
type string;
}
leaf version {
description "Version";
type string;
}
container workflows {
container init {
/* Needs expansion */
}
container terminate {
/* Needs expansion */
}
container graceful-shutdown {
/* Needs expansion */
}
}
list connection-points {
description "External interfaces exposed by this VNF";
key connection-point-id;
leaf connection-point-id {
type string;
}
leaf description {
type string;
description "Description of the Connection Point";
}
}
list internal-virtual-link {
key connectivity-id;
leaf connectivity-id {
type string;
}
leaf connectivity-description {
type string;
description "Description";
}
leaf-list connection-end-points {
type leafref {
path "../../flavours/vdus/network-interfaces/name";
}
}
}
list monitoring-params {
key param-id;
leaf param-id {
ETSI
172
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
type string;
}
leaf description {
type string;
description "Description";
}
}
list flavours {
key flavour-id;
leaf flavour-id {
description "Flavour id";
type string;
}
leaf description {
description "Description of the flavour";
type string;
}
list assurance-params {
description "Flavour key";
key param-id;
leaf param-id {
type leafref {
path "../../../monitoring-params/param-id";
}
}
leaf value {
type uint32;
}
}
/* TODO: Add licensing and pricing info */
list vdus {
description "Virtual Deployable Units";
key vdu-id;
leaf vdu-id {
description "";
type string;
}
leaf num-instances {
description "";
type uint16;
}
container workflows {
container init {
/* Needs expension */
}
container terminate {
/* Needs expension */
}
container graceful-shutdown {
/* Needs expension */
}
}
container vm-spec {
choice image {
case image-uri {
leaf pkg-uri {
type inet:uri;
}
}
case image-uuid {
leaf pkg-uuid {
type yang:uuid;
}
}
}
}
ETSI
173
list storage {
key storage-id;
leaf storage-id {
type string;
}
leaf size-gb {
type uint16;
}
list properties {
description "Properties for the storage";
key property-id;
leaf property-id {
type string;
}
leaf property-description {
type string;
description
"Description of the property";
}
leaf property-value {
type string;
}
}
}
container cpu {
leaf num-vpu {
type int16;
}
list properties {
description "Properties for the cpu";
key property-id;
leaf property-id {
type string;
}
leaf property-description {
type string;
description
"Description of the property";
}
leaf property-value {
type string;
}
}
}
container memory {
leaf total-memory-gb {
type int16;
}
list properties {
description "Properties for the memory";
key property-id;
leaf property-id {
type string;
}
leaf property-description {
type string;
description
"Description of the property";
}
leaf property-value {
type string;
}
}
}
list other-constraints {
description "Other Constraints for the VDU";
key constraint-id;
leaf constraint-id {
type string;
}
leaf constraint-description {
type string;
description
ETSI
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
174
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
"Description of the constraint";
}
leaf constraint-value {
type string;
}
}
list network-interfaces {
key name;
leaf name {
type string;
}
leaf description {
type string;
}
choice connection-point {
case connection-point {
leaf connection-point-ref {
type leafref {
path "../../../../connection-points/connection-point-id";
}
}
}
case internal-connectivity {
leaf internal-connectivity-ref {
type leafref {
path "../../../../internal-connectivity/connectivity-id";
}
}
}
}
list properties {
description "Properties for the interface";
key id;
leaf id {
type string;
}
leaf value {
type string;
}
}
}
}
}
container dependencies {
list vdu-dependency-list {
key vdu-id;
leaf vdu-id {
type leafref {
path "../../../flavours/vdus/vdu-id";
}
}
leaf depends-on {
type leafref {
path "../../../flavours/vdus/vdu-id";
}
}
}
}
list autoscaling-policies {
key scaling-policy-id;
leaf scaling-policy-id {
type string;
}
}
}
}
F.3.3
YANG schema for VLD
module vld {
namespace "urn:ietf:params:xml:ns:yang:nfvo";
prefix "vld";
ETSI
175
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
import ietf-yang-types {
prefix yang;
}
import ietf-inet-types {
prefix inet;
}
organization
"example.org";
contact
"Firstname Lastname - [email protected]";
description
"Virtual Link Descriptor";
revision 2013-12-23 {
description
"Initial revision";
}
container vld {
leaf name {
description "Description of the Virtual Link Descriptor";
type string;
}
leaf vld-id {
description "UUID of the VLD";
type yang:uuid;
}
leaf provider {
description "Provider";
type string;
}
leaf description {
description "Description of the VNF";
type string;
}
leaf version {
description "Version";
type string;
}
leaf latency-assurance {
type uint16;
}
leaf max-end-points {
type uint16;
}
list properties {
description "Vendor specific properties";
key property-id;
leaf property-id {
type string;
}
leaf property-description {
type string;
description
"Description of the property";
}
leaf property-value {
type string;
}
}
}
}
ETSI
176
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Annex G (informative):
TM Forum SID service model
NFV-based network functions (VNFs) need to co-exist with traditional non NFV-based NFs (this includes PNFs as well
as logical NFs that are not NFV-based) to enable simple and scalable gradual deployment of VNFs and other NFV
concepts.
