User guide | Alcatel-Lucent 4504 Switch User Manual

Release Notes
OmniSwitch 6800/6850/9000
Release 6.1.3.R01
These release notes accompany release 6.1.3.R01 software for the OmniSwitch 6800, 6850, and 9000
hardware. They provide important information on individual software features and hardware modules.
Since much of the information in these release notes is not included in the hardware and software user
manuals, it is important that you read all sections of this document before installing new hardware or loading new software.
Contents
• Related Documentation, see page 2.
• System Requirements, see page 4.
•Memory Requirements, see page 4.
•UBoot, FPGA, MiniBoot, BootROM, and Upgrade Requirements, see page 4.
• New Hardware Supported, see page 5.
• Supported Hardware/Software Combinations, see page -9.
• New Software Features, see page -11.
• Software Supported, see page 13.
• Supported Traps, see page 37.
• Unsupported Software Features, see page 42.
• Unsupported CLI Commands, see page 42.
• Unsupported MIBs, see page 44.
• Open Problem Reports, and Feature Exceptions, see page 48.
•Switch Management, see page 48.
•Layer 2, see page 53.
•Layer 3, see page 65.
•Advanced Routing, see page 69.
•Quality of Service, see page 72.
•Security, see page 74.
•System, see page 78.
• Technical Support, see page 86.
Release Notes
Part Number 032067-10, Rev D
Copyright 2006 Alcatel Internetworking, Inc.
page 1
Related Documentation
Related Documentation
These release notes should be used in conjunction with the OmniSwitch 6800, 6850, and 9000. The
following are the titles and descriptions of the user manuals that apply to the OmniSwitch 6800, 6850, and
9000.
Note. User manuals can be downloaded at http://www.alcatel.com/enterprise/en/resource_library/
user_manuals.html.
• OmniSwitch 6800 Series Getting Started guide
Describes the hardware and software procedures for getting an OmniSwitch 6800 Series switch up and
running.
• OmniSwitch 6850 Series Getting Started Guide
Describes the hardware and software procedures for getting an OmniSwitch 6850 Series switch up and
running.
• OmniSwitch 9000 Series Getting Started Guide
Describes the hardware and software procedures for getting an OmniSwitch 9000 Series switch up and
running.
• OmniSwitch 6800 Series Hardware User Guide
Complete technical specifications and procedures for all OmniSwitch 6800 Series chassis, power
supplies, and fans.
• OmniSwitch 6850 Series Hardware User Guide
Complete technical specifications and procedures for all OmniSwitch 6850 Series chassis, power
supplies, and fans.
• OmniSwitch 9000 Series Hardware User Guide
Complete technical specifications and procedures for all OmniSwitch 9000 Series chassis, power
supplies, and fans.
• OmniSwitch CLI Reference Guide
Complete reference to all CLI commands supported on the OmniSwitch. Includes syntax definitions,
default values, examples, usage guidelines, and CLI-to-MIB variable mappings.
• OmniSwitch 6800/6850/9000 Network Configuration Guide
Includes network configuration procedures and descriptive information on all the major software
features and protocols included in the base software package. Chapters cover Layer 2 information
(Ethernet and VLAN configuration), Layer 3 information (routing protocols), security options (Authenticated Switch Access (ASA)), Quality of Service (QoS), link aggregation.
• OmniSwitch 6800/6850/9000 Series Switch Management Guide
Includes procedures for readying an individual switch for integration into a network. Topics include the
software directory architecture, software rollback protections, authenticated switch access, managing
switch files, system configuration, using SNMP, and using web management software (WebView).
page 2
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Related Documentation
• OmniSwitch 6800/6850/9000 Series Advanced Routing Configuration Guide
Includes network configuration procedures and descriptive information on all the software features and
protocols included in the advanced routing software package. Chapters cover multicast routing
(DVMRP and PIM), BGP, OSPF, and OSPFv3.
• Upgrade Instructions for 6.1.3.R01
Provides instructions for upgrading the OmniSwitch 6800, 6850, 9000 to 6.1.3.R01.
• Technical Tips, Field Notices
Contracted customers can visit our customer service website at: http://eservice.ind.alcatel.com.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 3
System Requirements
System Requirements
Memory Requirements
• OmniSwitch 6800 Series Release 6.1.3.R01 requires 256 MB of SDRAM and 64MB of flash memory.
This is the standard configuration shipped.
• OmniSwitch 6850 Series Release 6.1.3.R01 requires 256 MB of SDRAM and 64MB of flash memory.
This is the standard configuration shipped.
• OmniSwitch 9000 Series Release 6.1.3.R01 requires 256 MB of SDRAM and 128MB of flash
memory. This is the standard configuration shipped.
Configuration files and the compressed software images—including web management software
(WebView) images—are stored in the flash memory. Use the show hardware info command to determine your SDRAM and flash memory.
UBoot, FPGA, Miniboot, BootROM, and Upgrade Requirements
Note. Field upgrade of existing OmniSwitch units is required prior to loading this release. See the Upgrading OmniSwitch 6800, 6850, and 9000 Series Switches to 6.1.3.R01 upgrade instructions for more information.
