Juniper / Cisco Interoperability Cookbook

Juniper / Cisco Interoperability Cookbook
 Juniper / Cisco Interoperability Cookbook August 2014 Juniper Networks/Cisco Systems Switch Interoperability Cookbook
TABLE
OF CONTENTS
Introduction ...................................................................................................................................... 3 Interoperability testing ...................................................................................................................... 5 Cisco Discovery Protocol (CDP) passthrough ............................................................................. 5 Generic Routing Encapsulation (GRE) ........................................................................................ 9 Jumbo frame routing .................................................................................................................. 13 Jumbo frame switching .............................................................................................................. 16 Layer-3 virtual private networks (L3 VPNs) ............................................................................. 19 Link aggregation ........................................................................................................................ 27 Link-Layer Discovery Protocol (LLDP) .................................................................................... 30 Multi-channel link aggregation group (MC-LAG) .................................................................... 34 Multicast routing ........................................................................................................................ 41 Multicast switching .................................................................................................................... 44 Real-Time Performance Monitoring (RPM) .............................................................................. 48 Redundant Trunk Group (RTG) ................................................................................................. 50 Spanning tree case 1: Rapid spanning tree protocol (RSTP) ..................................................... 54 Spanning tree case 2: Multiple spanning tree protocol (MSTP) ................................................ 59 Spanning tree case 3: VLAN spanning tree protocol (VSTP) and Per-VLAN Spanning Tree
Plus (PVST+) ............................................................................................................................. 65 Virtual LAN (VLAN) trunking .................................................................................................. 70 Virtual Router Redundancy Protocol (VRRP) interoperability ................................................. 74 Wi-Fi passthrough ...................................................................................................................... 77 Appendix A: Sample Configuration Files ..................................................................................... 83 Appendix B: Software Versions Tested ........................................................................................ 83 Appendix C: Disclaimer ................................................................................................................. 83 ILLUSTRATIONS
Figure 1: CDP passthrough topology ............................................................................... 6 Figure 2: GRE validation topology ................................................................................. 10 Figure 3: Jumbo frame routing topology ....................................................................... 14 Figure 4: Jumbo frame switching topology ................................................................... 17 Figure 5: L3 VPN validation topology ............................................................................ 20 Figure 6: Link aggregation validation topology ............................................................ 28 Figure 7: LLDP validation topology ............................................................................... 31 Figure 8: MC-­‐LAG validation topology........................................................................... 35 Figure 9: Multicast routing validation topology ........................................................... 42 Figure 10: Multicast switching validation topology ..................................................... 45 Figure 11: Real-­‐Time Performance Monitoring validation topology .......................... 48 Figure 12: Redundant Trunk Group validation topology ............................................ 51 Figure 13: RSTP validation topology ............................................................................. 56 Figure 14: MSTP validation topology ............................................................................ 60 Figure 15: VSTP-­‐PVST+ validation topology ................................................................. 66 Figure 16: VLAN trunking validation topology ............................................................. 71 Figure 17: VRRP validation topology............................................................................. 75 Figure 18: Wi-­‐Fi passthrough validation topology ....................................................... 78 Version 2014081200. Copyright © 2009-2014 Network Test Inc. All rights reserved.
2
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Introduction 3
Objectives This configuration guide aims to help networking professionals successfully interconnect
Juniper Networks and Cisco Systems switches using a variety of popular Layer 2 and
Layer 3 protocols. By following the step-by-step procedures described in this document,
it should be possible to verify interoperability and to pass traffic between the two
vendors’ switches.
Intended audience This configuration guide is intended for any network architect, administrator, or engineer
who needs to interconnect Juniper and Cisco Ethernet switches.
This document assumes familiarity with basic Ethernet and TCP/IP networking concepts,
as well as at least limited experience with the Juniper and Cisco command-line interfaces
(CLIs). No previous experience is assumed for the protocols discussed in this document.
For beginning readers unfamiliar with Juniper or Cisco CLI syntax, both companies’ web
sites offer free access to extensive software documentation. In addition, several excellent
books on Juniper Junos Software and Cisco IOS configuration are available.
For Juniper Junos operating system configuration, these titles include Junos Enterprise
Switching by Harry Reynolds and Doug Marschk; Day One: Exploring the Junos CLI by
Cathy Gadecki and Michael Scruggs, available in free PDF format or in book format; and
the widely used Junos Cookbook by Aviva Garrett.
Popular titles on Cisco IOS configuration include Cisco LAN Switching Fundamentals by
David Barnes and Basir Sakandar; Cisco Routers for the Desperate by Michael W.
Lucas; and Routing TCP/IP, Volume 1 by Jeff Doyle and Jennifer Carroll.
For basic TCP/IP networking concepts, the standard references are Internetworking with
TCP/IP, Volume 1 by Douglas E. Comer and TCP/IP Illustrated, Volume 1 by Kevin R.
Fall and W. Richard Stevens.
For IP multicast topics, Interdomain Multicast Routing: Practical Juniper Networks and
Cisco Systems Solutions by Brian M. Edwards, Leonard A. Giuliano, and Brian R. Wright
offers in-depth explanations of multicast routing protocols and numerous configuration
examples using Juniper and Cisco routers.
Devices covered in this document Using the commands given in this document, Network Test has verified interoperability
between the Juniper EX4300, QFX5100, and Juniper EX9200 Ethernet switches and
Cisco Catalyst 3850 and Cisco Nexus 7000 series Ethernet switches. The Layer-3 VPN
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
interoperability section uses a Juniper MX80 router as well as the other devices
previously mentioned. The Wi-Fi interoperability section also uses a Cisco 5508
controller and Cisco 3602 and Cisco 3702 access points. Appendix B lists software
versions tested. Except where specifically noted, command syntax for the Juniper and
Cisco switches does not change across product lines.
Conventions used in this document The typographical syntax in this document follows that used in the Juniper Complete
Software Guide for Junos Software for EX Switches.
The following table lists text and syntax conventions.
Convention
Description
Examples
Bold type
Represents text that you type
To enter configuration mode, type the
configure command:
Fixed-width text like this
Represents output that appears on the
terminal screen
Italic text like this
•
•
•
•
Introduces important new
terms
Identifies book titles
Identifies RFC and Internetdraft titles
Identifies variables (options
for which you substitute a
value) in commands or
configuration statements.
[email protected]> configure
[email protected]> show chassis
alarms
No alarms currently active
•
•
•
•
< > angle brackets
Enclose optional keywords or variables.
| (pipe symbol)
Indicates a choice between the mutually
exclusive keywords or variables on
either side of the symbol. The set of
choices is often enclosed in parentheses
for clarity.
Indicates a comment specified on the
same line as the configuration statement
to which it appears.
Enclose a variable for which you can
substitute one or more values.
Identify a level in the configuration
hierarchy.
# (pound sign)
[ ] (square braces)
Indention and braces ( { } )
; (semicolon)
Identifies a leaf statement at a
configuration hierarchy level.
A policy term is a named
structure that defines match
conditions and actions.
Junos System Basics
Configuration Guide
RFC 4814, Hash and
Stuffing: Overlooked Factors
in Network Device
Benchmarking
[email protected]# set system
domain-name domainname
stub <default-metric
metric>;
broadcast | multicast
(string1 | string2 |
string3)
rsvp { # Required for
dynamic MPLS only
community name members [
community-ids]
[edit]
routing-options {
static {
route default {
nexthop address;
retain;
}
}
}
nexthop address;
4
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Interoperability testing For each interoperability test described here, this document uses a five-section format
consisting of objective, technical background, Juniper configuration, Cisco configuration
and test validation.
Cisco Discovery Protocol (CDP) passthrough Objective To verify the ability of a Juniper switch to forward Cisco Discovery Protocol (CDP)
traffic between two Cisco devices.
Background The proprietary Cisco Discovery Protocol (CDP) allows sharing of information, such as
IP address, model number, and power requirements among connected Cisco devices.
Cisco devices use CDP messages to transmit information about their capabilities to
other Cisco products in the network. Accordingly, an interoperability requirement for
any Juniper switch in the path between two Cisco devices is the ability to “pass
through” CDP traffic without affecting CDP operation.
No extra configuration of Juniper or Cisco switches is required for CDP passthrough.
Because Juniper EX Series and QFX Series switches forward CDP messages in regular
Ethernet frames, a standard Ethernet switching configuration will work. Similarly, CDP
is enabled by default on most Cisco devices, so no additional configuration is needed.
Topology In this example, Cisco Catalyst 3850 and Cisco Nexus 7010 switches will use CDP to
exchange model numbers and interface information across two Juniper EX9208 switches
in a Virtual Chassis configuration. Though not required for this test, the inter-switch links
also used link aggregation to bundle one or more physical interfaces into a single logical
pipe. There is a separate section in this document describing link aggregation
configuration.
The interfaces used are as follows:
•
•
•
Cisco Catalyst 3850: TenGigabitEthernet1/1/3, TenGigabitEthernet1/1/4, and
Port-channel2 (t1/1/3, t1/1/4 and po2)
Juniper Virtual Chassis with EX9208: xe-5/3/1, xe-12/3/0, and ae1 (to Catalyst
3850); and xe-5/0/5, xe-12/0/5, and ae2 (to Nexus 7010)
Cisco Nexus 7010: Ethernet3/9, Ethernet3/10, and port-channel1 (e3/9, e3/10, and
po1)
5
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
All devices are configured as switches and all inter-switch links act as VLAN trunks.
However, even without VLAN configuration CDP traffic will be forwarded just as in this
example. Figure 1 shows the topology for CDP passthrough.
Figure 1: CDP passthrough topology Juniper configuration In this example, the inter-switch ports use link aggregation and VLAN trunking. Both
steps are optional; if desired, physical switch ports can instead be configured for Ethernet
switching.
These steps define two VLANs; again, this is optional:
[email protected]> configure
[email protected]# set vlans v2001 vlan-id 2001
[email protected]# set vlans v2002 vlan-id 2002
These steps create link aggregation groups ae1 to the Cisco Catalyst 3850 and ae2 to the
Cisco Nexus 7010:
[email protected]#
[email protected]#
[email protected]#
[email protected]#
set
set
set
set
interfaces
interfaces
interfaces
interfaces
ae1
ae1
ae2
ae2
aggregated-ether-options lacp active
description "linkagg to 3850"
aggregated-ether-options lacp active
description “linkagg to 7010"
These steps configure the newly created
trunk ports:
ae
interfaces as Ethernet switching and VLAN
[email protected]# set interfaces ae1 unit 0 family ethernet-switching
interface-mode trunk
[email protected]# set interfaces ae1 unit 0 family ethernet-switching vlan
members v2001
[email protected]# set interfaces ae1 unit 0 family ethernet-switching vlan
members v2002
6
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
[email protected]# set interfaces ae2 unit 0 family ethernet-switching
interface-mode trunk
[email protected]# set interfaces ae2 unit 0 family ethernet-switching vlan
members v2001
[email protected]# set interfaces ae2 unit 0 family ethernet-switching vlan
members v2002
These steps assign physical ports to membership in the link aggregation groups:
[email protected]#
[email protected]#
[email protected]#
[email protected]#
[email protected]#
[email protected]#
[email protected]#
[email protected]#
set
set
set
set
set
set
set
set
interfaces
interfaces
interfaces
interfaces
interfaces
interfaces
interfaces
interfaces
xe-5/3/1 ether-options 802.3ad ae1
xe-5/3/1 description "ae1 linkagg to 3850"
xe-12/3/0 ether-options 802.3ad ae1
xe-12/3/0 description "ae1 linkagg to 3850"
xe-5/0/5 ether-options 802.3ad ae2
xe-5/0/5 description "ae2 linkagg to 7010"
xe-12/0/5 ether-options 802.3ad ae2
xe-12/0/5 description "ae2 linkagg to 7010"
The spanning tree protocol must be either disabled on all switches, or disabled on all
switches. This command will enable rapid spanning tree on a Juniper EX Series switch:
[email protected]# set protocols rstp
[email protected]# commit
To disable rapid spanning tree on a Juniper EX Series switch:
[email protected]# set protocols rstp disable
[email protected]# commit
Cisco commands Since CDP is enabled by default on Cisco devices, no additional configuration is needed.
The steps given here are to define link aggregation and VLAN trunking, but both are
optional for purposes of validating CDP passthrough.
On the Catalyst 3850, CDP and rapid spanning tree (called Rapid PVST-Plus in Cisco
documentation) will be enabled. All that remains is to (optionally) define a link
aggregation group and VLAN trunking.
These commands will create two VLANs:
Cat3850# configure terminal
Cat3850(config)# vlan 2001-2002
Cat3850(config-vlan)# exit
These steps will create a link aggregation group (called a Port-channel in Cisco parlance)
and configure it for VLAN trunking:
Cat3850(config)# interface Port-channel1
Cat3850(config-if)# description to EX9200
Cat3850(config-if)# switchport trunk allowed vlan 2001-2002
Cat3850(config-if)# switchport mode trunk
Cat3850(config-if)# exit
7
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
These steps assign physical ports to membership in the link aggregation groups:
Cat3850(config)# interface TenGigabitEthernet1/1/3
Cat3850(config-if)# description po1 to 9200
Cat3850(config-if)# switchport trunk allowed vlan 2001-2002
Cat3850(config-if)# switchport mode trunk
Cat3850(config-if)# channel-group 1 mode passive
Cat3850(config)# interface TenGigabitEthernet1/1/4
Cat3850(config-if)# description po1 to 9200
Cat3850(config-if)# switchport trunk allowed vlan 2001-2002
Cat3850(config-if)# switchport mode trunk
Cat3850(config-if)# channel-group 1 mode passive
Cat3850(config-if)# end
Then issue similar commands on the Nexus 7010. First create multiple VLANs:
Nexus7010# configure terminal
Nexus7010(config)# vlan 2001-2002
Nexus7010(config-vlan)# exit
Then create a link aggregation group and configure it for VLAN trunking:
Nexus7010(config)# interface port-channel2
Nexus7010(config-if)# description linkagg to ex9200 ae2
Nexus7010(config-if)# switchport
Nexus7010(config-if)# switchport mode trunk
Nexus7010(config-if)# switchport trunk allowed vlan 2001-2002
Finally, assign physical ports to the link aggregation group and configure it for VLAN
trunking. Note that Cisco Nexus ports are in shutdown mode by default, and must be
explicitly enabled:
Nexus7010(config-if)#
Nexus7010(config-if)#
xe-12/0/5
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
interface Ethernet3/9
description linkagg to ex9200 ae2 xe-5/0/5 and
switchport
switchport mode trunk
switchport trunk allowed vlan 2001-2003
channel-group 2 mode active
no shutdown
end
Validation To verify that a Juniper EX Series switch will forward CDP messages between two Cisco
devices, use the show cdp neighbors command on either Cisco device.
8
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
The Catalyst 3850 will recognize the Nexus 7010 via CDP:
9
Cat3850#show cdp neighbors
Capability Codes: R - Router, T - Trans Bridge, B - Source Route Bridge
S - Switch, H - Host, I - IGMP, r - Repeater, P - Phone,
D - Remote, C - CVTA, M - Two-port Mac Relay
Device ID
Local Intrfce
Nexus7010(TBM13093202)
Ten 1/1/4
3/10
Nexus7010(TBM13093202)
Ten 1/1/3
3/10
Holdtme
Capability
Platform
Port ID
138
R S C
N7K-C7010 Eth
138
R S C
N7K-C7010 Eth
And the Nexus 7010 similarly will recognize the Catalyst 3850:
Nexus7010# show cdp
Capability Codes: R
S
V
s
neighbors
- Router, T - Trans-Bridge, B - Source-Route-Bridge
- Switch, H - Host, I - IGMP, r - Repeater,
- VoIP-Phone, D - Remotely-Managed-Device,
- Supports-STP-Dispute
Device-ID
Local Intrfce
Cat3850.englab.juniper.net
Eth3/9
Cat3850.englab.juniper.net
Eth3/9
Hldtme Capability
Platform
Port ID
156
S I
WS-C3850-48P
Ten1/1/3
170
S I
WS-C3850-48P
Ten1/1/4
Note that in both cases, the Cisco switches correctly identified the hostname (“Device
ID”), model number (“Platform”) and interface (“Port ID”) of the remote Cisco device.
All this information is learned via CDP, which is forwarded without any additional
configuration needed on Juniper switches running Junos.
Generic Routing Encapsulation (GRE) Objective To verify the ability of Juniper and Cisco switches to tunnel traffic over an IP backbone
using GRE.
Background As described in IETF RFC 2784, GRE provides a method of encapsulating traffic into IP
packets for transmission across a routed network. On the receiving end, the traffic is
decapsulated and forwarded in its original form.
Although this configuration example uses IP-in-IP encapsulation, GRE can carry
virtually any protocol, including non-routable traffic such as raw Ethernet frames, across
a routed IP network.
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Topology Figure 2 shows the GRE validation test bed. This example uses a GRE tunnel between a
Juniper Virtual Chassis (here, comprising two Juniper EX9208 switches) and a Cisco
Nexus 7010 switch. The GRE tunnel endpoints are in the 192.18.44.0/24 subnet, and the
tunnel uses a 10-gigabit Ethernet link between switches.
Even though a different subnet (192.18.38.0/24) is configured on the physical interfaces
of both switches, traffic will traverse the GRE tunnel. In this example, the Juniper and
Cisco devices use OSPF to learn about network reachability.
Figure 2: GRE validation topology Juniper commands 1. Define IP addresses on the inter-switch link and the link to the Spirent TestCenter
traffic generator/analyzer. Note that these addresses are associated with the physical
interfaces, not the GRE tunnel:
[email protected]> configure
[email protected]# set interfaces xe-5/0/5 description "physical interface to
Nexus 7010 e3/9"
[email protected]# set interfaces xe-5/0/5 unit 0 family inet address
192.18.38.1/24
[email protected]# set interfaces ge-2/0/0 description "physical interface to
STC"
10
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
[email protected]# set interfaces ge-2/0/0 unit 0 family inet address
192.18.5.1/24
2. Define a GRE tunnel, in this case called gr-5/0/0. In Junos, the required commands
include a tunnel source, destination, and endpoint IP address:
[email protected]# set
Nexus 7010”
[email protected]# set
[email protected]# set
[email protected]# set
192.18.44.1/24
interfaces gr-5/0/0 description "GRE tunnel endpoint to
interfaces gr-5/0/0 unit 0 tunnel source 192.18.38.1
interfaces gr-5/0/0 unit 0 tunnel destination 192.18.38.2
interfaces gr-5/0/0 unit 0 family inet address
3. Enable OSPF and enable it on the tunnel interface:
[email protected]# set protocols ospf area 0.0.0.0 interface gr-5/0/0.0
[email protected]# commit
Cisco commands 1. Ensure the tunnel feature is enabled. This step is required for Cisco Nexus 7000
switches, but is not required for Cisco Catalyst switches:
Nexus7010# configure terminal
Nexus7010(config)# feature tunnel
2. Define IP addresses on the inter-switch link and the link to the Spirent TestCenter
traffic generator/analyzer. Note that these addresses are associated with the physical
interface, not the GRE tunnel:
Nexus7010(config)# interface Ethernet3/9
Nexus7010(config-if)# description physical interface to EX9208 xe-5/0/5
Nexus7010(config-if)# ip address 192.18.38.2/24
Nexus7010(config-if)# no shutdown
Nexus7010(config-if)# interface Ethernet3/13
Nexus7010(config-if)# description physical interface to STC
Nexus7010(config-if)# ip address 192.18.98.1/24
Nexus7010(config-if)# no shutdown
Note that Cisco Nexus ports are in shutdown mode by default, and must be explicitly
enabled. This step is not required for Cisco Catalyst switches.