NFV-based portions of the network also need to integrate with OSS and BSS functions. Integration of operations and
processes is on a network wide basis. Hence, the integration crosses the boundaries between NFV-based and non-NFVbased functions.
When replacing traditional NFs by VNFs, the impact on any major E2E processes that are affected should be
minimized; examples include:
•
Service Fulfilment.
•
Service Provisioning.
•
Billing.
•
Etc.
E2E processes driven by the OSS/BSS functions need to cross NFV and non-NFV boundaries.
As described in the NFV End-to-End Architecture document, an end-to-end Network Service (e.g. mobile voice/data,
Internet access, or a virtual private network) may be defined as a forwarding graph of network functions (NFs) and end
points/terminals, which is "what" an operator provides to customers.
This is illustrated by figure G.1.
End-to-end network service
Network function forwarding graph
NF
1
End
Point A
Infrastructure
Network 1
NF
2
Infrastructure
Network 2
NF
3
End
Point B
Infrastructure
Network 3
Figure G.1: Graph representation of an end-to-end Network Service
The SID Service Model from the TM Forum is used by most OSS/BSS systems.
To provide easy integration with existing OSS/BSS systems, end-to-end Network Services that include VNFs or VNF
Forwarding Graphs shall be able to be mapped to the SID Service Model.
This would help the Operators to simplify the NFV roll-out, reducing the impact on OSS/BSS. This Service Model shall
be used on the interface between OSS/BSS and Management & Orchestration (Os-Ma) and also on the interface
between OSS/BSS and EM.
Figure G.2 is showing a simplified view of the SID Service Model.
ETSI
177
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Figure G.2: Simplified SID Service Model
NFV defines a number of key NFV concepts. At a minimum, the concepts of a VNF, Network Service, Network
Service Descriptor (NSD) and VNF Graph need to be introduced in the Service Model above to provide the needed
integration.
A VNF is a type of SID LogicalResource. As a SID LogicalResource, VNF implements ResourceFacingServices
(RFSs) that are required to provide CustomerFacingServices (CFSs).
A Network Service has two important characteristics:
1)
it defines a collection of LogicalResources (e.g. VNFs) that have a defined topology; and
2)
the Network Service as a whole may offer a set of interfaces and services that result from arranging the VNFs
in a particular topology and grouping them together as a component.
A Network Service can provide CFSs and/or RFSs (both are optional!). A Network Service shall contain at least 1 VNF.
A Network Service shall contain at least 1 VNF Forwarding Graph. A VNF Forwarding Graph shall have at least one
node, and at least one node shall be a VNF.
Figure G.3 is showing those changes.
CustomerFacingService
ResourceFacingService
CFSserviceRequiresRFServices
1..n
0..n
0..n
0..n
0..n
LogicalResourcesmplementRFS
NetworkServiceProvidesRFS
NetworkServiceProvidesCFS
1..n
0..n
0..n
LogicalResource
NetworkService
1
1..n
1..n
0..n
0..n
VNFFGHasGraphNode
1..n
0..n
VNFFGHasGraphEdge
Legend
0..n
GraphNode
VNF
VNFFGConnectsVNF
VNF Forwarding Graph
SID Model
GraphEdge
VNF Implementation in Service Model
Figure G.3: VNF Implementation in SID Service Model
ETSI
178
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Note that figure G.3 is mainly illustrative and intended as an input to the TM Forum SID.