The software versions listed in this section are the minimum required, except where otherwise noted.
OmniSwitch 6800 Series
• Miniboot: 6.1.2.49.R03 or later
• BootROM: 6.1.2.49.R03 or later
OmniSwitch 6850 Series
• UBoot: 6.1.3.601.R01 or later
• Miniboot.uboot: 6.1.3.601.R01 or later
OmniSwitch 9000 Series
• UBoot NI: 6.1.1.167.R02 or later
• UBoot CMM: 6.1.1.167.R02 or later
• Miniboot.uboot CMM: 6.1.1.167.R02 or later
• FPGA CMM: Major Revision: 2 Minor Revision: 25 (displays as 0x19)
POE Firmware
• 5.01
page 4
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
New Hardware Supported
New Hardware Supported
The following new hardware is supported subject to the feature exceptions and problem reports described
later in these release notes.
New Chassis
The following new OmniSwitch 6850 Series chassis are available in this release.
Note. The USB port on the OmniSwitch 6850 Series units is not supported in this release.
OS6850-24L
The OmniSwitch 6850-24L is a stackable edge/workgroup switch offering 20 unshared 10/100Base-T
ports, as well as four combo ports individually configurable to be 10/100/1000Base-T or 1000 Base-X
high speed connections.
OS6850-48L
The OmniSwitch 6850-48L is a stackable edge/workgroup switch offering 44 unshared 10/100Base-T
ports, as well as four combo ports individually configurable to be 10/100/1000Base-T or 1000Base-X high
speed connections.
OS6850-P24L
The OmniSwitch 6850-P24L is a stackable edge/workgroup switch offering 20 unshared 10/100Base-T
Power over Ethernet (PoE) ports, as well as four combo ports individually configurable to be
10/100/1000 Base-T PoE or 1000 Base-X high speed connections.
OS6850-P48L
The OmniSwitch 6850-P48L is a stackable edge/workgroup switch offering 44 unshared 10/100Base-T
Power over Ethernet (PoE) ports, as well as four combo ports individually configurable to be
10/100/1000 Base-T PoE or 1000 Base-X high speed connections.
OS6850-U24X
The OmniSwitch 6850-U24X is a stackable edge/workgroup switch offering 22 1000Base-X MiniGBIC
SFP ports, two (2) 10 Gigabit XFP slots, as well as two combo ports individually configurable to be
10/100/1000 Base-T ports.
Note. OmniSwitch 6850 units and OmniSwitch 6800 units should not be mixed in the same stack. See the
OmniSwitch 6850 Series Hardware Users Guide for more information on OmniSwitch 6850 hardware
features.
The following new OmniSwitch 9000 Series chassis are available in this release:
OS9800
The OmniSwitch 9800 chassis supports a high-performance switch fabric and provides 16 slots for Ethernet, Gigabit Ethernet, and/or 10 Gigabit Ethernet Network Interface (NI) modules. An additional two slots
are reserved for primary and redundant Chassis Management Modules (CMMs). The OmniSwitch 9800
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 5
New Hardware Supported
supports a maximum of four power supplies and up to 384 10/100/1000 copper ports and/or 1000 Mbps
fiber ports. It is suitable for wiring closet installations. It can also be equipped with up to 96 10 Gigabit
Ethernet ports for use as the core switch.
Note. OmniSwitch 9700 NI modules and OmniSwitch 7700 NI modules should not be mixed in the same
chassis. See the OmniSwitch 9000 Series Hardware Users Guide for more information on OmniSwitch
9000 hardware features.
New Chassis Management Module (CMM)
The following CMM is available in this release:
OS9800-CMM
The Chassis Management Module (CMM) is the management unit for OmniSwitch 9000 switches. In its
role as the management unit, the CMM provides key system services, including:
• Console, USB, and Ethernet management port connections to the switch
• Software and configuration management, including the Command Line Interface (CLI)
• Web-based management (WebView)
• SNMP management
• Power distribution
• Switch diagnostics
• Important availability features, including redundancy (when used in conjunction with another CMM),
software rollback, temperature management, and power management
• The CMM also contains the switch fabric unit for the OmniSwitch 9000. Data passing from one NI
module to another passes through the CMM fabric. When two CMMs are installed, both fabrics are
normally active.
Note. The USB port on the front panel of the CMM is not supported in this release.
New Network Interface (NI) Modules
The following NI modules are available in this release:
OS9-GNI-P24
Provides 24 auto-sensing twisted-pair Power over Ethernet (PoE) ports, individually configurable as
10BaseT, 100BaseTX, or 1000BaseT.
OS9-XNI-U6
Provides six 10 Gigabit XFP slots.
page 6
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
New Hardware Supported
New Power over Ethernet Components/Enhancements
OmniSwitch 9000 OS-IP-SHELF PoE Power Shelf
The OS-IP-SHELF unit is the peripheral power shelf for the chassis that provides the electrical current
used for PoE ports. Up to four 600W OS-IP-SHELF specific power supply modules are available. These
are load sharing modules that can provide up to a maximum of 2400W PoE power source.