3. Define a GRE tunnel, in this case called Tunnel0. In NX-OS, the required commands
include a tunnel source, destination, and endpoint IP address:
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
interface Tunnel0
ip address 192.18.44.2/24
tunnel source 192.18.38.2
tunnel destination 192.18.38.1
no shutdown
exit
3. Enable OSPF and enable it on the tunnel interface’s network:
11
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
12
Nexus7010(config)# feature ospf
Nexus7010(config)# router ospf 1
Nexus7010(config-rtr)# network 192.18.44.0/24 area 0.0.0.0
Nexus7010(config-rtr)# log-adjacency-changes
Nexus7010(config-rtr)# end
Validation The command “show interfaces gr-5/0/0” will verify that the GRE tunnel is
up. If run while the Juniper and Cisco devices are forwarding traffic, this command’s
output will include traffic statistics:
[email protected]> show interfaces gr-5/0/0
Physical interface: gr-5/0/0, Enabled, Physical link is Up
Interface index: 833, SNMP ifIndex: 1665
Type: GRE, Link-level type: GRE, MTU: Unlimited, Speed: 100000mbps
Device flags
: Present Running
Interface flags: Point-To-Point SNMP-Traps
Input rate
: 0 bps (0 pps)
Output rate
: 0 bps (0 pps)
Logical interface gr-5/0/0.0 (Index 818) (SNMP ifIndex 1673)
Flags: Up Point-To-Point SNMP-Traps 0x4000
IP-Header 192.18.38.2:192.18.38.1:47:df:64:0000000000000000
Encapsulation: GRE-NULL
Copy-tos-to-outer-ip-header: Off
Gre keepalives configured: Off, Gre keepalives adjacency state: down
Input packets : 0
Output packets: 0
Protocol inet, MTU: 9154
Flags: Sendbcast-pkt-to-re
Addresses, Flags: Is-Preferred Is-Primary
Destination: 192.18.44/24, Local: 192.18.44.1, Broadcast:
192.18.44.255
gr-5/0/0, Enabled, Link is Up
Encapsulation: GRE, Speed: 100000mbps
Traffic statistics:
delta
Input bytes:
3281648740 (79102480 bps)
[98256668]
Output bytes:
2233966800 (39551240 bps)
[49103600]
Input packets:
6643094 (20015 pps)
[198901]
Output packets:
4522200 (10007 pps)
[99400]
Current
On Cisco devices, the equivalent command is “show interface tunnel
<tunnel number>”.
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Jumbo frame routing 13
Objective To validate the ability of Juniper and Cisco switches to correctly route bidirectional
traffic with packet lengths greater than 1,500 bytes.
Background Some routing protocols such as open shortest path first (OSPF) require that both routers
agree on the same maximum transmission unit (MTU) when exchanging routing
information. For Ethernet interfaces, the requirement for matched MTUs applies equally
to jumbo frames (those larger than 1,518 bytes) as to standard-length frames.
In part because of the lack of a standard length for jumbo frames, there is confusion in the
marketplace about the maximum frame length possible. Older Linux drivers for Ethernet
interfaces in servers support a maximum length of around 7,000 bytes, though most
Linux drivers now allow frame lengths of 9,000 bytes or more. Ethernet interfaces of
switches and routers typically support a larger protocol data unit (PDU) but there is some
confusion as to whether that PDU should be a maximum of 9,000 bytes or 9,216 bytes.
Adding to the confusion, implementations differ as to whether the 4-byte cyclic
redundancy check (CRC) should or should not be included when stating the maximum
frame length.
Juniper and Cisco switches typically support 9,216-byte jumbo frames, including CRC1.
This section explains how to configure both vendors’ devices to set up an OSPF routing
session using jumbo frames.
Topology In this example, a Juniper Virtual Chassis Fabric (comprising two Juniper QFX5100s and
one Juniper EX4300) configured as an OSPF router exchanges jumbo frames with a
Cisco Nexus 7010 switch. This example uses OSPF because it requires both sides to use
the same MTU when exchanging database description messages2.
Figure 3 illustrates the configuration used to validate jumbo frame routing. In this
example, an OSPF routing session will be established between the Juniper and Cisco
devices. Both interfaces have IP addresses in the 192.18.64.0/24 subnet.
1
The Juniper EX9200 supports a maximum Ethernet frame length of 9,192 bytes, or 9,152 bytes when
configured in Virtual Chassis mode, to account for a 40-byte internal header. Other Juniper devices,
including those given in the examples here, support a maximum Ethernet frame length of 9,216 bytes.
2
This requirement is specified in RFC 2328, section 10.6.
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
14
Figure 3: Jumbo frame routing topology Juniper commands Jumbo frame support is enabled by adding the mtu keyword when configuring interfaces.
Note that the mtu keyword applies to the physical interface and not the logical unit
interface where an IPv4 address is assigned.
Note that the Junos mtu keyword does not include the Ethernet CRC. Thus, to pass
9,216-byte Ethernet frames (including CRC), the routing interface will take a command
of mtu 9212.
These commands assign MTU and IP address to interface xe-0/1/0:
[email protected]>
[email protected]#
[email protected]#
[email protected]#
configure
set interfaces xe-1/0/2 mtu 9212
set interfaces xe-1/0/2 description "Nexus 7010 e3/3"
set interfaces xe-1/0/2.0 family inet address 192.18.64.2/24
Next, this command starts OSPF routing on interface xe-1/0/2.0. In this example, the
interface is a member of OSPF area 0:
[email protected]# set protocols ospf area 0.0.0.0 interface xe-1/0/2.0
[email protected]# commit
Cisco commands Cisco devices also use the mtu keyword in the interface configuration context to enable
switching of jumbo frames. Cisco IOS has separate commands for mtu, describing the
maximum transmission unit for the Ethernet frame and for the ip mtu, describing the
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
MTU for the IP packet. Cisco NX-OS, as in the Nexus 7010, uses only the mtu keyword
to cover Ethernet frame length:
Nexus7010# configure terminal
Nexus7010(config)# system jumbomtu 9216
Nexus7010(config)# interface Ethernet3/3
Nexus7010(config-if)# description to Juniper VCF xe-1/0/2
Nexus7010(config-if)# mtu 9216
Nexus7010(config-if)# ip address 192.18.64.1/24
Nexus7010(config-if)# no shutdown
Nexus7010(config-if)# exit
Nexus7010(config)# router ospf 1
Nexus7010(config-rtr)# log-adjacency-changes
Nexus7010(config-rtr)# network 192.18.64.0 0.0.0.255 area 0
Nexus7010(config-rtr)# end
The above example is for Nexus 7000 series switches. On Catalyst 3850 switches, the
global configuration system mtu routing command sets IP MTU size:
Cat3850# configure terminal
Cat3850(config)# system mtu routing 9198
Cat3850(config)# interface TenGigabitEthernet1/0/1
Cat3850(config-if)# no switchport
Cat3850(config-if)# ip address 10.0.0.1 255.255.255.0
Cat3850(config-if)# exit
Cat3850(config)# router ospf 1
Cat3850(config-rtr)# log-adjacency-changes
Cat3850(config-rtr)# network 10.0.0.0 0.0.0.255 area 0
Cat3850(config-rtr)# end
The commands above have been verified with Catalyst 3850 and 3750-E switches routing
jumbo frames. On some versions of IOS, the Catalyst 3750 may instead use the global
system mtu jumbo <value> command.
Validation Unless both Juniper and Cisco interfaces agree on MTU size, OSPF routing adjacencies
will remain in ExStart state, and will never transition to OSPF “full” state. To verify that
an OSPF adjacency has entered OSPF “full” state on Juniper switches, use the show
ospf neighbor command:
[email protected]> show ospf neigbhor
Address
Interface
192.18.64.1
xe-1/0/2.0
State
Full
ID
192.18.64.1
Pri
1
Dead
32
On the Cisco device, use the show ip ospf neighbor command:
Nexus7010# show ip ospf neighbor
Neighbor ID
192.18.64.2
Ethernet3/3
Pri
128
State
FULL/BDR
Dead Time
00:00:35
Address
192.18.64.2
Interface
15
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
The fact that both routers are in OSPF “Full” state indicates they have agreed to exchange
IP packets up to 9,198 bytes long (9,216 bytes, including Ethernet header and CRC).
OSPF routing sessions will not be fully established unless both sides agree on MTU size.
Jumbo frame switching Objective To validate the ability of Juniper and Cisco switches to correctly switch bidirectional
traffic consisting of jumbo frames.
Background For many years the IEEE Ethernet specification has defined the maximum length of an
Ethernet frame to be 1,518 bytes (or 1,522 bytes with an 802.1Q VLAN tag). The use of
jumbo frames – those larger than 1,518 bytes – remains nonstandard3.
In part because of the lack of a standard length for jumbo frames, there is confusion in the
marketplace about the maximum frame length possible. Older Linux drivers for Ethernet
interfaces in servers support a maximum length of around 7,000 bytes, though most
Linux drivers now allow frame lengths of 9,000 bytes or more. Ethernet interfaces of
switches and routers typically support a larger protocol data unit (PDU) but there is some
confusion as to whether that PDU should be a maximum of 9,000 bytes or 9,216 bytes.
Adding to the confusion, implementations differ as to whether the 4-byte cyclic
redundancy check (CRC) should or should not be included when stating the maximum
frame length.
Juniper and Cisco switches typically support 9,216-byte jumbo frames, including CRC4.
This section explains how to configure both vendors’ switches to exchange jumbo
frames.
Topology In this example, a Juniper Virtual Chassis Fabric switch (comprising two Juniper
QFX5100s and one Juniper EX4300) exchanges jumbo frames with a Cisco Nexus 7010.
As commonly used in many organizations, VLAN trunk ports connect the switches and
VLAN access ports at the edge accept untagged jumbo frames. However, the ability to
switch jumbo frames does not depend on VLAN tagging. This example would also work
with all interfaces passing untagged traffic.
3
Recent versions of the 802.3 Ethernet specification have extended the maximum “envelope” frame length
to 2,000 bytes to allow for multiple VLAN headers and various encapsulation methods. However, the
specification’s maximum “basic” frame length remains at 1,518 bytes.
4
The Juniper EX9200 supports a maximum Ethernet frame length of 9,192 bytes, or 9,152 bytes when
configured in Virtual Chassis mode, to account for a 40-byte internal header. Other Juniper devices,
including those given in the examples here, support a maximum Ethernet frame length of 9,216 bytes.
16
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Figure 4 illustrates the configuration used to validate jumbo frame switching. As noted in
the configuration sections below, all interfaces explicitly support switching of jumbo
frames. On the Juniper Virtual Chassis Fabric, two clients (one apiece attached to the
Juniper QFX5100 and EX4300) use an untagged VLAN ID of 2001. The 10-Gbit/s
Ethernet interface xe-1/0/2 is a trunk port, conveying tagged traffic to the Cisco Nexus
7010 switch. On the Cisco side, interface Ethernet3/3 is also a trunk port. A server is
attached to access port Ethernet3/13.
Figure 4: Jumbo frame switching topology Juniper commands Jumbo frame support is enabled by adding the mtu keyword when configuring interfaces.
Note that the mtu keyword applies to the physical interface and not the logical unit
interface where VLAN membership is assigned.
Note that the Junos mtu keyword does not include the Ethernet CRC. Thus, to pass
9,216-byte Ethernet frames (including CRC), untagged (access) ports will take a
command of mtu 9212, while trunk ports will take a command of mtu 9216 (to
accommodate the 4-byte VLAN tag).
In this example, MTU and VLAN settings are configured separately. First, MTU settings
are applied to each interface. Again, note that interface xe-1/0/1 takes a larger MTU value
to accommodate VLAN tagging:
[email protected]>
[email protected]#
[email protected]#
[email protected]#
[email protected]#
[email protected]#
[email protected]#
e3/3"
configure
set interfaces
set interfaces
set interfaces
set interfaces
set interfaces
set interfaces
xe-1/0/40 mtu 9212
xe-1/0/40 description "to client on QFX5100"
ge-2/0/0 mtu 9212
ge-2/0/0 description "to client on EX4300"
xe-1/0/2 mtu 9216
xe-1/0/2 description "VLAN trunk to Nexus 7010
17
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Next, a VLAN is created and interfaces are assigned to the VLAN. In this example, the
client-attached interfaces accept untagged traffic, while interface xe-1/0/2 passes tagged
traffic to the Cisco switch:
[email protected]# set vlans v2001 vlan-id 2001
[email protected]# set interfaces xe-1/0/40.0 family ethernet-switching vlan
members v2001
[email protected]# set interfaces ge-2/0/0.0 family ethernet-switching vlan members
v2001
[email protected]# set interfaces xe-1/0/2.0 family ethernet-switching port-mode
trunk vlan members v2001
The spanning tree protocol must be either enabled or disabled on all switches. This
command will enable rapid spanning tree on a Juniper switch:
[email protected]# set protocols rstp
[email protected]# commit
To disable rapid spanning tree on a Juniper switch:
[email protected]# set protocols rstp disable
[email protected]# commit
Cisco commands Cisco devices also use the mtu keyword in the interface configuration context to enable
switching of jumbo frames. As with the Juniper configuration, VLANs are created
separately. Unlike the Juniper example, MTU size and VLAN membership are both
associated with the physical interface, and MTU size also is set systemwide. Also, with
Cisco IOS and NX-OS devices, the mtu keyword does include the Ethernet CRC:
Nexus7010# configure terminal
Nexus7010(config)# system jumbomtu 9216
Nexus7010(config)# vlan 2001
Nexus7010(config-vlan)# exit
Nexus7010(config)# interface Ethernet3/3
Nexus7010(config-if)# description VLAN trunk to Juniper VCF
Nexus7010(config-if)# mtu 9216
Nexus7010(config-if)# switchport
Nexus7010(config-if)# switchport mode trunk
Nexus7010(config-if)# switchport trunk allowed vlan 2001
Nexus7010(config-if)# no shutdown
Nexus7010(config-if)# interface Ethernet3/13
Nexus7010(config-if)# description to server
Nexus7010(config-if)# mtu 9216
Nexus7010(config-if)# switchport
Nexus7010(config-if)# switchport mode access
Nexus7010(config-if)# switchport access vlan 2001
Nexus7010(config-if)# no shutdown
Nexus7010(config-if)# end
18
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
The “no shutdown” command is mandatory for the Nexus 7010 and other devices
running the NX-OS operating system. It is optional for the Cisco Catalyst 3850 and other
switches running the IOS operating system.
Validation Generating a known quantity of jumbo frames between the client and server will validate
the ability of the switches to exchange jumbo traffic. This can be verified by examining
the interface counters on the switch ports where clients and servers are attached.
Alternatively, a test instrument can generate bidirectional jumbo frame traffic between
the switches. Both switches should forward all jumbo frames with zero frame loss.
Layer-­‐3 virtual private networks (L3 VPNs) Objective To verify the ability of Juniper and Cisco switches to use BGP to create a VPN tunnel
over an MPLS-based network.
To verify the ability of Juniper and Cisco switches to forward traffic over an L3 VPN.
Background MPLS-based VPNs provide virtual, routable IP tunnels across an MPLS network. As
described in IETF RFC 4364, customers use BGP routing to set up these tunnels, with no
visibility of the service provider’s underlying MPLS transport.
MPLS-based VPNs offer advantages over conventional IP-based VPNs for enterprises
and service providers in terms of scalability and ease of configuration. Conventional
VPNs typically require fully meshed networking among all sites requiring VPN
connectivity. Moreover, each router must be reconfigured each time a site is added or
deleted. In contrast, a change in one site in an MPLS-based VPN requires reconfiguration
only of the service provider’s edge router at that site.
Topology This example models a service provider’s MPLS network with provider edge (PE) and
provider (P) devices, as well as customer edge (CE) devices representing an enterprise
network. In this case, a Juniper MX80 router acts as the P device. A Juniper Virtual
Chassis (comprising two Juniper EX9208 switches) and a Cisco Nexus 7010 act as PE
devices. A Juniper Virtual Chassis Fabric (comprising two Juniper QFX5100 and one
Juniper EX4300 switches) and a Cisco Catalyst 3850 act as CE devices.
The PE devices run MP-BGP (multiprotocol BGP) and OSPF routing protocols,
redistributing routes learned from the L3 VPN into OSPF for use by the CE devices. The
PE and P devices also run MPLS Label Distribution Protocol (LDP). The PE devices also
19
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
run one unique virtual routing and forwarding (VRF) instance for each customer; this
allows customers to maintain unique, and possibly overlapping, routing tables (for
example, with multiple customers each using the same RFC 1918 private address space).
The CE devices run OSPF.
20
This table lists the IP networks in use in the customer and service provider networks.
With L3 VPNs, the customer’s equipment has no visibility into the service provider
network. Instead, all traffic appears to be routed between customer networks using BGP
and OSPF.
Customer networks 192.18.1.0/24, 192.18.18.0/24, 192.18.40.0/24, 192.18.68.0/24, 192.18.69.0/24, 192.18.98.0/24, 192.18.99.0/24 Service provider networks 192.18.38.0/24, 192.38.70.0/24 Figure 5 shows the L3 VPN validation test bed.
Figure 5: L3 VPN validation topology Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Juniper configuration Juniper Virtual Chassis Fabric (CE device):
With L3 VPNs, no special configuration is needed for CE routers. Indeed, unlike
conventional VPNs, the CE routers do not require fully meshed connectivity with all
other sites using VPN tunneling.
In this example, the CE devices run OSPF to exchange information with the PE routers.
The CE devices also could use static routes or other dynamic routing protocols such as
IS-IS or BGP. Again, no MPLS or VPN awareness is required.