ETSI
179
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Annex H (informative):
Open Virtualisation Format
The Distributed Management Task Force (DMTF) Open Virtualisation Format (OVF) specification version 2.0.1,
document number DSP0243, dated 2013-08-22 and the related OVF Envelope XML Schema Document (XSD)
version 2.0.1, dated 2013-08-22, DSP8023 were developed to describe the packaging and distribution of software to be
run in a virtual machine. The document status is identified as a DMTF standard. The authors of the OVF specification
identified several characteristics that the specification complies with which include:
•
Distribution considerations including verification and integrity checking.
•
Single and multiple VM configurations.
•
Virtualisation platform neutrality.
•
Vendor and platform neutrality.
•
Design extensibility.
•
Localizable.
•
Openness.
OVF relies on a number of DMTF Common Information Model (CIM) standards. CIM has defined ~1378
classes/subclasses, with another ~61 Problem Resolution Standard (PRS) classes that can be treated as CIM extensions.
This contribution references the CIM V2.38.0 schema.
The OVF specification relies on a number of Common Information Model (CIM) classes in the VirtualHardwareSection
such as the CIM_VirtualSystemSettingData, CIM_ResourceAllocationSettingData,
CIM_EthernetPortAllocationSettingData, and CIM_StorageAllocationSettingData. These
classes have many subclasses.
OVF and the CIM classes may also be used to describe multiple VM based solutions via the
VirtualSystemCollection elements. This may or may not align with some topology considerations for the
VNFD.
Although the specification was not developed with explicit considerations for some of the more exacting requirements
expected by NFV deployments, the broad adoption of support for OVF in cloud related products and its extensible
design means that it may be considered a reasonable starting point for VDU elements and may be considered as one of
the standards that the NFV-interested community can refer to in order to accelerate NFV deployments, drive
interoperability between NFV deployments and in particular, focus on how VDU elements may be described in a
standards compliant way.
The annex in the OVF standard document combined with a number of schemas and the CIM profiles listed on the
DMTF website contain examples of OVF descriptors.
ETSI
180
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Annex I (informative):
Other information
I.1
OpenStack
I.1.1
Introduction
OpenStack provides an Infrastructure as a Service (IaaS) solution through a set of interrelated services. Each service
offers an Application Programming Interface (API) that facilitates their integration.
OpenStack Programs represent core functionalities that are essential to the completion of OpenStack's vision and are
provided by services. Projects represent the implemented services and belong to a Program.
Figure I.1 shows the relationships among the OpenStack core services. For more information about OpenStack services
refer to [i.9]. Note that in the present clause, OpenStack information refer to Havana release (October 2013) [i.10].
Figure I.1: OpenStack conceptual architecture (source [i.9])
I.1.2
NFV-related features in OpenStack community
Following is a list of projects in the OpenStack community which are related to ETSI NFV. Note that the information
contained herein refers to Havana release (October 2013) [i.10] and projects/programs status as of February 2014.
ETSI
181
Project
Snabb NFV
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
Description
Open source NFV system for OpenStack Neutron. The design
goals of the project are:
•
Robustness: no bottleneck, no single point of failure.
•
High capacity (40+ Gbps Ethernet per host).
•
Transparency: connect to operator's own network.
References
For more information:
https://github.com/SnabbCo/snab
bswitch/tree/snabbnfvreadme/src/designs/nfv
Snabb NFV consists of two pieces of software:
•
Snabb mechanism driver: an extension of the
OpenStack Neutron modular layer 2 (ML2) plugin. ML2
plugin is a framework that allows simultaneous
utilization of several L2 networking technologies.
•
Snabb Switch: software switch on compute hosts.
Climate
Configuration and management is performed using a subset of
the standard Neutron API.
Climate is an OpenStack service for virtual and physical
resource reservation. The service aims at handling a calendar
of reservations for various resources based on various
reservation criteria. The project is split into three main
development tracks:
•
Core part development, including leases/reservation
management, plugins, etc.
•
Physical resource reservations.
•
Virtual resource reservations.
For more information:
https://launchpad.net/climate
Following is a list of programs in the OpenStack community which are of particular interest due to specific
requirements from ETSI NFV.
Program
Telemetry
(code-name
Ceilometer)
Baremetal
(code-name
Ironic)
Description
Ceilometer focuses its activity on the following primary
purposes:
• Collection of metering data, e.g. CPU and network
costs.
• Collection of data by monitoring notifications sent by
services, or by polling the infrastructure.
• Configuration of the type of collected data based on
various operating requirements.
• Access of the metering data through the REST APIs.