1. Configure IP addresses for interfaces xe-1/0/2 (connected with the Cisco Nexus 7000)
and xe-1/0/40 and ge-2/0/2 (connected with the Spirent TestCenter test instrument via the
Juniper QFX5100 and Juniper EX4300 switches, respectively, within the Virtual Chassis
Fabric):
VCF>
VCF#
VCF#
VCF#
VCF#
VCF#
VCF#
configure
set interfaces
set interfaces
set interfaces
set interfaces
set interfaces
set interfaces
xe-1/0/2 description "to Nexus 7010 int e3/3"
xe-1/0/2 unit 0 family inet address 192.18.68.2/24
xe-1/0/40 description "VCF5100 to stc"
xe-1/0/40 unit 0 family inet address 192.18.99.1/24
ge-2/0/2 description "VCF4300 to stc"
ge-2/0/2 unit 0 family inet address 192.18.98.1/24
2. Enable OSPF on the interfaces defined in the previous step:
VCF#
VCF#
VCF#
VCF#
set protocols ospf area 0.0.0.0 interface xe-1/0/2.0
set protocols ospf area 0.0.0.0 interface xe-1/0/40.0
set protocols ospf area 0.0.0.0 interface ge-2/0/2.0
commit
The PE devices will redistribute into BGP any routes learned via OSPF from the CE
devices, and the L3 VPN in turn will forward these routes to other sites with VPN
tunnels.
On Juniper Virtual Chassis (PE device):
With L3 VPNs, the PE devices require the most extensive configuration. The steps
involved include the following:
•
•
•
•
•
•
•
•
•
Configuring a loopback address
Configuring addresses on interfaces connected to CE and P devices
Enabling LDP
Enabling MPLS
Configuring a backbone IGP (in this case, OSPF)
Configuring MP-BGP
Enabling at least one VRF instance
Configuring routing protocols for each VRF instance (or static routes)
Redistributing routes learned from CE devices into MP-BGP
21
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
1. Define loopback interface lo0 and configure an IP address for that interface. This
address will serve as the PE device’s router ID for BGP and LDP:
EX9208> configure
EX9208# set interfaces lo0 unit 0 family inet address 192.18.40.1/32
2. Define IP addresses for interfaces xe-5/0/5 (connected to the Juniper MX80 acting as a
P device) and xe-5/3/1 (connected to the Cisco Catalyst 3850 CE device):
EX9208#
0/0/1"
EX9208#
EX9208#
EX9208#
set interfaces xe-5/0/5 description "to Juniper MX80 P device xeset interfaces xe-5/0/5 unit 0 family inet address 192.18.38.1/24
set interfaces xe-5/3/1 description "to c3850 CE device int t1/1/3"
set interfaces xe-5/3/1 unit 0 family inet address 192.18.1.1/24
4. Configure LDP as the MPLS backbone label distribution protocol:
EX9208# set protocols ldp interface xe-5/0/5.0
5. Enable MPLS on interfaces lo0 and xe-5/0/5:
EX9208# set protocols mpls interface xe-5/0/5.0
EX9208# set protocols mpls interface lo0.0
EX9208# set interfaces xe-5/0/5 unit 0 family mpls
6. Enable OSPF on the service provider network (IS-IS also would work as an IGP):
EX9208# set protocols ospf traffic-engineering
EX9208# set protocols ospf area 0.0.0.0 interface xe-5/0/5.0
EX9208# set protocols ospf area 0.0.0.0 interface lo0.0
7. Configure MP-BGP to exchange routes with other PE devices. This example defines a
“Juniper-to-Cisco” BGP group in which the service provider network uses
autonomous system 100 (AS 100):
EX9208# set protocols bgp
EX9208# set protocols bgp
EX9208# set protocols bgp
EX9208# set protocols bgp
family inet-vpn unicast
EX9208# set protocols bgp
as 100
group
group
group
group
Juniper-to-Cisco
Juniper-to-Cisco
Juniper-to-Cisco
Juniper-to-Cisco
type internal
local-address 192.18.40.1
local-as 100
neighbor 192.18.69.1
group Juniper-to-Cisco neighbor 192.18.69.1 peer-
8. Configure a VRF instance. In this example, the VRF’s name is “VPN1”. The “routedistinguisher” command uniquely identifies this VRF’s network. RDs prevent
traffic misrouting when multiple customers use the same network space (for example,
when two customers both use net-10 addresses):
EX9208#
EX9208#
EX9208#
EX9208#
set
set
set
set
routing-instances
routing-instances
routing-instances
routing-instances
VPN1
VPN1
VPN1
VPN1
instance-type vrf
route-distinguisher 100:2
vrf-target target:100:2
vrf-table-label
22
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
9. Configure the VRF instance defined in the previous step on interface xe-5/3/1, which
connects with the CE device:
EX9208# set routing-instances VPN1 interface xe-5/3/1.0
10. Configure OSPF for routing between CE and PE devices. This step binds the VRF
instance called VPN1 to the routing protocols or static routes used at customer sites:
EX9208# set routing-instances VPN1 protocols ospf area 0.0.0.0 interface
xe-5/3/1.0
EX9208# set routing-instances VPN1 protocols ospf export default-export
EX9208# set routing-instances VPN1 protocols ospf import default-import
11. Configure the PE device to redistribute routes learned from CE devices into MPBGP. This example uses the routes learned from OSPF or BGP, but other routing
protocols or static routing could be used with other VRF instances. This example
redistributes routes from a policy statement called “Tsunami1”:
EX9208# set policy-options policy-statement Tsunami1-export-policy
from protocol ospf
EX9208# set policy-options policy-statement Tsunami1-export-policy
from protocol bgp
EX9208# set policy-options policy-statement Tsunami1-export-policy
then community add Tsunami1
EX9208# set policy-options policy-statement Tsunami1-export-policy
then accept
EX9208# set policy-options policy-statement Tsunami1-export-policy
then reject
EX9208# set policy-options policy-statement Tsunami1-import-policy
from protocol bgp
EX9208# set policy-options community Tsunami1 members target:100:2
EX9208# commit
term 1
term 1
term 1
term 1
term 2
term 1
On Juniper MX80 (P device):
The steps involved for configuration of a P device include the following:
•
•
•
•
•
Configuring a loopback address
Configuring addresses on interfaces connected to CE and P devices
Enabling LDP
Enabling MPLS
Configuring a backbone IGP (in this case, OSPF)
1. Define loopback interface lo0 and configure an IP address for that interface. This
address will serve as the PE device’s router ID for BGP and LDP:
mx80> configure
mx80# set groups global interfaces lo0 unit 0 family inet address
10.255.3.56/32 primary
23
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
2. Define IP addresses for interfaces xe-0/0/1 and xe-0/0/2 (connected to the Juniper
Virtual Chassis and Cisco Nexus 7010, respectively, each acting as PE devices):
mx80# set
5/0/5"
mx80# set
mx80# set
mx80# set
interfaces xe-0/0/1 description "to Juniper Virtual Chassis xeinterfaces xe-0/0/1 unit 0 family inet address 192.18.38.2/24
interfaces xe-0/0/2 description "to Cisco Nexus 7010 e3/10"
interfaces xe-0/0/2 unit 0 family inet address 192.18.70.2/24
3. Configure LDP as the MPLS backbone label distribution protocol:
mx80# set protocols ldp interface xe-0/0/1.0
mx80# set protocols ldp interface xe-0/0/2.0
mx80# set protocols ldp interface lo0.0
4. Enable MPLS on the interfaces connected to PE devices:
mx80#
mx80#
mx80#
mx80#
set
set
set
set
protocols mpls interface
protocols mpls interface
interfaces xe-0/0/1 unit
interfaces xe-0/0/2 unit
xe-0/0/1.0
xe-0/0/2.0
0 family mpls
0 family mpls
5. Enable OSPF on the service provider network. IS-IS would also work as an IGP:
mx80#
mx80#
mx80#
mx80#
set protocols ospf area 0.0.0.0 interface xe-0/0/1.0
set protocols ospf area 0.0.0.0 interface xe-0/0/2.0
set protocols ospf area 0.0.0.0 interface lo0.0
commit
Cisco configuration Cisco Catalyst 3850 (CE device):
1. Configure IP addresses for interfaces t1/1/3 (connected with the Juniper Virtual
Chassis) and g1/0/2 (connected with the Spirent TestCenter test instrument):
c3850# configure terminal
c3850(config)# interface TenGigabitEthernet1/1/3
c3850(config-if)# description to Juniper Virtual Chassis int xe-5/3/1
c3850(config-if)# no switchport
c3850(config-if)# ip address 192.18.1.2 255.255.255.0
c3850(config-if)# no shutdown
c3850(config-if)# interface GigabitEthernet1/0/2
c3850(config-if)# description to stc
c3850(config-if)# no switchport
c3850(config-if)# ip address 192.18.18.1 255.255.255.0
c3850(config-if)# no shutdown
c3850(config-if)# exit
2. Enable OSPF on the interfaces defined in the previous step:
c3850(config)# ip routing
c3850(config)# router ospf 1
c3850(config-rtr)# network 192.18.1.0 0.0.0.255 area 0
24
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
c3850(config-rtr)# network 192.18.18.0 0.0.0.255 area 0
c3850(config-rtr)# log-adjacency-changes
c3850(config-rtr)# end
Cisco Nexus 7010 (PE device):
1. Enable the various NX-OS feature sets required for L3 VPNs:
Nexus7010# configure terminal
Nexus7010(config)# feature ospf
Nexus7010(config)# feature bgp
Nexus7010(config)# feature mpls l3vpn
Nexus7010(config)# feature mpls ldp
2. Define a loopback interface and configure an IP address for that interface. This address
will serve as the PE device’s router ID for BGP and LDP:
Nexus7010(config)# interface loopback0
Nexus7010(config-int)# ip address 192.18.69.1/32
Nexus7010(config-int)# no shutdown
Note that an explicit no shutdown command is mandatory for Cisco Nexus 7000
devices.
3. Define IP addresses for interfaces e3/10 (connected to the Juniper MX80 acting as a P
device) and e3/3 (connected to the Juniper Virtual Chassis Fabric):
Nexus7010(config-int)#
Nexus7010(config-int)#
Nexus7010(config-int)#
Nexus7010(config-int)#
Nexus7010(config-int)#
Nexus7010(config-int)#
Nexus7010(config-int)#
Nexus7010(config-int)#
Nexus7010(config-int)#
interface e3/10
description to Juniper MX80 P device int xe-0/0/2
ip address 192.18.70.1/24
no shutdown
interface e3/3
description to Juniper Virtual Chassis Fabric
ip address 192.18.68.1/24
no shutdown
exit
4. Configure LDP as the MPLS backbone label distribution protocol:
Nexus7010(config)# mpls
Nexus7010(config-mpls)#
Nexus7010(config-mpls)#
Nexus7010(config-mpls)#
ldp configuration
router-id Eth3/10
neighbor 192.18.38.1 targeted
exit
5. Enable MPLS on interface e3/10 (connected to the P device):
Nexus7010(config)# interface e3/10
Nexus7010(config-int)# mpls ip
Nexus7010(config-int)# exit
6. Enable OSPF on the service provider network (IS-IS also would work as an IGP):
Nexus7010(config)# router ospf 1
Nexus7010(config-rtr)# router-id 192.18.69.1
25
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Nexus7010(config-rtr)#
Nexus7010(config-rtr)#
Nexus7010(config-rtr)#
Nexus7010(config-rtr)#
Nexus7010(config-rtr)#
network 192.18.69.1/32 area 0.0.0.0
network 192.18.70.0/24 area 0.0.0.0
mpls ldp autoconfig area 0.0.0.0
mpls ldp sync
exit
7. Enable BGP and configure MP-BGP to exchange routes with other PE devices. In this
example, the service provider network uses autonomous system 100 (AS 100):
Nexus7010(config)# router bgp 100
Nexus7010(config-rtr)# router-id 192.18.69.1
Nexus7010(config-rtr)# neighbor 192.18.40.1 remote-as 100
Nexus7010(config-rtr)# address-family vpnv4 unicast
Nexus7010(config-rtr)# send-community extended
Nexus7010(config-rtr)# exit
8. Configure a VRF instance. In this example, the VRF’s name is “VPN1”. The “rd”
command is a route distinguisher that uniquely identifies this VRF’s network. RDs
prevent traffic misrouting when multiple customers use the same network space (for
example, when two customers both use net-10 addresses):
Nexus7010(config)# vrf
Nexus7010(config-vrf)#
Nexus7010(config-vrf)#
Nexus7010(config-vrf)#
Nexus7010(config-vrf)#
Nexus7010(config-vrf)#
context VPN1
rd 100:2
address-family ipv4 unicast
route-target import 100:2
route-target export 100:2
exit
9. Configure the VRF instance defined in the previous step on interface e3/3, which
connects with the CE device:
Nexus7010(config)# int e3/3
Nexus7010(config-int)# vrf member VPN1
Nexus7010(config-int)# exit
10. Configure OSPF for routing between CE and PE devices. This step binds the VRF
instance to the routing protocols or static routes used at customer sites. This example
places the learned routes into a route map called rmap1:
Nexus7010(config)# route-map rmap1 permit 10
Nexus7010(config)# router ospf 1
Nexus7010(config-rtr)# vrf VPN1
Nexus7010(config-rtr)# network 192.18.68.0/24 area 0.0.0.0
Nexus7010(config-rtr)# redistribute bgp 100 route-map rmap1
Nexus7010(config-rtr)# exit
11. Configure the PE device to redistribute routes learned from CE devices into MPBGP. This example uses the routes learned from OSPF process 1, but other routing
protocols or static routing could be used with other VRF instances. This example
redistributes routes from the route map called rmap1:
Nexus7010(config)# router bgp 100
Nexus7010(config-router)# vrf VPN1
26
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Nexus7010(config-router-vrf)# address-family ipv4 unicast
Nexus7010(config-router-vrf)# redistribute direct route-map rmap1
Nexus7010(config-router-vrf)# redistribute ospf 1 route-map rmap1
Nexus7010(config)# end
Validation The Junos command “show route protocol bgp” will display routes learned via
MP-BGP across the service provider’s network. The equivalent command for the Cisco
Nexus 7010 is “show ip route bgp”. On both devices, the existence of routes
learned across the service provider’s MPLS network validates that BGP is working in the
L3 VPN.
On the data plane, a traffic generator such as Spirent TestCenter should be able to reach
sites across the service provider’s network across the L3 VPN tunnel.
Link aggregation Objective To validate the ability of Juniper and Cisco switches to correctly forward traffic over a
logical connection created using IEEE 802.3ad link aggregation.
To verify the ability of Juniper and Cisco switches to use the link aggregation control
protocol (LACP) to dynamically remove a member from a link aggregation group (LAG).
Background The IEEE 802.3ad link specification defines a standards-based method for aggregating
multiple physical Ethernet links into a single logical link. The logical link, known as a
link aggregation group (LAG), is comprised of multiple members (individual pairs of
physical interfaces on each switch). LAGs may be defined statically or dynamically, the
latter using the link aggregation control protocol (LACP). With LACP enabled, 802.3adcompliant switches can dynamically add or remove one or more members to a LAG.
Especially when used with LACP, link aggregation adds redundancy to network
connections. Depending on the number of flows and the hashing technique used, link
aggregation may also boost available bandwidth.
Link aggregation groups also can be defined across multiple chassis, as discussed in the
“Multi-Channel Link Aggregation Group” section in this document.
Topology In this example, a Cisco Catalyst 3850 switch uses a two-member LAG to exchange
traffic with a Juniper Virtual Chassis comprised of two Juniper EX9208 switches. The
Virtual Chassis could also be a Juniper EX4300, a Juniper QFX510, or both of these
27
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
combined into a Virtual Chassis Fabric; these all use the same link aggregation
commands given here. Note that LAG member interfaces can reside on different physical
switches in a Virtual Chassis or Virtual Chassis Fabric configuration; see the Junos
Software Configuration Guide for more details.
Interfaces xe-5/3/1 and xe-12/3/0 on the Juniper switch make up the members of the
LAG. On the Cisco switch, the LAG members are interfaces TenGigabitEthernet1/1/3
and TenGigabitEthernet1/1/4. LACP is enabled on all LAG members.
Figure 6 shows the topology used to validate link aggregation and LACP functionality.
Figure 6: Link aggregation validation topology Juniper commands Juniper Junos uses the ae interface notation to define each “aggregated Ethernet”
instance. The procedure is as follows:
1. Configure the desired number of link aggregation instances (just one, in this example):
[email protected]> configure
[email protected]# set chassis aggregated-devices ethernet device-count 1
2. Specify the members to be included within the aggregated Ethernet bundle:
[email protected]# set interfaces xe-5/3/1 ether-options 802.3ad ae1
[email protected]# set interfaces xe-12/3/0 ether-options 802.3ad ae1
3. Enable LACP on the aggregated Ethernet instance:
[email protected]# set interfaces ae1 aggregated-ether-options lacp active
[email protected]# set interfaces ae1 description "linkagg to 3850"
28
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
5. (Optional) Assign the link aggregation interface to be a member of a VLAN. The
following example assigns interface ae1.0 to access mode and allows traffic for
VLAN v2001:
[email protected]# set interfaces ae1.0 family ethernet-switching interface-mode
trunk
[email protected]# set interfaces ae1.0 family ethernet-switching vlan members
v2001
Link aggregation interfaces also can be configured in VLAN trunking mode to carry
tagged traffic from multiple VLANs. The following example assigns interface ae1.0 to
trunk-mode membership to carry traffic from VLANs v2001 and v2002:
[email protected]# set interfaces ae1.0 family ethernet-switching interface-mode
trunk
[email protected]# set interfaces ae1.0 family ethernet-switching vlan members
v2001
[email protected]# set interfaces ae1.0 family ethernet-switching vlan members
v2002
6. To disable or re-enable a member of the LAG, disable that member:
[email protected]# set interfaces xe-5/3/1 disable
Delete the disable command to re-enable the LAG member:
[email protected]# delete interfaces xe-0/1/1 disable
[email protected]# commit
Cisco commands 1. Create the link aggregation group (called a port-channel in Cisco IOS
terminology):
Cat3850# configure terminal
Cat3850(config)# interface Port-channel1
Cat3850(config-if)# switchport mode access
2. Add interfaces to the link aggregation group. The command “channel-group 1”
adds an interface to the link aggregation group defined in the previous step, while “mode
active” enables LACP (or “mode passive” if the other side of the LAG uses active
mode):
Cat3850(config)# interface TenGigabitEthernet1/1/3
Cat3850(config-if)# switchport mode access
Cat3850(config-if)# channel-group 1 mode passive
Cat3850(config)# interface TenGigabitEthernet1/1/4
Cat3850(config-if)# switchport mode access
Cat3850(config-if)# channel-group 1 mode passive
Cat3850(config-if)# end
29
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
These commands are for a Cisco Catalyst 3850. On Cisco Nexus devices running NXOS, the main difference is that each physical interface must be explicitly enabled with a
“no shutdown” command. For example, these commands put interface Ethernet3/9
into Port-channel 2:
Nexus7010(config)# interface Ethernet3/9
Nexus7010(config-if)# description linkagg to ex9200 xe-5/0/5 and xe-12/0/5
Nexus7010(config-if)# switchport
Nexus7010(config-if)# switchport mode trunk
Nexus7010(config-if)# channel-group 2 mode passive
Nexus7010(config-if)# no shutdown
Nexus7010(config-if)# end
Validation The Junos command “show lacp interfaces <aggregated Ethernet
interface>” will show LAG state. The following command was run after disabling
interface xe-12/3/0, and validates that LACP on both switches dynamically removed the
second member of the LAG. Note that interface xe-12/3/0 is in “Detached” state:
[email protected]# run show lacp interfaces ae1
Aggregated interface: ae1
LACP state:
Role
Exp
Def Dist Col Syn Aggr Timeout Activity
xe-5/3/1
Actor
No
No
Yes Yes Yes
Yes
Fast
Active
xe-12/3/0
Partner
No
No
Yes Yes Yes
Yes
Slow
Active
xe-5/3/1
Actor
No
Yes
No
No
No
Yes
Fast
Active
xe-12/3/0
Partner
No
Yes
No
No
No
Yes
Fast
Passive
LACP protocol:
Receive State Transmit State
Mux State
xe-5/3/1
Current
Slow periodic Collecting distributing
xe-12/3/0
Port disabled
No periodic
Detached
Link-­‐Layer Discovery Protocol (LLDP) Objective To verify the ability of Juniper and Cisco switches to exchange capabilities information
using LLDP.
Background LLDP, as described in the IEEE 802.1AB specification, is a standards-based method of
exchanging device capabilities. Unlike Cisco Discovery Protocol (CDP), covered
elsewhere in this document, LLDP is an open standard, and thus allows multiple
vendors’ devices to exchange capabilities data.
30
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Topology In this example, a Juniper Virtual Chassis comprising two Juniper EX9208 switches uses
LLDP to learn the MAC address (chassis ID), port information, and system name of a
Cisco Catalyst 3850.
Figure 7 shows the LLDP validation topology. A Juniper Virtual Chassis (comprised of
two Juniper EX9208 switches) connects to a Cisco Catalyst 3850 via two 10-gigabit
Ethernet interfaces. In this example, each switch is connected with two ports using a link
aggregation group (LAG), The LAG also serves as a VLAN trunk port, with traffic using
VLAN IDs 2001-2003 allowed. LLDP would also work without link aggregation, and
with the two switch ports configured in access mode. Also note that this example
assumes spanning tree protocol (STP) has been disabled on both switches, although
LLDP would also work with STP enabled.
Figure 7: LLDP validation topology Juniper commands 1. Define VLANs v2001, v2002, and v2003:
[email protected]>
[email protected]#
[email protected]#
[email protected]#
configure
set vlans v2001 vlan-id 2001
set vlans v2002 vlan-id 2002
set vlans v2003 vlan-id 2003
2. (Optional) Place interfaces xe-5/1/0 and xe-12/3/0 into trunk mode and allow tagged
traffic for the VLANs defined in the previous step. VLAN trunking is optional, and is not
required for LLDP to work:
[email protected]# set interfaces xe-5/3/1 unit 0 family ethernet-switching
interface-mode trunk
[email protected]# set interfaces xe-5/3/1 unit 0 family ethernet-switching vlan
members v2001
31
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
[email protected]# set interfaces
members v2002
[email protected]# set interfaces
members v2003
[email protected]# set interfaces
interface-mode trunk
[email protected]# set interfaces
vlan members v2001
[email protected]# set interfaces
vlan members v2002
[email protected]# set interfaces
vlan members v2003
xe-5/3/1 unit 0 family ethernet-switching vlan
xe-5/3/1 unit 0 family ethernet-switching vlan
xe-12/3/0 unit 0 family ethernet-switching
xe-12/3/0 unit 0 family ethernet-switching
xe-12/3/0 unit 0 family ethernet-switching
xe-12/3/0 unit 0 family ethernet-switching
3. (Optional) Create link aggregation group (LAG) ae1 and add interfaces to the LAG.
Link aggregation is optional, and is not required for LLDP to work:
[email protected]#
[email protected]#
[email protected]#
[email protected]#
set protocols lldp interface all
set interfaces ae1 description "linkagg to 3850"
set interfaces ae1 aggregated-ether-options lacp active
set interfaces ae1 unit 0 family ethernet-switching interface-
mode trunk
[email protected]# set interfaces ae1 unit 0 family ethernet-switching vlan members
v2001
[email protected]# set interfaces ae1 unit 0 family ethernet-switching vlan members
v2002
[email protected]# set interfaces ae1 unit 0 family ethernet-switching vlan members
v2003
[email protected]#
[email protected]#
[email protected]#
[email protected]#
set
set
set
set
interfaces
interfaces
interfaces
interfaces
xe-5/3/1 description "ae1 linkagg to 3850"
xe-5/3/1 ether-options 802.3ad ae1
xe-12/3/0 description "ae1 linkagg to 3850"
xe-12/3/0 ether-options 802.3ad ae1
4. (Optional) Enable LLDP. On Juniper switches, LLDP is enabled by default on all
interfaces; if LLDP has not been disabled, skip this step. The following (optional)
command enables LLDP on all interfaces but it also can be set on a per-interface basis:
[email protected]# set protocols lldp interface all
5. (Optional) Disable rapid spanning tree protocol (RSTP). In this example RSTP is
disabled on all interfaces but it also can be set on a per-interface basis:
[email protected]# delete protocols rstp
[email protected]# commit
Cisco commands 1. Define VLANs 2001-2003:
Cat3850# configure terminal
Cat3850(config)# vlan 2001
Cat3850(config-vlan)# exit
Cat3850(config)# vlan 2002
Cat3850(config-vlan)# exit
Cat3850(config)# vlan 2002
Cat3850(config-vlan)# exit
32
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
2. (Optional) Define link aggregation group Port-channel1 and configure it for
VLAN trunking. Link aggregation and VLAN trunking are optional, and are not required
for LLDP to work:
Cat3850(config)# interface Port-channel1
Cat3850(config-if)# description linkagg to EX9208
Cat3850(config-if)# switchport trunk allowed vlan 2001-2003
Cat3850(config-if)# switchport mode trunk
3. Define interfaces TenGigabitEthernet1/1/3 and TenGigabitEthernet1/1/4 as VLAN
trunk ports and add them to Port-channel1:
Cat3850(config)# interface range TenGigabitEthernet1/1/3-4
Cat3850(config-if-range)# description linkagg to EX9208
Cat3850(config-if-range)# switchport trunk allowed vlan 2001-2003
Cat3850(config-if-range)# switchport mode trunk
Cat3850(config-if-range)# channel-group 1 mode passive
Cat3850(config-if-range)# exit
4. Enable LLDP. On the Cisco Catalyst 3850, this command applies systemwide:
Cat3850(config)# lldp run
For a Cisco Nexus 7010, the LLDP feature must be enabled. It too applies systemwide:
Nexus7010(config)# feature lldp
5. Disable spanning tree for VLANs 2001-2003:
Cat3850(config)# no spanning-tree vlan 2001-2003
Cat3850(config)# end
33
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Validation On the Juniper switch, the command show lldp neighbors will verify that the
Cisco switch is attached to interfaces xe-5/3/1 and xe-12/3/0:
[email protected]> show lldp neighbors
Local Interface
Parent Interface
System Name
xe-5/3/1
dc-tme-c3850-01.englab.juniper.net
xe-12/3/0
ae1
dc-tme-c3850-01.englab.juniper.net
Chassis Id
Port info
0c:27:24:ce:95:80
Te1/1/3
0c:27:24:ce:95:80
Te1/1/4
Cisco Catalyst and Nexus switches also use the command show lldp neighbors:
Cat3850# show lldp neighbors
Capability codes:
(R) Router, (B) Bridge, (T) Telephone, (C) DOCSIS Cable Device
(W) WLAN Access Point, (P) Repeater, (S) Station, (O) Other
Device ID
EX9208
EX9208
Local Intf
Te1/1/4
Te1/1/3
Hold-time
120
120
Capability
B,R
B,R
Port ID
1621
1591
Multi-­‐channel link aggregation group (MC-­‐LAG) Objective To verify the ability of multiple Juniper switches to provide multi-channel link
aggregation groups, presenting a single logical interface to a Cisco Catalyst access switch
and to a Cisco core switch.
Background To attached switches, an MC-LAG looks and functions the same as standard 802.3ad link
aggregation: One or more physical interfaces on the attached switch bond together to
form a single logical interface.
The difference with MC-LAG is on the other end: A single logical interface spans
multiple physical switches, adding extra resiliency even if an entire switch fails. MCLAG uses standard LACP messages inside the link aggregation group (LAG), and passes
messages between switch chassis to monitor device state.
Switches attached to an MC-LAG require no special configuration beyond the usual link
aggregation commands.
34
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Topology In this example, two Juniper EX9208 chassis form two MC-LAGs, one apiece to a Cisco
Catalyst 3850 access switch and to a Cisco Nexus 7010 core switch. Notably, the Juniper
core switches do not use Juniper Virtual Chassis technology in this example. Instead,
each EX9208 is a standalone switch.
Figure 8 illustrates the topology used to validate MC-LAG interoperability.
Figure 8: MC-­‐LAG validation topology Juniper commands In this example, the steps required are:
•
•
•
Configure an inter-switch link to carry Inter-Control Center Communications
Protocol (ICCP) messages between EX9208 chassis
Configure an MC-LAG with the Cisco access switch
Configure an MC-LAG with the Cisco core switch
These steps repeated for Switches A and B.
35
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
1. On switch A, begin by configuring an IP address on the loopback interface lo0. This
will ensure MC-LAG will continue to function even if the inter-switch link or other
EX9208 fails:
EX9208A> configure
EX9208A# set interfaces lo0 unit 0 family inet address 192.18.40.2/24
2. Next, configure the inter-switch link:
EX9208A# set chassis aggregated-devices ethernet device-count 10
EX9208A# set interfaces xe-0/0/0 description "ICCP port to EX9208B"
EX9208A# set interfaces xe-0/0/0 unit 0 family inet address 192.18.39.1/24
EX9208A# set multi-chassis multi-chassis-protection 192.18.39.2 interface xe0/0/0
EX9208A# set protocols iccp local-ip-addr 192.18.39.1
EX9208A# set protocols iccp peer 192.18.39.2 session-establishment-hold-time 50
EX9208A# set protocols iccp peer 192.18.39.2 redundancy-group-id-list 1
EX9208A# set protocols iccp peer 192.18.39.2 backup-liveness-detection backuppeer-ip 192.18.40.1
EX9208A# set protocols iccp peer 192.18.39.2 liveness-detection minimumreceive-interval 60
EX9208A# set protocols iccp peer 192.18.39.2 liveness-detection transmitinterval minimum-interval 60
3. Configure a link aggregation group ae1 for connection to the Cisco Catalyst 3850. In
this case the LAG is also a trunk that allows traffic for all VLANs, though VLAN
configuration is not required for MC-LAG to work:
EX9208A# set interfaces
EX9208A# set interfaces
EX9208A# set interfaces
EX9208A# set interfaces
00:01:02:03:04:05
EX9208A# set interfaces
EX9208A# set interfaces
EX9208A# set interfaces
1
EX9208A# set interfaces
EX9208A# set interfaces
active
EX9208A# set interfaces
active
EX9208A# set interfaces
trunk
EX9208A# set interfaces
all
xe-0/3/0 description "MC-LAG to 3850"
xe-0/3/0 ether-options 802.3ad ae1
ae1 aggregated-ether-options lacp active
ae1 aggregated-ether-options lacp system-id
ae1 aggregated-ether-options lacp admin-key 3
ae1 aggregated-ether-options mc-ae mc-ae-id 3
ae1 aggregated-ether-options mc-ae redundancy-group
ae1 aggregated-ether-options mc-ae chassis-id 0
ae1 aggregated-ether-options mc-ae mode activeae1 aggregated-ether-options mc-ae status-control
ae1 unit 0 family ethernet-switching interface-mode
ae1 unit 0 family ethernet-switching vlan members
4. Configure a link aggregation group ae2 for connection to the Cisco Nexus 7010. In
this case the LAG is also a trunk that allows traffic for all VLANs, though VLAN
configuration is not required for MC-LAG to work:
EX9208A# set interfaces
EX9208A# set interfaces
EX9208A# set interfaces
EX9208A# set interfaces
00:01:02:03:04:05
xe-0/0/5 description "MC-LAG to 7010"
xe-0/0/5 ether-options 802.3ad ae2
ae2 aggregated-ether-options lacp active
ae2 aggregated-ether-options lacp system-id
36
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
EX9208A#
EX9208A#
EX9208A#
1
EX9208A#
EX9208A#
active
EX9208A#
active
EX9208A#
trunk
EX9208A#
all
EX9208A#
set interfaces ae2 aggregated-ether-options lacp admin-key 3
set interfaces ae2 aggregated-ether-options mc-ae mc-ae-id 2
set interfaces ae2 aggregated-ether-options mc-ae redundancy-group
set interfaces ae2 aggregated-ether-options mc-ae chassis-id 0
set interfaces ae2 aggregated-ether-options mc-ae mode activeset interfaces ae2 aggregated-ether-options mc-ae status-control
set interfaces ae2 unit 0 family ethernet-switching interface-mode
set interfaces ae2 unit 0 family ethernet-switching vlan members
commit
5. On switch B, begin by configuring an IP address on the loopback interface lo0. This
will ensure MC-LAG will continue to function even if the inter-switch link or other
EX9208 fails:
EX9208B> configure
EX9208B# set interfaces lo0 unit 0 family inet address 192.18.40.1/24
6. Next, configure the inter-switch link:
EX9208B# set chassis aggregated-devices ethernet device-count 10
EX9208B# set interfaces xe-5/0/0 description "ICCP port to EX9208A"
EX9208B# set interfaces xe-5/0/0 unit 0 family inet address 192.18.39.2/24
7. Configure a link aggregation group ae1 for connection to the Cisco Catalyst 3850. In
this case the LAG is also a trunk that allows traffic for all VLANs, though VLAN
configuration is not required for MC-LAG to work:
EX9208B# set interfaces
EX9208B# set interfaces
EX9208B# set interfaces
EX9208B# set interfaces
00:01:02:03:04:05
EX9208B# set interfaces
EX9208B# set interfaces
EX9208B# set interfaces
1
EX9208B# set interfaces
EX9208B# set interfaces
active
EX9208B# set interfaces
standby
EX9208B# set interfaces
trunk
EX9208B# set interfaces
all
xe-5/3/1 description "linkagg to 3850"
xe-5/3/1 ether-options 802.3ad ae1
ae1 aggregated-ether-options lacp active
ae1 aggregated-ether-options lacp system-id
ae1 aggregated-ether-options lacp admin-key 3
ae1 aggregated-ether-options mc-ae mc-ae-id 3
ae1 aggregated-ether-options mc-ae redundancy-group
ae1 aggregated-ether-options mc-ae chassis-id 1
ae1 aggregated-ether-options mc-ae mode activeae1 aggregated-ether-options mc-ae status-control
ae1 unit 0 family ethernet-switching interface-mode
ae1 unit 0 family ethernet-switching vlan members
37
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
8. Configure a link aggregation group ae2 for connection to the Cisco Nexus 7010. In
this case the LAG is also a trunk that allows traffic for all VLANs, though VLAN
configuration is not required for MC-LAG to work:
EX9208B# set interfaces
EX9208B# set interfaces
EX9208B# set interfaces
EX9208B# set interfaces
00:01:02:03:04:05
EX9208B# set interfaces
EX9208B# set interfaces
EX9208B# set interfaces
1
EX9208B# set interfaces
EX9208B# set interfaces
active
EX9208B# set interfaces
standby
EX9208B# set interfaces
trunk
EX9208B# set interfaces
all
EX9208B# commit
xe-5/0/5 description "linkagg to 7010"
xe-5/0/5 ether-options 802.3ad ae2
ae2 aggregated-ether-options lacp active
ae2 aggregated-ether-options lacp system-id
ae2 aggregated-ether-options lacp admin-key 3
ae2 aggregated-ether-options mc-ae mc-ae-id 2
ae2 aggregated-ether-options mc-ae redundancy-group
ae2 aggregated-ether-options mc-ae chassis-id 1
ae2 aggregated-ether-options mc-ae mode activeae2 aggregated-ether-options mc-ae status-control
ae2 unit 0 family ethernet-switching interface-mode
ae2 unit 0 family ethernet-switching vlan members
Cisco commands As noted, no special MC-LAG configuration is needed on Cisco devices. The only
required steps are to define link aggregation groups (called “port channels” in Cisco
parlance) on each device, as previously discussed in the “link aggregation” section.