Ceilometer is an integrated project and it is available in the
OpenStack Havana release.
The program aims at supporting baremetal host provisioning in
OpenStack.
The main outcome is the "baremetal driver", a hypervisor driver
for OpenStack Nova Compute that allows provisioning and
management of physical hardware using common cloud APIs
and tools.
Current implementation of the driver provides functionality for
managing the baremetal database, power control of managed
hardware via IPMI, PXE boot of baremetal nodes, etc.
As of February 2014, Ironic is incubation stage, i.e. not yet
integrated in the OpenStack release.
ETSI
References
For more information:
https://wiki.openstack.org/wiki/Cei
lometer
For more information:
https://wiki.openstack.org/wiki/Ba
remetal
and
https://wiki.openstack.org/wiki/Iro
nic
182
Annex J (informative):
Authors & contributors
The following people have contributed to the present document:
Rapporteur:
Jürgen Quittek, NEC
Other contributors:
Prashant Bauskar, Tech Mahindra
Tayeb BenMeriem, Orange
Andy Bennett, Cisco
Michel Besson, Amdocs
Ron Breault, Wind River Systems
Michael Brenner, Alcatel-Lucent
Peter Michael Bruun, HP
Jean-Marie Calmel, Oracle
Bruno Chatras, Orange
Sachin Chickballapur, Juniper
Junsheng Chu, ZTE
Murray Cooke, Intel
Elena Demaria, Telecom Italia
Vinay Devadatta, Wipro
Thomas Dietz, NEC
Dave Duggle, EnterpriseWeb
Mehmet Ersü, NSN
Marc Flauw, HP
Aijuan Feng, Huawei
Chengyan Feng, Huawei
Stefano Galli, ASSIA
Gerardo Garcia-de-Blas, Telefonica
Deepanshu Gautam, Huawei
Kalyanjeet Gogoi, Juniper Networks
Gao Gongying, China Unicom
Xia Haitao, Huawei
Patrice Hédé, Huawei
Shai Herzog, Amdocs
Adrian Hoban, Intel
Mary Hoover, AT&T
Jenny Huang, AT&T
Fred Huve, HP
Chunxiu Jia, Huawei
Miao Jie, China Unicom
Sudarshan Kandi, Verizon
Ashiq Khan, Docomo
Naseem Khan, Verizon
Norm Kincl, HP
Michael Klotz, Deutsche Telekom
Laurent Laporte, Sprint
Monica Lazer, AT&T
Zhu Lei, Huawei
Gang Liang, Huawei
Jianning Liu, Huawei
Diego Lopez, Telefonica
Paul Mannion, Intel
Veena Mendiratta, Alcatel-Lucent
Raquel Morera, Verizon
Al Morton, AT&T
ETSI
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
183
Tetsuya Nakamura, Docomo
Adrian Neal, Vodafone
Zong Ning, Huawei
Thinh Nguyenphu, NSN
Neal Oliver, Intel
Francois Ozog, 6Wind
Zhan Peng, Huawei
David Pérez-Caparrós, Docomo
Francisco-Javier Ramón Salguero, Telefonica
Sankar Ramamoorthi, Juniper
Uwe Rauschenbach, NSN
Andy Reid, BT
Evelyne Roch, Huawei
René Robert, Orange
Susana Sabater, Vodafone
Raj Sahakari, tech Mahindra
Rajeev Seth, Sonus
Mukhtiar Shaikh, Brocade
Myung-Ki Shin, ETRI
Marcus Schoeller, NEC
Bertrand Souville, Docomo
John Strassner, Huawei
Marco Stura, Ooredoo
Shivani Sud, Intel
Markku Tuohino, NSN
Joan Triay Marques, Docomo
Koji Tsubouchi, Fujitsu
Hrishikesh Venkataraman, Tech Mahindra
Bill Walker, Huawei
Anni Wei, Huawei
Peter Wörndle, Ericsson
Marcus Wong, Huawei
Amanda Xiang, Huawei
Jong-Hwa Yi, ETRI
Valerie Young, Intel
Fang Yu, Huawei
Lei Zhu, Huawei
ETSI
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)
184
History
Document history
V1.1.1
December 2014
Publication
ETSI
ETSI GS NFV-MAN 001 V1.1.1 (2014-12)

advertisement

Was this manual useful for you? Yes No
Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Related manuals

Download PDF

advertisement