Here are the steps required to configure link aggregation on the Cisco Catalyst 3850:
1. Define a port channel to connect with the MC-LAG. In this example, the interface
Port-channel1 is also a trunk that allows traffic for all VLANs, though VLAN
configuration is not required for link aggregation to work:
c3850# configure terminal
c3850(config)# interface Port-channel1
c3850(config-if)# switchport mode trunk
c3850(config-if)# exit
2. Assign interfaces to Port-channel1:
c3850(config)# interface TenGigabitEthernet1/1/3
c3850(config-if)# description linkagg to EX9208A
c3850(config-if)# switchport mode trunk
c3850(config-if)# channel-group 1 mode passive
c3850(config-if)# interface TenGigabitEthernet1/1/4
c3850(config-if)# description linkagg to EX9208B
c3850(config-if)# switchport mode trunk
c3850(config-if)# channel-group 1 mode passive
c3850(config-if)# end
38
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Here are the steps required to configure link aggregation on the Cisco Nexus 7010:
3. Enable interface configuration and LACP support:
Nexus7010# configure terminal
Nexus7010(config)# feature interface-vlan
Nexus7010(config)# feature lacp
4. Define a port channel to connect with the MC-LAG. In this example, the interface
Port-channel2 is also a trunk that allows traffic for VLANs 2001-2003, though
VLAN configuration is not required for link aggregation to work:
Nexus7010(config)# interface Port-channel2
Nexus7010(config-if)# description linkagg to EX9208
Nexus7010(config-if)# switchport
Nexus7010(config-if)# switchport mode trunk
Nexus7010(config-if)# switchport trunk allowed vlan 2001-2003
Nexus7010(config-if)# exit
5. Assign interfaces to Port-channel3. Note that switches running NX-OS require an
explicit “no shutdown” command to enable an interface:
Nexus7010(config)# interface Ethernet3/9
Nexus7010(config-if)# description linkagg to EX9208A
Nexus7010(config-if)# switchport
Nexus7010(config-if)# switchport mode trunk
Nexus7010(config-if)# switchport trunk allowed vlan 2001-2003
Nexus7010(config-if)# channel-group 2 mode passive
Nexus7010(config-if)# no shutdown
Nexus7010(config-if)# interface Ethernet3/10
Nexus7010(config-if)# description linkagg to EX9208B
Nexus7010(config-if)# switchport
Nexus7010(config-if)# switchport mode trunk
Nexus7010(config-if)# switchport trunk allowed vlan 2001-2003
Nexus7010(config-if)# channel-group 2 mode active
Nexus7010(config-if)# no shutdown
Nexus7010(config-if)# end
39
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Validation On the Juniper EX9208 switches, the command “show interfaces mc-ae” will display
the status of configured MC-LAGs. For example, here is the output of that command on
EX9208A:
[email protected]# run show interfaces mc-ae
Member Link
: ae1
Current State Machine's State: mcae active state
Local Status
: active
Local State
: up
Peer Status
: active
Peer State
: up
Logical Interface
: ae1.0
Topology Type
: bridge
Local State
: up
Peer State
: up
Peer Ip/MCP/State
: 192.18.39.2 xe-0/0/0.0 up
Member Link
: ae2
Current State Machine's State: mcae active state
Local Status
: active
Local State
: up
Peer Status
: active
Peer State
: up
Logical Interface
: ae2.0
Topology Type
: bridge
Local State
: up
Peer State
: up
Peer Ip/MCP/State
: 192.18.39.2 xe-0/0/0.0 up
Similarly, the command “show iccp” will display the status of the inter-switch link:
[email protected]# run show iccp
Redundancy Group Information for peer 192.18.39.2
TCP Connection
: Established
Liveliness Detection : Up
Backup liveness peer status: Up
Redundancy Group ID
Status
1
Up
Client Application: l2ald_iccpd_client
Redundancy Group IDs Joined: 1
Client Application: lacpd
Redundancy Group IDs Joined: 1
Also, as with regular link aggregation, the command “show interfaces <aeX>
detail” will display information on the status of LAG X, where X is the LAG ID.
In the event of a link failure between Juniper and Cisco switches, these same commands
will also indicate a change in LAG status.
On Cisco Catalyst 3850 switches , the command “show etherchannel summary”
will display the status of LAG members. On Cisco Nexus 7010 switches, the equivalent
40
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
command is “show port-channel summary”. Again, no MC-LAG awareness or
configuration is required on either Cisco switch.
Multicast routing Objective To verify the ability of a network comprised of Juniper and Cisco devices to learn
multicast routing information using the PIM-SM protocol.
To verify the ability of a network comprised of Juniper and Cisco devices to correctly
forward multicast traffic based on routing information learned via PIM-SM.
Background Protocol Independent Multicast-Sparse Mode (PIM-SM) is a popular choice for multicast
routing. Devices running PIM-SM can learn topology information from other PIM-SM
routers and make forwarding decisions based on that information.
Like all multicast protocols, PIM-SM uses reverse path forwarding (RPF) lookups to
determine which router interface is closest to the multicast source. Because PIM-SM does
not include a mechanism to populate an RPF table, it relies on a unicast routing protocol
such as Open Shortest Path First (OSPF) or Intermediate System-to-Intermediate System
(IS-IS) for this purpose.
Topology This example is similar to that used in the “IP multicast switching” section, with one
important change: Here, both Juniper and Cisco devices act as routers running PIM-SM
and OSPF. There are no switches, and therefore no IGMP snooping, in this scenario.
In this example, a video server generates multicast traffic to one subnet of a Juniper
Virtual Chassis (comprised of two Juniper EX9208s). The Juniper device uses PIM-SM
to propagate routing information about that network to other networks, including one in
which a Cisco Catalyst 3850 switch, also running PIM-SM, is attached.
Both the Juniper and Cisco devices use PIM-SM and OSPF to propagate routing
information. Multicast subscribers attached to routed interfaces, each in a different IP
subnet, receive traffic from the streaming video server. The subscriber interfaces also use
IGMP (not IGMP snooping) to build a multicast forwarding table.
Figure 9 illustrates the topology used to validate IP multicast routing functionality. PIMSM and OSPF routing is enabled on both Juniper and Cisco devices.
41
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
42
Figure 9: Multicast routing validation topology Juniper commands 1. Assign IP addresses to the interfaces:
[email protected]>
[email protected]#
[email protected]#
[email protected]#
[email protected]#
configure
set interfaces
set interfaces
set interfaces
set interfaces
ge-2/0/0
ge-2/0/0
xe-5/3/1
ge-5/3/1
description "to video server"
unit 0 family inet address 192.18.32.1/24
description "to Cisco Catalyst 3850"
unit 0 family inet address 192.18.1.1/24
If the interfaces previously used the ethernet-switching keyword, it should be
deleted first with the “delete interfaces <name> family ethernetswitching” command.
2. Enable PIM-SM on both interfaces:
[email protected]# set protocols pim interface ge-2/0/0.0 mode sparse
[email protected]# set protocols pim interface ge-5/3/1.0 mode sparse
3. Enable OSPF on both interfaces. This step is not strictly necessary for IP multicast
forwarding, but it is required for PIM-SM routing to build an RPF table:
[email protected]# set protocols ospf area 0.0.0.0 interface ge-2/0/0.0
[email protected]# set protocols ospf area 0.0.0.0 interface xe-5/3/1.0
4. Configure the Juniper device to act as a rendezvous point (RP), in this case by
statically assigning an IP address:
[email protected]# set protocols pim rp local address 192.18.32.1
[email protected]# commit
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
There are also dynamic methods of discovering and assigning RPs, but those are beyond
the scope of this document. Interdomain Multicast Routing, mentioned in the “Intended
audience” section, covers dynamic RP configuration for Juniper and Cisco devices.
Cisco commands The following commands apply to a Cisco Catalyst 3850. Except where noted, the syntax
is similar for NX-OS devices such as the Cisco Nexus 7010.
1. Enable IP routing and IP multicast routing:
Cat3850# configure terminal
Cat3850(config)# ip routing
Cat3850(config)# ip multicast-routing
On the Cisco Nexus 7010, the commands are different:
Nexus7010# configure terminal
Nexus7010(config)# feature ospf
Nexus7010(config)# feature pim
2. Configure IP addresses and PIM-SM for each interface in use:
Cat3850(config)# interface TenGigabitEthernet1/1/3
Cat3850(config-if)# description to EX9200 Virtual Chassis
Cat3850(config-if)# no switchport
Cat3850(config-if)# ip address 192.18.1.2 255.255.255.0
Cat3850(config-if)# ip pim sparse-mode
Cat3850(config-if)# ip igmp version 3
Cat3850(config-if)# no shutdown
Cat3850(config-if)# interface GigabitEthernet1/0/1
Cat3850(config-if)# description to EX9200 Virtual Chassis
Cat3850(config-if)# no switchport
Cat3850(config-if)# ip address 192.18.4.1 255.255.255.0
Cat3850(config-if)# ip pim sparse-mode
Cat3850(config-if)# ip igmp version 3
Cat3850(config-if)# no shutdown
Cat3850(config-if)# interface GigabitEthernet1/0/2
Cat3850(config-if)# description to EX9200 Virtual Chassis
Cat3850(config-if)# no switchport
Cat3850(config-if)# ip address 192.18.5.1 255.255.255.0
Cat3850(config-if)# ip pim sparse-mode
Cat3850(config-if)# ip igmp version 3
Cat3850(config-if)# no shutdown
Cat3850(config-if)# interface GigabitEthernet1/0/3
Cat3850(config-if)# description to EX9200 Virtual Chassis
Cat3850(config-if)# no switchport
Cat3850(config-if)# ip address 192.18.6.1 255.255.255.0
Cat3850(config-if)# ip pim sparse-mode
Cat3850(config-if)# ip igmp version 3
Cat3850(config-if)# no shutdown
Cat3850(config-if)# exit
43
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
The “no shutdown” command is optional with the Catalyst 3850, but mandatory for
the Nexus 7010 and other devices running the NX-OS operating system.
Also, note that in this example IP addresses are configured directly on physical
interfaces. If desired, IP addresses and routing information can instead be assigned to
VLAN interfaces. In that case, each physical interface should be put into switchport
mode and assigned to a VLAN. The “VLAN Trunking” section of this document has
more details.
3. Configure OSPF. This step is not strictly necessary for IP multicast forwarding, but it
is required for PIM-SM routing to build an RPF table:
Cat3850(config)# router ospf 1
Cat3850(config-rtr)# log-adjacency-changes
Cat3850(config-rtr)# network 192.18.1.0 0.0.0.255
Cat3850(config-rtr)# network 192.18.4.0 0.0.0.255
Cat3850(config-rtr)# network 192.18.5.0 0.0.0.255
Cat3850(config-rtr)# network 192.18.6.0 0.0.0.255
Cat3850(config-if)# exit
area
area
area
area
0
0
0
0
4. Configure a PIM rendezvous point (RP), which in this case was statically defined on
the Juniper device:
Cat3850(config)# ip pim rp-address 192.18.32.1
Cat3850(config-if)# end
Validation Once subscribers attached to the Cisco Catalyst 3850 have joined multicast groups by
sending IGMPv3 reports with join messages, any multicast traffic for those groups
offered to interface xe-1/0/40 on the Juniper Virtual Chassis will be forwarded to all
subscriber ports on the Cisco Catalyst 3850.
The Junos command show pim neighbors brief also will verify that the
Juniper and Cisco devices see one another and can exchange topology updates. On
Cisco devices, the equivalent command is “show ip pim neighbors”.
Multicast switching Objective To verify the ability of Juniper and Cisco switches to correctly forward traffic from a
network using IGMP snooping for multicast switching.
44
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Background Ethernet switches use Internet group management protocol (IGMP) snooping to
determine where a switch should forward multicast traffic. With IGMP snooping enabled,
a switch listens for IGMP reports from attached devices that wish to receive multicast
traffic. The switch then maps subscribed multicast group address(es) to the interface on
which the subscriber is attached. When the switch receives traffic destined for an IP
multicast group address, it will forward it only to those interfaces from which it has heard
membership reports.
Topology In this example, both Juniper and Cisco switches operate purely in Layer-2 mode, with no
multicast or unicast routing protocols configured. This test case assumes routing is
handled elsewhere in the network.
A video server generates multicast traffic that is routed across a network and reaches the
two switches described here. The switches, in turn, use their IGMP snooping tables to
determine which ports should and should not receive multicast traffic.
The streaming video server sends traffic to 10 multicast group addresses in the range of
225.0.1.0 through 225.0.1.9. Subscribers attached to the Juniper and Cisco switches join
all 10 multicast groups.
Figure 10 illustrates the topology used to validate IP multicast switching functionality.
Both the Juniper and Cisco switches use IGMP snooping.
Figure 10: Multicast switching validation topology Juniper commands In this example, IGMP snooping is enabled on a per-VLAN basis. For example, these
commands create a VLAN called “v2001” and then enable IGMP snooping on that
VLAN:
45
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
[email protected]> configure
[email protected]# set vlans v2001 vlan-id 2001
[email protected]# set protocols igmp-snooping vlan v2001
If desired, IGMP snooping can be enabled on all VLANs using the all keyword:
[email protected]# set protocols igmp-snooping vlan all
Next, configure both Juniper ports to use VLAN trunking and to allow traffic from
VLAN “v2001”:
[email protected]# set
[email protected]# set
interface-mode
[email protected]# set
members v2001
[email protected]# set
[email protected]# set
interface-mode
[email protected]# set
members v2001
interfaces xe-1/0/10 description "to 3850"
interfaces xe-1/0/10 unit 0 family ethernet-switching
trunk
interfaces xe-1/0/10 unit 0 family ethernet-switching vlan
interfaces xe-1/0/40 description "to router"
interfaces xe-1/0/40 unit 0 family ethernet-switching
trunk
interfaces xe-1/0/40 unit 0 family ethernet-switching vlan
Finally, enable IGMP version 3, the most recent version of the protocol on both
interfaces:
[email protected]# set protocols igmp interface xe-1/0/10.0 version 3
[email protected]# set protocols igmp interface xe-1/0/40.0 version 3
[email protected]# commit
Cisco commands The following commands apply to a Cisco Catalyst 3850. Except where noted, the syntax
is similar for Nexus 7010 switches.
1. Create VLAN 2001:
Cat3850# configure terminal
Cat3850(config)# vlan 2001
Cat3850(config-vlan)# exit
2. Configure interface TenGigabitEthernet1/1/3 (connected to the Juniper Virtual Chassis
Fabric) for VLAN trunking and to allow VLAN 2001:
Cat3850(config)# interface TenGigabitEthernet1/1/3
Cat3850(config-if)# switchport mode trunk
Cat3850(config-if)# switchport trunk allowed vlan 2001
Cat3850(config-if)# no shutdown
Cat3850(config-if)# exit
The “no shutdown” command is optional with the Catalyst 3850, but mandatory for
the Nexus 7010 and other devices running the NX-OS operating system.
46
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
3. Set the IP address of the IGMP querier. Usually this is a multicast-enabled router near
the switches:
Cat3850(config)# ip igmp snooping querier address 192.18.36.1
Cat6500(config)# end
Validation On the Juniper Virtual Chassis Fabric, results of the command show igmpsnooping membership will verify that the switch has correctly mapped multicast
groups to the appropriate subscriber interfaces.
[email protected]> show igmp snooping membership | no-more
Instance: default-switch
Vlan: v2001
Learning-Domain: default
Interface: xe-1/0/10.0, Groups: 10
Group: 225.0.1.0
Group mode: Exclude
Source: 0.0.0.0
Last reported by: 10.205.0.2
Group timeout:
223 Type: Dynamic
Group: 225.0.1.1
Group mode: Exclude
Source: 0.0.0.0
Last reported by: 10.205.0.2
Group timeout:
223 Type: Dynamic
..
Interface: xe-1/0/40.0, Groups: 0
The output for interface xe-1/0/10 will continue through multicast group address
225.0.1.9. Note that interface xe-1/0/40 shows no associated multicast groups. This is
because all subscribers are connected to the Cisco switch. There are no multicast group
subscribers directly attached to the Juniper switch. Thus, incoming packets with a
destination address of an IP multicast group should only be forwarded through interface
xe-1/0/10.
The command show interface <name> extensive will verify correct
forwarding of multicast traffic. The “Multicast packets” counter (under “MAC
statistics”) will increment on interfaces with multicast subscribers attached, and will not
increment on other interfaces.
On Cisco devices, the command “show ip igmp snooping groups” will
display information about multicast group membership.
47
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Real-­‐Time Performance Monitoring (RPM) Objective To validate the ability of Juniper EX Series switches to perform real-time health checks
on attached devices.
Background Juniper’s Real-Time Performance Monitoring (RPM) feature can perform “health
checks” on attached network devices and servers using ICMP, HTTP, TCP and UDP
probes and requests. These active probes monitor devices across any network and
investigate reachability problems. RPM keeps a history of the most recent 50 probes;
such monitoring over time can be useful in troubleshooting and capacity planning.
Topology In this example, a Juniper Virtual Chassis comprising two Juniper EX9208 switches uses
uses ICMP probes to monitor round-trip times between it and a Cisco Catalyst 3850E
switch. Note that no RPM-specific commands are needed on the Cisco switch. The
same RPM configuration on the Juniper switch will work with any Cisco switch, or
indeed any remote device capable of responding to pings (ICMP probe requests).
Figure 11 illustrates the RPM validation test bed. The Juniper Virtual Chassis in this
example uses an IP address of 192.18.0.10/24 assigned to interface xe-5/3/1. This
interface sends ICMP probe requests to a Cisco switch with an address of
192.18.0.1/24. The same RPM configuration would work in a Juniper switch
configuration in which a VLAN is created and an IP address is assigned to the VLAN.
Figure 11: Real-­‐Time Performance Monitoring validation topology 48
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Juniper configuration 1. Assign an IP address of 192.18.0.10/24 to interface xe-5/3/1:
[email protected]> configure
[email protected]# set interfaces ge-0/0/22 unit 0 family inet address
192.18.0.10/24
2. Define an RPM probe and test for the Cisco switch at 192.18.0.1:
[email protected]#
timestamp
[email protected]#
192.18.0.1
[email protected]#
[email protected]#
[email protected]#
[email protected]#
[email protected]#
set services rpm probe myprobe test t1 probe-type icmp-pingset services rpm probe myprobe test t1 target address
set services
set services
set services
set services
commit
rpm
rpm
rpm
rpm
probe
probe
probe
probe
myprobe
myprobe
myprobe
myprobe
test
test
test
test
t1
t1
t1
t1
probe-count 10
probe-interval 1
test-interval 15
data-size 128
Note that although only ICMP is used here, a single probe can encompass multiple tests
using multiple types of health checks.
Cisco configuration 1. Assign an IP address of 192.18.0.1/24 to the monitored interface (in this case
TenGigabitEthernet1/1/3):
Cat3850# configure terminal
Cat3850(config)# interface GigabitEthernet1/1/3
Cat3850(config-if)# no switchport
Cat3850(config-if)# ip address 192.18.0.1 255.255.255.0
Cat3850(config-if)# end
Cisco Nexus 7000 switches running the NX-OS operating system also require an explicit
“no shutdown” command to enable an interface.
No RPM-specific configuration is needed on the Cisco switch, or on any other device
monitored using RPM.
49
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Validation The command show services rpm history-results will display up to 50
results of RPM probes and tests:
[email protected]# run show services rpm history-results
Owner, Test
myprobe, t1
Probe received
Fri Apr 18 00:31:20 2014
Round trip time
2802
myprobe, t1
Fri Apr 18 00:31:21 2014
5441
myprobe, t1
Fri Apr 18 00:31:22 2014
9909
myprobe, t1
Fri Apr 18 00:31:23 2014
2754
myprobe, t1
Fri Apr 18 00:31:24 2014
3334
usec
usec
usec
usec
usec
Redundant Trunk Group (RTG) Objective To validate failover functionality of Juniper’s Redundant Trunk Group (RTG) feature
between Juniper and Cisco switches.
Background Juniper’s Redundant Trunk Group (RTG) feature allows definition of primary and
secondary paths between switches and redirects traffic across the secondary trunk if the
primary link fails. RTG provides an alternative to spanning tree bridging for
redundancy. RTG works in mixed Juniper-Cisco environments with no additional
configuration needed on Cisco switch ports. Up to 16 redundant trunk groups can be
defined on a standalone switch or Juniper Virtual Chassis Fabric.
Topology In this example, three devices – one from Juniper and two Cisco – form a ring topology.
The devices are the Juniper Virtual Chassis Fabric (in turn comprised of two Juniper
QFX5100 and one Juniper EX4300 switches); a Cisco Nexus 7010; and a Cisco Catalyst
3850. At test time, RTG was not supported in the Juniper EX9200, so it was not included
in this test.
In this example, the inter-switch ports are VLAN trunk ports allowing traffic from VLAN
IDs 2001-2003. However, RTG works equally well with or without VLAN trunking.
Spirent TestCenter traffic generator/analyzers offer frames to access ports on each switch.
50
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Spanning tree, which is enabled by default on Juniper and Cisco switches, is disabled in
this example. Instead RTG configured on the Juniper Virtual Chassis Fabric sets up
primary and secondary traffic paths. When trunk links are configured as part of an RTG,
they cannot be part of a spanning tree topology.
Figure 12 shows the RTG test bed topology.
Figure 12: Redundant Trunk Group validation topology Initially, ports xe-1/0/2 on the Juniper Virtual Chassis Fabric is defined as the primary
path for the RTG. If a link failure occurs, the Virtual Chassis Fabric will use the other
trunk port on the same switch.
Juniper commands 1. Define three VLANs with VLAN IDs of 2001-2003:
[email protected]>
[email protected]#
[email protected]#
[email protected]#
configure
set vlans v2001 vlan-id 2001
set vlans v2002 vlan-id 2002
set vlans v2003 vlan-id 2003
2. Define interfaces xe-1/0/2.0 and xe-1/0/10.0 as VLAN trunk ports allowing traffic
from the three VLANs defined in the previous step:
51
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
[email protected]# set
[email protected]# set
interface-mode
[email protected]# set
members v2001
[email protected]# set
members v2002
[email protected]# set
members v2003
[email protected]# set
[email protected]# set
interface-mode
[email protected]# set
members v2001
[email protected]# set
members v2002
[email protected]# set
members v2003
interfaces xe-1/0/2 description "RTG to Nexus 7010 e3/3"
interfaces xe-1/0/2 unit 0 family ethernet-switching
trunk
interfaces xe-1/0/2 unit 0 family ethernet-switching vlan
interfaces xe-1/0/2 unit 0 family ethernet-switching vlan
interfaces xe-1/0/2 unit 0 family ethernet-switching vlan
interfaces xe-1/0/10 description "RTG to Cat3850 t1/1/4"
interfaces xe-1/0/10 unit 0 family ethernet-switching
trunk
interfaces xe-1/0/10 unit 0 family ethernet-switching vlan
interfaces xe-1/0/10 unit 0 family ethernet-switching vlan
interfaces xe-1/0/10 unit 0 family ethernet-switching vlan
Note that this example uses trunk ports. Access ports also can be members of redundant
trunk groups.
3. Disable rapid spanning tree, which is enabled by default:
[email protected]# delete protocols rstp
4. Define an RTG named “rtg0” and set interface xe-1/0/2.0 as the primary path and
interface xe-1/0/10.0 as the secondary path:
[email protected]# set switch-options redundant-trunk-group group rtg0 interface
xe-1/0/2.0 primary
[email protected]# set switch-options redundant-trunk-group group rtg0 interface
xe-1/0/10.0
[email protected]# commit
Cisco commands Cisco Nexus 7010:
1. NX-OS switch configuration does not require any RTG-specific commands. Simply
define VLANs; disable spanning tree on those VLANs; and assign switch ports to be
trunk-mode members of those VLANs:
Nexus7010# configure terminal
Nexus7010(config)# vlan 2001-2003
Nexus7010(config-vlan)# exit
Nexus7010# no spanning-tree vlan 2001-2003
Nexus7010# interface Ethernet3/3
Nexus7010(config-if)# description RTG to VCF xe-1/0/2
Nexus7010(config-if)# switchport
Nexus7010(config-if)# switchport mode trunk
Nexus7010(config-if)# switchport trunk allowed vlan 2001-2003
Nexus7010(config-if)# no shutdown
Nexus7010# interface Ethernet3/7
Nexus7010(config-if)# description RTG to Cat3850 t1/1/4
52
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
switchport
switchport mode trunk
switchport trunk allowed vlan 2001-2003
no shutdown
end
Note that Cisco Nexus 7000 switches require an explicit “no shutdown” command to
enable an interface.
Cisco Catalyst 3850:
1. Cisco Catalyst switches do not require any RTG-specific configuration. Simply define
VLANs; disable spanning tree on those VLANs; and assign switch ports to be trunkmode members of those VLANs:
Cat3850# configure terminal
Cat3850(config)# vlan 2001-2003
Cat3850(config-vlan)# exit
Cat3850(config)# no spanning-tree vlan 2001-2003
Cat3850(config)# interface TenGigabitEthernet1/1/3
Cat3850(config-if)# description RTG to Juniper VCF xe-1/0/10
Cat3850(config-if)# switchport
Cat3850(config-if)# switchport mode trunk
Cat3850(config-if)# switchport trunk allowed vlan 2001-2003
Cat3850(config)# interface TenGigabitEthernet1/1/4
Cat3850(config-if)# description RTG to Nexus 7010 e3/7
Cat3850(config-if)# switchport
Cat3850(config-if)# switchport mode trunk
Cat3850(config-if)# switchport trunk allowed vlan 2001-2003
Cat3850(config-if)# end
Validation The show redundant-trunk-group command indicates the current RTG state.
This example is from the Juniper Virtual Chassis Fabric:
[email protected]# run show redundant-trunk-group group rtg0
Interface
Flap
State
Bandwidth
Time of last flap
10 Gbps
Never
10 Gbps
2014-04-18 01:38:23
count
xe-1/0/2.0 Up/Pri/Act
0
xe-1/0/10.0 Up
6
(00:01:55 ago)
Note that interface xe-1/0/2.0 is the primary path. After offering traffic from Spirent
TestCenter, the packet counters for interfaces xe-1/0/2.0 and xe-1/0/10.0 will indicate that
the switch forwarded all traffic to interface xe-1/0/2.0, the primary path in the RTG.
To verify correct operation of RTG redundancy, disable the primary path:
53
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
54
[email protected]# set interfaces xe-1/0/2 disable
[email protected]# commit
Now the show redundant-trunk-group command will indicate the primary
interface is down while the secondary interface remains up:
[email protected]# show redundant-trunk-group
[email protected]# run show redundant-trunk-group group rtg0
Interface
Flap
State
Bandwidth
Time of last flap
10 Gbps
2014-04-18 01:47:25
(00:00:11 ago)
10 Gbps
2014-04-18 01:38:23
(00:09:13 ago)
count
xe-1/0/2.0 Dwn/Pri
1
xe-1/0/10.0 Up/Act
6
Note also that the command output indicates when the primary interface went down and
that its flap count has incremented by 1.
Again, after offering traffic from Spirent TestCenter, the packet counters for both
interfaces will indicate that the switch forwarded all traffic to interface xe-1/0/10.0, the
secondary path in the RTG.
Spanning tree case 1: Rapid spanning tree protocol (RSTP) Objective To verify interoperability of a rapid spanning tree topology between Juniper and Cisco
switches.
To measure convergence time of a rapid spanning tree topology between Juniper and
Cisco switches after link failure.
Background The spanning tree protocol is widely used in Ethernet networks for loop prevention and
redundancy. Rapid spanning tree, defined in IEEE 802.1w, provides much faster
convergence time after a link or device failure than the original 802.1D spanning tree
specification.
Topology This example uses redundant links between two Juniper switches and one Cisco switch.
Junos running on Juniper EX switches supports rapid spanning tree protocol (RSTP) by
default. Cisco switches also support spanning tree by default; although Cisco IOS defines
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
the spanning tree mode as that vendor’s proprietary “PVST Plus” mode, it is
interoperable with other vendors’ rapid spanning tree implementations.
Figure 13 illustrates the RSTP validation test bed. This example involves a standalone
Juniper EX4300 switch; a Juniper Virtual Chassis Fabric (comprised of two Juniper
QFX5100 and one Juniper EX4300 switches); and a Cisco Nexus 7010.
All inter-switch links use VLAN trunk ports that allow tagged traffic with a VLAN ID of
2001. Rapid spanning is enabled by default on the Juniper switches. VLAN trunking is
not required by spanning tree; it would work equally well with interfaces in access mode.
Cisco’s “PVST Plus,” enabled by default on the Nexus 7010, is interoperable with
standard rapid spanning tree. Traffic offered from the Spirent TestCenter traffic
generator/analyzer verifies the spanning tree topology.
55
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
56
Figure 13: RSTP validation topology Juniper commands Juniper Virtual Chassis Fabric:
1. Create a VLAN with an ID of 2001. Then assign interfaces to carry traffic from that
VLANs in trunking mode:
[email protected]> config
[email protected]# set vlans v2001 vlan-id 2001
[email protected]# set interfaces xe-1/0/2 description “RSTP to Nexus 7010 e3/3”
[email protected]# set interfaces xe-1/0/2 unit 0 family ethernet-switching
interface-mode trunk
[email protected]# set interfaces xe-1/0/2 unit 0 family ethernet-switching vlan
members v2001
[email protected]# set interfaces ge-0/0/12 description “RSTP to Juniper EX4300 ge0/0/12”
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
[email protected]# set
interface-mode
[email protected]# set
members v2001
[email protected]# set
v2002
interfaces ge-0/0/12 unit 0 family ethernet-switching
trunk
interfaces ge-0/0/12 unit 0 family ethernet-switching vlan
interfaces ge-0/0/12 unit 0 family ethernet-switching vlan
2. Enable rapid spanning tree. On a new switch configuration, this step should be
unnecessary since rapid spanning tree is enabled by default:
[email protected]# set protocols rstp
[email protected]# commit
Juniper EX4300:
1. Create a VLAN with an ID of 2001. Then assign interfaces to carry traffic from that
VLANs in trunking mode:
[email protected]> config
[email protected]# set vlans v2001 vlan-id 2001
[email protected]# set interfaces xe-0/2/0 description “RSTP to Nexus 7010 e3/1”
[email protected]# set interfaces xe-0/2/0 unit 0 family ethernet-switching
interface-mode trunk
[email protected]# set interfaces xe-0/2/0 unit 0 family ethernet-switching vlan
members v2001
[email protected]# set interfaces ge-0/0/12 description “RSTP to Juniper EX4300
ge-0/0/12”
[email protected]# set interfaces ge-0/0/12 unit 0 family ethernet-switching
interface-mode trunk
[email protected]# set interfaces ge-0/0/12 unit 0 family ethernet-switching
vlan members v2001
2. Enable rapid spanning tree. On a new switch configuration, this step should be
unnecessary since rapid spanning tree is enabled by default:
[email protected]# set protocols rstp
[email protected]# commit
Cisco commands 1. Create a VLAN with an ID of 2001. Then assign interfaces to carry traffic from that
VLANs in trunking mode:
Nexus7010# configure terminal
Nexus7010(config)# vlan 2001
Nexus7010(config-vlan)# exit
Nexus7010(config)# interface Ethernet3/1
Nexus7010(config-if)# description STP to EX4300 xe-0/2/0
Nexus7010(config-if)# switchport
Nexus7010(config-if)# switchport mode trunk
Nexus7010(config-if)# switchport trunk allowed vlan 2001
Nexus7010(config-if)# no shutdown
Nexus7010(config)# interface Ethernet3/3
Nexus7010(config-if)# description STP to VCF xe-1/0/2
Nexus7010(config-if)# switchport
57
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
switchport mode trunk
switchport trunk allowed vlan 2001
no shutdown
exit
Note that Cisco Nexus 7000 switches require the explicit “no shutdown” command to
enable interfaces. This command is not required in Cisco Catalyst switches.
2. Enable PVST Plus. On a new switch configuration, this step should be unnecessary
since PVST Plus is enabled by default:
Nexus7010(config)# spanning-tree mode rstp
Nexus7010(config)# end
On Cisco Catalyst switches, the equivalent command is “spanning-tree mode
rapid-pvst”.
Validation The command show spanning-tree bridge brief will display a summary of
spanning tree parameters:
[email protected]# run show spanning-tree bridge brief
STP bridge parameters
Routing instance name
: GLOBAL
Context ID
: 0
Enabled protocol
: RSTP
...
On the Cisco Nexus 7010, the equivalent command is “show spanning-tree
<vlan ID>”:
Nexus7010# show spanning-tree vlan 2001
VLAN2001
Spanning tree enabled protocol rstp
Root ID
Priority
34769
Address
0024.f718.9ec1
This bridge is the root
Hello Time 2 sec Max Age 20 sec
Bridge ID
Priority
Address
Hello Time
Interface
---------------Eth3/1
Eth3/3
Role
---Desg
Back
Sts
--FWD
BLK
Forward Delay 15 sec
34769 (priority 32768 sys-id-ext 2001)
0024.f718.9ec1
2 sec Max Age 20 sec Forward Delay 15 sec
Cost
--------2
2
Prio.Nbr
-------128.385
128.387
Type
-------------------------------P2p
P2p
Note that interface e3/1 is in forwarding state, and e3/3 is in blocking state.
58
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
To verify all switches send traffic only over the spanning tree interfaces in forwarding
state, generate a known quantity of frames from Spirent TestCenter or other source and
compare switch interface packet counters with those sent and received on each interface.
Interfaces in blocking state will receive spanning tree BPDU frames but should transmit
no frames.
To determine convergence time, disable one of the spanning tree interfaces in forwarding
state while offering a known quantity of frames from Spirent TestCenter or other traffic
generator. Convergence time can be derived from frame loss. For example, if Spirent
TestCenter generates traffic at a rate of 1,000 frames per second, each dropped frame is
equivalent to 1 millisecond of convergence time. If the switches drop 47 frames, then
rapid spanning tree convergence time is 47 ms.
Spanning tree case 2: Multiple spanning tree protocol (MSTP) Objective To verify interoperability of a multiple spanning tree topology between Juniper and Cisco
switches.
To measure convergence time of a multiple spanning tree topology between Juniper and
Cisco switches after link failure.
Background As defined in IEEE specification 802.1s, the multiple spanning tree protocol (MSTP)
adds loop prevention and redundancy on a per-VLAN basis. With MSTP, individual
spanning tree topologies can be configured for each VLAN.
Topology This example uses redundant links between two Juniper switches and one Cisco switch.
VLAN IDs of 2001 and 2002 have been defined on both the Juniper and Cisco switches,
and MSTP is enabled on all switches.
Figure 14 illustrates the MSTP validation test bed. This example involves a standalone
Juniper EX4300 switch; a Juniper Virtual Chassis Fabric (comprised of two Juniper
QFX5100 and one Juniper EX4300 switches); and a Cisco Nexus 7010.
As shown in the figure, the links interconnecting each switch are trunk ports that allow
tagged traffic from VLAN IDs 2001 and 2002. Two access-mode ports are configured on
each switch: One apiece for VLAN IDs 2001 and 2002. Traffic offered from the Spirent
TestCenter traffic generator/analyzer verifies the spanning tree topology in each VLAN.
59
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
60
Figure 14: MSTP validation topology Juniper commands On both Juniper devices:
1. Create VLANs “v2001” and “v2002”:
[email protected]> config
[email protected]# set vlans v2001 vlan-id 2001
[email protected]# set vlans v2002 vlan-id 2002
2. Configure access-mode interfaces for VLANs v2001 and v2002, respectively. The
Spirent TestCenter traffic generator/analyzer will attach to these ports.
On the Juniper Virtual Chassis Fabric:
[email protected]#
[email protected]#
v2001
[email protected]#
[email protected]#
v2002
set interfaces ge-2/0/0 description “v2001 to Spirent”
set interfaces ge-2/0/0.0 family ethernet-switching vlan members
set interfaces ge-2/0/2 description “v2002 to Spirent”
set interfaces ge-2/0/1.0 family ethernet-switching vlan members
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
On the Juniper EX4300:
[email protected]#
[email protected]#
members v2001
[email protected]#
[email protected]#
members v2002
61
set interfaces ge-0/0/24 description “v2001 to Spirent”
set interfaces ge-0/0/24.0 family ethernet-switching vlan
set interfaces ge-0/0/25 description “v2002 to Spirent”
set interfaces ge-0/0/25.0 family ethernet-switching vlan
3. Configure trunk ports that allow tagged traffic from VLANs v2001 and v2002.
On the Juniper Virtual Chassis Fabric:
[email protected]# set
[email protected]# set
interface-mode
[email protected]# set
members v2001
[email protected]# set
members v2002
[email protected]# set
0/0/12”
[email protected]# set
interface-mode
[email protected]# set
members v2001
[email protected]# set
v2002
interfaces xe-1/0/2 description “RSTP to Nexus 7010 e3/3”
interfaces xe-1/0/2 unit 0 family ethernet-switching
trunk
interfaces xe-1/0/2 unit 0 family ethernet-switching vlan
interfaces xe-1/0/2 unit 0 family ethernet-switching vlan
interfaces ge-0/0/12 description “RSTP to Juniper EX4300 geinterfaces ge-0/0/12 unit 0 family ethernet-switching
trunk
interfaces ge-0/0/12 unit 0 family ethernet-switching vlan
interfaces ge-0/0/12 unit 0 family ethernet-switching vlan
On the Juniper EX4300:
[email protected]# set interfaces
[email protected]# set interfaces
interface-mode trunk
[email protected]# set interfaces
members v2001
[email protected]# set interfaces
members v2002
[email protected]# set interfaces
ge-0/0/12”
[email protected]# set interfaces
interface-mode trunk
[email protected]# set interfaces
vlan members v2001
[email protected]# set interfaces
vlan members v2002
xe-0/2/0 description “RSTP to Nexus 7010 e3/1”
xe-0/2/0 unit 0 family ethernet-switching
xe-0/2/0 unit 0 family ethernet-switching vlan
xe-0/2/0 unit 0 family ethernet-switching vlan
ge-0/0/12 description “RSTP to Juniper EX4300
ge-0/0/12 unit 0 family ethernet-switching
ge-0/0/12 unit 0 family ethernet-switching
ge-0/0/12 unit 0 family ethernet-switching
4. Enable multiple spanning tree. This requires deleting rapid spanning tree (if enabled)
and then setting one multiple spanning tree instance (MSTI) per VLAN.
On the Juniper Virtual Chassis Fabric:
[email protected]#
[email protected]#
[email protected]#
[email protected]#
delete protocols rstp
set protocols mstp interface xe-1/0/2
set protocols mstp interface ge-0/0/12
set protocols mstp msti vlan v2001
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
[email protected]# set protocols mstp msti vlan v2002
[email protected]# commit
On the Juniper EX4300:
[email protected]#
[email protected]#
[email protected]#
[email protected]#
[email protected]#
[email protected]#
delete protocols rstp
set protocols mstp interface
set protocols mstp interface
set protocols mstp msti vlan
set protocols mstp msti vlan
commit
xe-0/2/0.0
ge-0/0/12.0
v2001
v2002
Cisco commands 1. Create VLANs 2001 and 2002:
Nexus7010# configure terminal
Nexus7010(config)# vlan 2001-2002
Nexus7010(config-vlan)# exit
2. Configure ports Ethernet3/13 and Ethernet3/14 as access-mode ports for VLANs 2001
and 2002, respectively:
Nexus7010(config)# interface Ethernet3/13
Nexus7010(config-if)# description MSTP to Spirent vlan 2001
Nexus7010(config-if)# switchport
Nexus7010(config-if)# switchport access vlan 2001
Nexus7010(config-if)# no shutdown
Nexus7010(config)# interface Ethernet3/14
Nexus7010(config-if)# description MSTP to Spirent vlan 2002
Nexus7010(config-if)# switchport
Nexus7010(config-if)# switchport access vlan 2002
Nexus7010(config-if)# no shutdown
Note that Cisco Nexus 7000 switches require the explicit “no shutdown” command to
enable interfaces. This command is not required in Cisco Catalyst switches.
3. Configure ports Ethernet3/1 and Ethernet3/3 as trunk ports that allow tagged traffic for
VLANs 2001 and 2002:
62
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
interface Ethernet3/1
description MSTP to Juniper VCF xe-1/0/2
switchport
switchport mode trunk
switchport trunk allowed vlan 2001-2002
no shutdown
interface Ethernet3/3
description MSTP to Juniper EX4300 xe-0/2/0
switchport
switchport mode trunk
switchport trunk allowed vlan 2001-2002
no shutdown
Note that Cisco Nexus 7000 switches require the explicit “no shutdown” command to
enable interfaces. This command is not required in Cisco Catalyst switches.
4. Enable multiple spanning tree. This requires defining multiple spanning tree as the
mode of operation and adding one multiple spanning tree instance (MSTI) per VLAN.
Nexus7010(config)# spanning-tree mode mst
Nexus7010(config)# spanning-tree mst configuration
Nexus7010(config-mst)# instance 1 vlan 2001
Nexus7010(config-mst)# instance 2 vlan 2002
Nexus7010(config-mst)# end
Validation The command show spanning-tree bridge brief will display a summary of
spanning tree parameters:
[email protected]> show spanning-tree bridge brief
STP bridge parameters
Context ID : 0
Enabled protocol : MSTI
...
The command show spanning-tree interface brief will display a
summary of spanning tree parameters on a per-interface and per-MSTI basis.
63
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
[email protected]> show spanning-tree interface brief
Interface
xe-1/0/2
ge-0/0/12
Port ID
Designated
port ID
128:387
128:490
128:491
128:616
Designated
bridge ID
32768.0024f7189ec1
32768.100e7ea6ffc1
Port
Cost
2000
20000
State
Role
FWD
BLK
ROOT
ALT
Port
Cost
2000
20000
State
Role
FWD
BLK
DESG
ALT
Port
Cost
2000
20000
State
Role
FWD
BLK
ROOT
ALT
Spanning tree interface parameters for instance 1
Interface
xe-1/0/40
ge-0/0/12
Port ID
Designated
port ID
128:493
128:490
128:493
128:616
Designated
bridge ID
32769.100e7eb06791
32769.100e7ea6ffc1
Spanning tree interface parameters for instance 2
Interface
xe-1/0/2
ge-0/0/12
Port ID
Designated
port ID
128:387
128:490
128:491
128:616
Designated
bridge ID
32770.0024f7189ec1
32770.100e7ea6ffc1
The equivalent command for the Cisco Nexus 7010 is “show spanning-tree
<vlan ID>”. In this example, the Nexus 7010 is the root bridge for VLAN 2001.
Nexus7010# show spanning-tree vlan 2001
MST0001
Spanning tree enabled protocol mstp
Root ID
Priority
32769
Address
0024.f718.9ec1
This bridge is the root
Hello Time 2 sec Max Age 20 sec
Bridge ID
Priority
Address
Hello Time
Interface
---------------Eth3/1
Eth3/3
Role
---Desg
Desg
Sts
--FWD
FWD
Forward Delay 15 sec
32769 (priority 32768 sys-id-ext 1)
0024.f718.9ec1
2 sec Max Age 20 sec Forward Delay 15 sec
Cost
--------2000
2000
Prio.Nbr
-------128.385
128.387
Type
-------------------------------P2p
P2p
To verify all switches send traffic only over the spanning tree interfaces in forwarding
state, generate a known quantity of frames from Spirent TestCenter or other source to
each VLAN and compare switch interface packet counters with those sent and received
on each interface. Interfaces in blocking state will receive spanning tree BPDU frames
but should transmit no frames.
To determine convergence time, disable one of the spanning tree interfaces in forwarding
state while offering a known quantity of frames from Spirent TestCenter or other traffic
generator. Convergence time can be derived from frame loss. For example, if Spirent
TestCenter generates traffic at a rate of 1,000 frames per second, each dropped frame is
equivalent to 1 millisecond of convergence time. If the switches drop 47 frames, then
rapid spanning tree convergence time is 47 ms.
64
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Spanning tree case 3: VLAN spanning tree protocol (VSTP) and Per-­‐VLAN Spanning Tree Plus (PVST+) Objective To verify interoperability of Juniper VLAN spanning tree protocol (VSTP) and Cisco
per-VLAN spanning tree protocol plus (PVST+) between Juniper and Cisco switches,
respectively.
To measure convergence time of a VSTP-PVST+ topology between Juniper and Cisco
switches after link failure.
Background With Juniper’s VLAN Spanning Tree Protocol (VSTP), Juniper switches can run one or
more spanning tree instances per VLAN. As stated in the Junos Software Guide, VSTP
“enables more intelligent tree spanning, because each VLAN can have interfaces enabled
or disabled depending on the paths available to that specific VLAN.”
The goal of this exercise is to demonstrate interoperability in a multiple-VLAN
environment using VSTP running on Juniper EX switches and PVST+ running on Cisco
Nexus or Catalyst switches.
Topology This example uses redundant links between two Juniper switches and one Cisco switch.
VLAN IDs of 2001 and 2002 have been defined on both the Juniper and Cisco switches.
VSTP is enabled on the Juniper switches, while the Cisco Nexus 7010 runs Rapid
PVST+.
Figure 15 illustrates the VSTP-PVST+ validation test bed. This example involves a
standalone Juniper EX4300 switch; a Juniper Virtual Chassis Fabric (comprised of two
Juniper QFX5100 and one Juniper EX4300 switches); and a Cisco Nexus 7010.
As shown in the figure, the links interconnecting each switch are trunk ports that allow
tagged traffic from VLAN IDs 2001 and 2002. Two access-mode ports are configured on
each switch: One apiece for VLAN IDs 2001 and 2002. Traffic offered from the Spirent
TestCenter traffic generator/analyzer verifies the spanning tree topology in each VLAN.
65
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
66
Figure 15: VSTP-­‐PVST+ validation topology Juniper commands On both Juniper devices:
1. Create VLANs “v2001” and “v2002”:
[email protected]> config
[email protected]# set vlans v2001 vlan-id 2001
[email protected]# set vlans v2002 vlan-id 2002
2. Configure access-mode interfaces for VLANs v2001 and v2002, respectively. The
Spirent TestCenter traffic generator/analyzer will attach to these ports.
On the Juniper Virtual Chassis Fabric:
[email protected]#
[email protected]#
v2001
[email protected]#
[email protected]#
v2002
set interfaces ge-2/0/0 description “v2001 to Spirent”
set interfaces ge-2/0/0.0 family ethernet-switching vlan members
set interfaces ge-2/0/2 description “v2002 to Spirent”
set interfaces ge-2/0/1.0 family ethernet-switching vlan members
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
On the Juniper EX4300:
[email protected]#
[email protected]#
members v2001
[email protected]#
[email protected]#
members v2002
67
set interfaces ge-0/0/24 description “v2001 to Spirent”
set interfaces ge-0/0/24.0 family ethernet-switching vlan
set interfaces ge-0/0/25 description “v2002 to Spirent”
set interfaces ge-0/0/25.0 family ethernet-switching vlan
3. Configure trunk ports that allow tagged traffic from VLANs v2001 and v2002.
On the Juniper Virtual Chassis Fabric:
[email protected]# set
[email protected]# set
interface-mode
[email protected]# set
members v2001
[email protected]# set
members v2002
[email protected]# set
0/0/12”
[email protected]# set
interface-mode
[email protected]# set
members v2001
[email protected]# set
v2002
interfaces xe-1/0/2 description “VSTP to Nexus 7010 e3/3”
interfaces xe-1/0/2 unit 0 family ethernet-switching
trunk
interfaces xe-1/0/2 unit 0 family ethernet-switching vlan
interfaces xe-1/0/2 unit 0 family ethernet-switching vlan
interfaces ge-0/0/12 description “VSTP to Juniper EX4300 geinterfaces ge-0/0/12 unit 0 family ethernet-switching
trunk
interfaces ge-0/0/12 unit 0 family ethernet-switching vlan
interfaces ge-0/0/12 unit 0 family ethernet-switching vlan
On the Juniper EX4300:
[email protected]# set interfaces
[email protected]# set interfaces
interface-mode trunk
[email protected]# set interfaces
members v2001
[email protected]# set interfaces
members v2002
[email protected]# set interfaces
ge-0/0/12”
[email protected]# set interfaces
interface-mode trunk
[email protected]# set interfaces
vlan members v2001
[email protected]# set interfaces
vlan members v2002
xe-0/2/0 description “VSTP to Nexus 7010 e3/1”
xe-0/2/0 unit 0 family ethernet-switching
xe-0/2/0 unit 0 family ethernet-switching vlan
xe-0/2/0 unit 0 family ethernet-switching vlan
ge-0/0/12 description “VSTP to Juniper EX4300
ge-0/0/12 unit 0 family ethernet-switching
ge-0/0/12 unit 0 family ethernet-switching
ge-0/0/12 unit 0 family ethernet-switching
4. Enable VLAN spanning tree protocol. This requires deleting rapid spanning tree or
multiple spanning tree (if enabled) and then enabling VSTP for each VLAN.
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
On both Juniper devices:
[email protected]#
[email protected]#
[email protected]#
[email protected]#
[email protected]#
68
delete protocols rstp
delete protocols mstp
set protocols vstp vlan v2001
set protocols vstp vlan v2002
commit
Cisco commands 1. Create VLANs 2001 and 2002:
Nexus7010# configure terminal
Nexus7010(config)# vlan 2001-2002
Nexus7010(config-vlan)# exit
2. Configure ports Ethernet3/13 and Ethernet3/14 as access-mode ports for VLANs 2001
and 2002, respectively:
Nexus7010(config)# interface Ethernet3/13
Nexus7010(config-if)# description RPVST+ to Spirent vlan 2001
Nexus7010(config-if)# switchport
Nexus7010(config-if)# switchport access vlan 2001
Nexus7010(config-if)# no shutdown
Nexus7010(config)# interface Ethernet3/14
Nexus7010(config-if)# description RPVST+ to Spirent vlan 2002
Nexus7010(config-if)# switchport
Nexus7010(config-if)# switchport access vlan 2002
Nexus7010(config-if)# no shutdown
Note that Cisco Nexus 7000 switches require the explicit “no shutdown” command to
enable interfaces. This command is not required in Cisco Catalyst switches.
3. Configure ports Ethernet3/1 and Ethernet3/3 as trunk ports that allow tagged traffic for
VLANs 2001 and 2002:
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
Nexus7010(config-if)#
interface Ethernet3/1
description RPVST+ to Juniper
switchport
switchport mode trunk
switchport trunk allowed vlan
no shutdown
interface Ethernet3/3
description RPVST+ to Juniper
switchport
switchport mode trunk
switchport trunk allowed vlan
no shutdown
VCF xe-1/0/2
2001-2002
EX4300 xe-0/2/0
2001-2002
Note that Cisco Nexus 7000 switches require the explicit “no shutdown” command to
enable interfaces. This command is not required in Cisco Catalyst switches.
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
2. Enable PVST Plus. On a new switch configuration, this step should be unnecessary
since PVST Plus is enabled by default.
Nexus7010(config)# spanning-tree mode rstp
Nexus7010(config)# end
On Cisco Catalyst switches, the equivalent command is “spanning-tree mode
rapid-pvst”.
Validation The command show spanning-tree bridge brief will display a summary of
spanning tree parameters:
[email protected]> show spanning-tree bridge brief
STP bridge parameters
Context ID : 0
Enabled protocol : VSTP
...
The command show spanning-tree interface brief will display a
summary of spanning tree parameters on a per-interface and per-VLAN basis:
[email protected]> show spanning-tree interface brief
Spanning tree interface parameters for VLAN 2001
Interface
ge-0/0/12
xe-1/0/2
Port ID
128:4234
128:4236
Designated
port ID
128:490
128:387
Designated
bridge ID
34769.100e7ea6ffc0
34769.0024f7189ec1
Port
Cost
20000
2000
State
Role
BLK
FWD
ALT
ROOT
Port
Cost
20000
2000
State
Role
BLK
FWD
ALT
ROOT
Spanning tree interface parameters for VLAN 2002
Interface
ge-0/0/12
xe-1/0/2
Port ID
128:4234
128:4236
Designated
port ID
128:490
128:387
Designated
bridge ID
34770.100e7ea6ffc0
34770.0024f7189ec1
The equivalent command for the Cisco Nexus 7010 is “show spanning-tree
<vlan ID>”. In this example, the Nexus 7010 is the root bridge for VLAN 2001:
69
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Nexus7010# show spanning-tree vlan 2001
LAN2001
Spanning tree enabled protocol rstp
Root ID
Priority
34769
Address
0024.f718.9ec1
This bridge is the root
Hello Time 2 sec Max Age 20 sec
Bridge ID
Priority
Address
Hello Time
Interface
---------------Eth3/1
Eth3/3
Role
---Desg
Desg
Sts
--FWD
FWD
70
Forward Delay 15 sec
34769 (priority 32768 sys-id-ext 2001)
0024.f718.9ec1
2 sec Max Age 20 sec Forward Delay 15 sec
Cost
--------2
2
Prio.Nbr
-------128.385
128.387
Type
-------------------------------P2p
P2p
To verify all switches send traffic only over the spanning tree interfaces in forwarding
state, generate a known quantity of frames from Spirent TestCenter or other source to
each VLAN and compare switch interface packet counters with those sent and received
on each interface. Interfaces in blocking state will receive spanning tree BPDU frames
but should transmit no frames.
To determine convergence time, disable one of the spanning tree interfaces in forwarding
state while offering a known quantity of frames from Spirent TestCenter or other traffic
generator. Convergence time can be derived from frame loss. For example, if Spirent
TestCenter generates traffic at a rate of 1,000 frames per second, each dropped frame is
equivalent to 1 millisecond of convergence time. If the switches drop 47 frames, then
rapid spanning tree convergence time is 47 ms.
Virtual LAN (VLAN) trunking Objective To verify interoperability of IEEE 802.1Q VLAN trunking between Juniper and Cisco
switches using tagged traffic.
To verify interoperability of IEEE 802.1Q VLAN trunking between Juniper and Cisco
switches using untagged (native) traffic.
Background The IEEE 802.1Q specification defines a method for defining virtual broadcast domains.
A 4-byte VLAN header, usually called a “tag,” allows definition of broadcast domains
that may differ from physical switch topology. Without VLANs, all switch ports are
members of the same broadcast domain; with VLAN tagging, a network manager can set
up multiple broadcast domains across switches, and restrict broadcasts for different
VLANs on different ports.
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Topology This configuration example will validate VLAN trunking interoperability between
Juniper and Cisco switches in three ways:
•
•
•
The switches will forward allowed tagged traffic from multiple VLANs across a
trunk port
The switches will forward allowed untagged traffic from a native VLAN across a
trunk port
The switches will not forward disallowed tagged traffic across a trunk port
The final example above is a negative test to verify that switches will forward only traffic
explicitly permitted by VLAN trunking configurations.
Figure 16 below illustrates the test bed used to verify VLAN trunking operation. In this
example, a VLAN trunk carries allowed VLAN traffic between a Juniper Virtual Chassis
(comprising two Juniper EX9208 switches) and a Cisco Nexus 7010. Both switches use
10-gigabit Ethernet interfaces for the trunk port in this example, though VLAN trunking
also would work on any matched pair of Ethernet interfaces. The trunk ports on each
switch will allow tagged traffic with VLAN IDs 2001 and 2002, and untagged (“native”)
traffic from ports with a VLAN ID of 2003. A fourth VLAN, with a ID of 2004, also
exists, but the trunk port is configured not to allow that traffic.
Figure 16: VLAN trunking validation topology 71
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Juniper commands 1. Define VLANs v2001 through v2004 with VLAN IDs of 2001 through 2004
respectively:
[email protected]>
[email protected]#
[email protected]#
[email protected]#
[email protected]#
[email protected]#
config
set vlans
set vlans
set vlans
set vlans
set vlans
v2001
v2002
v2003
v2004
v2005
vlan-id
vlan-id
vlan-id
vlan-id
vlan-id
2001
2002
2003
2004
2005
2. Define a VLAN trunk port that allows tagged traffic from VLANs v2001 and v2002
and untagged traffic from VLAN v2003:
[email protected]# set interfaces
e3/9"
[email protected]# set interfaces
members v2001
[email protected]# set interfaces
interface-mode trunk
[email protected]# set interfaces
members v2001
[email protected]# set interfaces
members v2002
[email protected]# set interfaces
members v2003
xe-5/0/5 description "VLAN trunk to Nexus 7010
xe-5/0/5 unit 0 family ethernet-switching vlan
xe-5/0/5 unit 0 family ethernet-switching
xe-5/0/5 unit 0 family ethernet-switching vlan
xe-5/0/5 unit 0 family ethernet-switching vlan
xe-5/0/5 unit 0 family ethernet-switching vlan
3. On the VLAN trunk port, allow native untagged traffic from VLAN v2003. Note that
the native-vlan-id command takes the VLAN ID and not the VLAN name as an
argument:
[email protected]# set interfaces xe-5/0/5 native-vlan-id 2003
4. Define access-mode interfaces allowing untagged traffic from VLANs v2001 through
v2004:
[email protected]# set interfaces
[email protected]# set interfaces
interface-mode access
[email protected]# set interfaces
members v2001
[email protected]# set interfaces
[email protected]# set interfaces
interface-mode access
[email protected]# set interfaces
members v2002
[email protected]# set interfaces
[email protected]# set interfaces
interface-mode access
[email protected]# set interfaces
members v2003
[email protected]# set interfaces
[email protected]# set interfaces
interface-mode access
ge-2/0/0 description "to stc v2001"
ge-2/0/0 unit 0 family ethernet-switching
ge-2/0/0 unit 0 family ethernet-switching vlan
ge-2/0/1 description "to stc v2002"
ge-2/0/1 unit 0 family ethernet-switching
ge-2/0/1 unit 0 family ethernet-switching vlan
ge-2/0/2 description "to stc v2003"
ge-2/0/2 unit 0 family ethernet-switching
ge-2/0/2 unit 0 family ethernet-switching vlan
ge-2/0/3 description "to stc v2004"
ge-2/0/3 unit 0 family ethernet-switching
72
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
[email protected]# set interfaces ge-2/0/3 unit 0 family ethernet-switching vlan
members v2004
[email protected]# commit
Cisco commands 1. Define VLANs 2001 through 2004:
Nexus7010# configure terminal
Nexus7010(config)# vlan 2001-2004
Nexus7010(config-vlan)# exit
2. Define a VLAN trunk port that allows tagged traffic from VLANs 2001 and 2002 and
native untagged traffic from VLAN 2003:
Nexus7010(config)# interface Ethernet3/9
Nexus7010(config-if)# description VLAN trunk to EX9208 xe-5/0/5
Nexus7010(config-if)# switchport
Nexus7010(config-if)# switchport mode trunk
Nexus7010(config-if)# switchport trunk native vlan 2003
Nexus7010(config-if)# switchport trunk allowed vlan 2001-2003
Nexus7010(config-if)# no shutdown
Nexus7010(config-if)# exit
Note that Cisco Nexus 7000 switches require the explicit “no shutdown” command to
enable interfaces. This command is not required in Cisco Catalyst switches.
3. Define access-mode interfaces allowing untagged traffic from VLANs 2001 through
2003:
Nexus7010(config)# interface Ethernet3/13
Nexus7010(config-if)# description to stc vlan 2001
Nexus7010(config-if)# switchport
Nexus7010(config-if)# switchport access vlan 2001
Nexus7010(config-if)# no shutdown
Nexus7010(config)# interface Ethernet3/14
Nexus7010(config-if)# description to stc vlan 2002
Nexus7010(config-if)# switchport
Nexus7010(config-if)# switchport access vlan 2002
Nexus7010(config-if)# no shutdown
Nexus7010(config)# interface Ethernet3/15
Nexus7010(config-if)# description to stc vlan 2003
Nexus7010(config-if)# switchport
Nexus7010(config-if)# switchport access vlan 2003
Nexus7010(config-if)# no shutdown
Nexus7010(config)# interface Ethernet3/16
Nexus7010(config-if)# description to stc vlan 2004
Nexus7010(config-if)# switchport
Nexus7010(config-if)# switchport access vlan 2004
Nexus7010(config-if)# no shutdown
Nexus7010(config-if)# end
Cisco Catalyst switches also will require the “switchport mode access”
command on each access-mode interface.
73
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Note that Cisco Nexus 7000 switches require the explicit “no shutdown” command to
enable interfaces. This command is not required in Cisco Catalyst switches.
Validation The Spirent TestCenter traffic generator/analyzer can be configured to offer bidirectional
traffic between pairs of access-mode interfaces on each switch. In all cases – involving
unicast, broadcast or multicast traffic – traffic will stay local to the VLAN in which it is
defined. For example, traffic offered to VLAN v2001 on the Juniper switch will be
forwarded only to interfaces in VLAN 2001 on the Cisco switch and vice-versa.
If desired, port mirroring can be enabled on either switch to verify that the trunk ports
carry tagged traffic for VLAN IDs 2001 and 2002 and untagged traffic for VLAN ID
2003.
As a final verification that VLANs limit broadcast domains, Spirent TestCenter can be
configured to offer traffic to the access ports with an VLAN ID of 2004. The trunk ports
on both switches will not forward this traffic.
Virtual Router Redundancy Protocol (VRRP) interoperability Objective To validate failover functionality of the virtual router redundancy protocol (VRRP)
between Juniper and Cisco switches configured as routers.
Background As described in RFC 5798, two or more routers can make use of VRRP to enhance
network availability. With VRRP, multiple routers share a single virtual IP address. One
router acts as the master (active) device, while all others act as backups. If the master
router fails (or if a link fails on the interface configured with the virtual IP address), one
of the backup routers takes over as master.
Topology In this example, Juniper and Cisco switches are both configured to route IP traffic. The
interfaces connecting the switches each have unique IP addresses configured –
192.18.38.1/24 for the Juniper Virtual Chassis and 192.18.38.2/24 for the Cisco Nexus
7010. The Juniper and Cisco devices also share a single virtual IP address of
192.18.38.254/24, with the Juniper device initially acting as VRRP master.
The PCs attached to the Juniper and Cisco devices each use the virtual IP address of
192.18.38.254/24 as their default gateway. In the event of a failure of the master (Juniper)
device, the backup (Cisco) device will take over as master.
74
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Figure 17 illustrates the VRRP validation test bed. In this example, both the Juniper and
Cisco devices advertise the virtual IP address of 192.18.38.254. On both the Juniper and
Cisco devices, the IP address is assigned to the physical interface rather than a VLAN
interface. However, VRRP would work equally well with IP addresses assigned to VLAN
interfaces.
Figure 17: VRRP validation topology Juniper commands Configure an IP address on interface xe-5/0/5, and also define vrrp-group 1 on that
interface:
[email protected]> config
[email protected]# set interfaces xe-5/0/5 description "VRRP link to Nexus 7010
e3/10"
[email protected]# set interfaces xe-5/0/5 unit 0 family inet address 192.18.38.1/24
vrrp-group 1 virtual-address 192.18.38.254
[email protected]# set interfaces xe-5/0/5 unit 0 family inet address 192.18.38.1/24
vrrp-group 1 priority 254
[email protected]# set interfaces xe-5/0/5 unit 0 family inet address 192.18.38.1/24
vrrp-group 1 preempt
75
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
[email protected]# set interfaces xe-5/0/5 unit 0 family inet address 192.18.38.1/24
vrrp-group 1 accept-data
[email protected]# commit
The optional priority 255 statement forces the Juniper switch’s virtual IP address to
become the master VRRP instance, assuming the Cisco Nexus 7010 device uses a lower
priority value. The legal range of VRRP priorities is 1 though 255, with 255 being
highest.
Cisco commands 1. Ensure the VRRP feature is installed. This step is not required with Cisco Catalyst
switches:
Nexus7010# configure terminal
Nexus7010(config)# feature vrrp
2. Define interface and VRRP virtual IP addresses on interface Ethernet3/10:
Nexus7010(config)# interface Ethernet3/10
Nexus7010(config-if)# description VRRP to EX9208 xe-5/0/5
Nexus7010(config-if)# no switchport
Nexus7010(config-if)# ip address 192.18.38.2/24
Nexus7010(config-if)# vrrp 1
Nexus7010(config-if-vrrp)# address 192.18.38.254
Nexus7010(config-if-vrrp)# no shutdown
Nexus7010(config-if-vrrp)# exit
Nexus7010(config-if)# no shutdown
Nexus7010(config-if)# end
Note the explicit “no shutdown” commands for both physical and virtual interfaces.
These commands are mandatory with Cisco Nexus 7000 switches, but are not required
with Cisco Catalyst switches.
Validation On Juniper devices, the show vrrp summary command will indicate the current
VRRP state on each system. In the following examples, the Juniper Virtual Chassis acts
as VRRP master and the Cisco Nexus 7010 acts as a backup.
On the Juniper Virtual Chassis:
[email protected]# run show vrrp summary
Interface
State
Group
VR state
xe-5/0/5.0
up
1
master
192.18.38.1
VR Mode
Active
Type
lcl
vip
192.18.38.254
The equivalent command on Cisco devices is show vrrp summary:
Address
76
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Nexus7010# show vrrp
Interface VR IpVersion Pri
Time Pre State
VR IP addr
--------------------------------------------------------------Ethernet3/10
1
IPV4
100
1 s Y Backup 192.18.38.254
77
Note that both devices agree the master router is 192.18.38.1 (on the Juniper Virtual
Chassis), and both use a virtual IP address of 192.18.38.254.
If a router or link fails, the backup router should take over as the master. In this example,
the Cisco Nexus 7010 is promoted to master statement by reducing the Juniper Virtual
Chassis’ priority to 10. Since the Cisco Nexus 7010 uses a priority of 250 by default, it
will take over as master once this Juniper Virtual Chassis configuration change is
committed:
[email protected]> configure
[email protected]# set interfaces xe-5/0/5 family inet address 192.18.38.1/24
vrrp-group 1 virtual-address 192.18.38.254 priority 10
[email protected]# commit
After this change, the Juniper device becomes the backup router, and the master router is
now 192.18.38.254 on the Cisco device:
[email protected]# run show vrrp summary
Interface
State
Group
VR state
xe-5/0/5.0
up
1
backup
192.18.38.1
VR Mode
Active
Type
lcl
Address
vip
192.18.38.254
The Cisco device agrees that it is now the VRRP master:
Nexus7010# show vrrp summary
VRRP Summary
-----------Total Number of Groups Configured: 1
Init : 0
Backup : 0
Master : 1
Wi-­‐Fi passthrough Objective To verify the ability of a Juniper switch to forward Wi-Fi management traffic between a
Cisco Wi-Fi controller and Cisco Wi-Fi access points.
Background Many enterprises provision wireless LAN access using one or more centrally managed
controllers that monitor and manage RF and data networking parameters in real time. In
this model, Wi-Fi access points (APs) distributed throughout the enterprise depend on the
controllers for their configurations. Thus, reliable connectivity between controllers and
access points is essential.
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Juniper switches can carry Cisco and other vendors’ Wi-Fi traffic transparently, with no
special configuration required.
Topology As shown in Figure 18, a centrally located Cisco 5508 Wi-Fi controller in a data center
must communicate with Cisco 3602 and Cisco 3702 APs in a campus network. The Cisco
controller connects with a Juniper EX9208 core switch. The Cisco APs connect with a
Juniper EX4300 access switch using the Power Over Ethernet Plus (PoE+) specification.
The Cisco controller and APs both use the default VLAN, which here uses a VLAN
ID of 1. Trunk ports between the Juniper switches allow traffic from all VLANs.
Figure 18: Wi-­‐Fi passthrough validation topology Juniper commands No special configuration is needed on the Juniper switches. In this example, both access
and trunk ports are members of the default VLAN, and the trunk port between switches
allows traffic from all VLANs.
On the Juniper EX9208 switch:
1. Create a VLAN called default and assign a VLAN ID of 1:
[email protected]> configure
[email protected]# set vlans default vlan-id 1
78
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
2. (Optional) Create a Layer-3 interface for the default VLAN and assign an IP
address to that interface. This step is not mandatory for Wi-Fi passthrough operation:
[email protected]# set vlans default l3-interface irb.1
[email protected]# set interfaces irb unit 1 family inet address 10.0.1.1/24
3. Configure the inter-switch link as a VLAN trunk and allow traffic from all VLANs:
[email protected]# set interfaces xe-5/2/1 description "trunk to EX4300 xe0/2/0"
[email protected]# set interfaces xe-5/2/1 unit 0 family ethernet-switching
interface-mode trunk
[email protected]# set interfaces xe-5/2/1 unit 0 family ethernet-switching vlan
members all
This example uses the Junos keyword all to permit all VLAN traffic across the trunk. If
desired, forwarding can be restricted to specific VLANs. For example, this command
would permit traffic only from the default VLAN:
[email protected]# set interfaces xe-5/2/1 unit 0 family ethernet-switching vlan
members default
4. Configure two interfaces to allow traffic from the Cisco controller, both as accessmode members of the default VLAN:
[email protected]# set interfaces
controller"
[email protected]# set interfaces
interface-mode access
[email protected]# set interfaces
members default
[email protected]# set interfaces
controller"
[email protected]# set interfaces
interface-mode access
[email protected]# set interfaces
members default
[email protected]# commit
ge-2/0/0 description "to Cisco 5508 Wi-Fi
ge-2/0/0 unit 0 family ethernet-switching
ge-2/0/0 unit 0 family ethernet-switching vlan
ge-2/0/1 description "to Cisco 5508 Wi-Fi
ge-2/0/1 unit 0 family ethernet-switching
ge-2/0/1 unit 0 family ethernet-switching vlan
On the Juniper EX4300 switch:
1. Create a VLAN called default and assign a VLAN ID of 1:
[email protected]> configure
[email protected]# set vlans default vlan-id 1
2. (Optional) Create a Layer-3 interface for the default VLAN and assign an IP
address to that interface. This step is not mandatory for Wi-Fi passthrough operation:
[email protected]# set vlans default l3-interface irb.1
[email protected]# set interfaces irb unit 1 family inet address 10.0.1.2/24
3. Configure the inter-switch link as a VLAN trunk and allow traffic from all VLANs:
79
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
[email protected]# set interfaces xe-0/2/0 description "trunk to EX9208 xe5/2/1"
[email protected]# set interfaces xe-0/2/0 unit 0 family ethernet-switching
interface-mode trunk
[email protected]# set interfaces xe-0/2/0 unit 0 family ethernet-switching vlan
members all
This example uses the Junos keyword all to permit all VLAN traffic across the trunk. If
desired, forwarding can be restricted to specific VLANs. For example, this command
would permit traffic only from the default VLAN:
[email protected]# set interfaces xe-0/2/0 unit 0 family ethernet-switching vlan
members default
4. Configure two interfaces to allow traffic from the Cisco APs, both as access-mode
members of the default VLAN:
[email protected]# set interfaces
[email protected]# set interfaces
interface-mode access
[email protected]# set interfaces
vlan members default
[email protected]# set interfaces
[email protected]# set interfaces
interface-mode access
[email protected]# set interfaces
vlan members default
ge-0/0/12 description "to Cisco AP 3602"
ge-0/0/12 unit 0 family ethernet-switching
ge-0/0/12 unit 0 family ethernet-switching
ge-0/0/13 description "to Cisco AP3702"
ge-0/0/13 unit 0 family ethernet-switching
ge-0/0/13 unit 0 family ethernet-switching
5. Enable PoE+ on the interfaces to which the Cisco APs are attached:
[email protected]# set poe interface ge-0/0/12
[email protected]# set poe interface ge-0/0/13
[email protected]# commit
If desired, PoE+ can be enabled on all interfaces using Junos’ all keyword:
[email protected]# set poe interface all
[email protected]# commit
Cisco commands Configuration of the Cisco Wi-Fi controllers involves numerous RF as well as data
networking parameters, and is beyond the scope of this document. For more information,
consult the Cisco Wireless LAN Configuration Guide.
Validation On the Cisco Wi-Fi controller attached to the Juniper EX9208, the command “show ap
summary” will display a list of all APs managed by the controller.
80
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
81
(Cisco Controller) >show ap summary
Number of APs.................................... 2
Global AP User Name.............................. Not Configured
Global AP Dot1x User Name........................ Not Configured
AP Name
Slots AP Model
Ethernet MAC
Location
Country IP Address
Clients
------------------ ----- -------------------- ------------------------- ------- --------------- ------AP1
2
AIR-CAP3602I-A-K9
c4:64:13:c0:7c:da
location US
10.0.1.53
3
APb838.61a6.8530
2
AIR-CAP3702I-A-K9
b8:38:61:a6:85:30
location US
10.0.1.59
5
------default
default
Again on the Cisco controller, the command “show client summary” will display
the total number of clients associated with the APs.
(Cisco Controller) show client summary
Number of Clients................................ 7
Number of PMIPV6 Clients......................... 0
GLAN/
RLAN/
WLAN Auth Protocol
MAC Address
AP Name
Slot Status
Port Wired PMIPV6 Role
----------------- ----------------- ---- ------------- ----- ---- --------------- ---- ----- ------ ---------------24:77:03:80:39:4c APb838.61a6.8530
0
Associated
1
No
802.11n(2.4 GHz) 1
No
No
Unassociated
24:77:03:80:4f:50 APb838.61a6.8530
0
Associated
1
No
802.11n(2.4 GHz) 1
No
No
Unassociated
7c:d1:c3:8b:96:a8 AP1
0
Associated
1
Yes
802.11n(2.4 GHz) 1
No
No
Local
7c:d1:c3:8b:b0:66 AP1
0
Associated
1
Yes
802.11n(2.4 GHz) 1
No
No
Local
7c:d1:c3:8c:ef:92 APb838.61a6.8530
0
Associated
1
Yes
802.11n(2.4 GHz) 1
No
No
Local
7c:d1:c3:8d:95:20 APb838.61a6.8530
0
Associated
1
Yes
802.11n(2.4 GHz) 1
No
No
Local
7c:d1:c3:8d:de:c8 AP1
0
Associated
1
Yes
802.11n(2.4 GHz) 1
No
No
Local
The command-line interface of the APs can verify correct PoE+ operation, including the
ability of the Juniper EX4300 switch to supply power greater than the 15.4-watt limit of
standard PoE.
APb838.61a6.8530>show controllers dot11Radio 0 powerreg
### Tx Power: Dot11 Config
Serving Channel
Active Level Index
Active Level
IEEE MIB format
11 (Type 0)
1 (Unitless)
23 dBm (OFDM 23 dBm)
TRUE
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Metric unit
HD Active
Low Power Mode
Cookie max
Client Power
Regulatory Limit
dBm
127
Set High (Active: Set High)
17
Max 23, Active Max 23 (Symmetric)
30
...
Note that the PoE+-capable AP draws 23 watts of power, well above the 15.4-watt limit
of standard PoE.
82
Juniper Networks/Cisco Systems Switch Interoperability Cookbook
Appendix A: Sample Configuration Files This appendix lists URLs for the Juniper and Cisco switch configuration files used to verify
interoperability. These files are freely available for download from a public Network Test
server.
A copy of this document, a brief interoperability report and all Juniper and Cisco configuration
files are available at http://networktest.com/jnpriop14.
Appendix B: Software Versions Tested This appendix lists the software versions used for all test bed devices.
Juniper EX9208 (in Virtual Chassis and standalone configurations): Junos 13.3R1.6
Juniper QFX5100 (in Virtual Chassis Fabric): Junos 13.2-20140409_x_132_x51
Juniper EX4300 (in standalone configuration): Junos 13.2X51-D15.5
Juniper MX80: Junos 13.3R1.6
Cisco Catalyst 3850: IOS-XE 03.02.01.SE
Cisco Nexus 7010: NX-OS 6.1(4a)
Cisco Catalyst 6509: IOS 12.2(33)SXH1
Cisco 5500 Wi-Fi controller: 7.6.110.0
Spirent TestCenter: 4.33.0086
Appendix C: Disclaimer Network Test Inc. has made every attempt to ensure that all test procedures were conducted
with the utmost precision and accuracy, but acknowledges that errors do occur. Network Test
Inc. shall not be held liable for damages which may result for the use of information contained
in this document.
All trademarks mentioned in this document are property of their respective owners.
83
Was this manual useful for you? yes no
Thank you for your participation!

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

Download PDF

advertisement