OmniSwitch 9600 Support for PS-510W AC/PS-360W AC
The 510W and 360W power supplies can be used as an alternate power source for PoE on the OmniSwitch
9600. A single 510W power supply allocates 380W for the PoE functionality; Similarly, a single 360W
power supply allocates 230W for the PoE functionality. Only one power supply module can be installed
per switch, not both. These power modules support load sharing across all the POE NIs in the OmniSwitch
9600 chassis.
Note that the 510W/360W power supplies are only supported on the OmniSwitch 9600 and not on the
OmniSwitch 9700 or OmniSwitch 9800.
New Ethernet Transceivers (SFPs)
The following Gigabit Ethernet, dual-speed Ethernet, and 100FX Ethernet transceivers are available in this
release:
Gigabit Ethernet Transceiver (SFP MSA)
SFP-GIG-T - 1000Base-T Gigabit Ethernet Transceiver (SFP MSA) - Supports category 5, 5E, and 6
copper cabling up to 100m. SFP only works in 1000 Mbit/s speed and full-duplex mode.
Dual Speed Ethernet Transceivers
SFP-DUAL-MM - Dual Speed 100Base-FX or 1000Base-X Ethernet optical transceiver (SFP MSA).
Supports multimode fiber over 1310nm wavelength (nominal) with an LC connector. Typical reach of
550m at Gigabit speed and 2km at 100Mbit speed.
Note:
• at 100Mbit speed, this SFP can interoperate with SFP-100-LC-MM or similar transceiver on the other
end.
• at Gigabit speed, this SFP cannot interoperate with SFP-GIG-SX or similar transceiver on the other end
running over 850nm wavelength.
SFP-DUAL-SM10 - Dual Speed 100Base-FX or 1000Base-X Ethernet optical transceiver (SFP MSA).
Supports single mode fiber over 1310nm wavelength (nominal) with an LC connector. Typical reach of
10km at Gigabit speed and 100Mbit speed.
Note:
• at 100Mbit speed, this SFP can interoperate with SFP-100-LC-SM15 or similar transceiver.
• at Gigabit speed, this SFP can interoperate with SFP-GIG-LX or similar transceiver.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 7
New Hardware Supported
100FX Ethernet Transceivers
SFP-100-BX20LT - 100Base-BX SFP transceiver with an SC type interface. This bi-directional transceiver is designed for use over single mode fiber optic on a single strand link up to 20KM point-to-point.
This transceiver is normally used in the central office (OLT) transmits 1550nm and receives 1310nm optical signal.
SFP-100-BX20NU - 100Base-BX SFP transceiver with an SC type interface. This bi-directional transceiver is designed for use over single mode fiber optic on a single strand link up to 20KM point-to-point.
This transceiver is normally used in the client (ONU) transmits 1310nm and receives 1550nm optical
signal.
SFP-100-LC-MM - 100Base-FX SFP transceiver. Supports multi mode fiber over 1310nm wavelength
(nominal) with an LC connector.
SFP-100-LC-SM15 - 100Base-FX SFP transceiver. Supports single mode fiber over 1310nm wavelength
(nominal) with an LC connector. Typical reach of 15 Km.
SFP-100-LC-SM40 - 100Base-FX SFP transceiver Supports single mode fiber over 1310nm wavelength
(nominal) with an LC connector. Typical reach of 40 Km.
Supported Configuration Matrix for New Ethernet Transceivers
The following table shows the Ethernet transceiver configurations supported with the 6.1.3.R01 release:
OS6800/OS6850
Combo Ports
OS6800-U24
Non-Combo
Ports
supported
supported
supported
supported
SFP-DUAL-MM - Dual Speed
100Base-FX or 1000Base-X Ethernet optical transceiver.
unsupported
unsupported
supported
supported
SFP-DUAL-SM10 - Dual Speed
100Base-FX or 1000Base-X Ethernet optical transceiver (SFP MSA)
unsupported
unsupported
supported
supported
SFP-100-BX20LT - 100Base-BX
SFP bi-directional transceiver.
unsupported
unsupported
supported
unsupported
SFP-100-BX20NU - 100Base-BX
SFP bidirectional transceiver.
unsupported
unsupported
supported
unsupported
SFP-100-LC-MM - 100Base-FX
SFP transceiver.
unsupported
unsupported
supported
unsupported
SFP-100-LC-SM15 - 100Base-FX
SFP transceiver.
unsupported
unsupported
supported
unsupported
SFP-100-LC-SM40 - 100Base-FX
SFP transceiver.
unsupported
unsupported
supported
unsupported
SFP
SFP-GIG-T - 1000Base-T Gigabit Ethernet Transceiver (SFP
MSA).
page 8
OS6850-U24X OS9-GNI-U24
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Supported Hardware/Software Combinations
Supported Hardware/Software Combinations
The following table shows the 6.1 software releases that support each of the listed OS9000, OS6850, and
OS6800 module types:
Module Type
Part Number
6.1.1.R01
6.1.1.R02
6.1.2.R03
6.1.3.R01
OS96/9700 CMM, REV B
902369
supported
supported
n/a
supported
OS96/9700 CMM, REV C
902444
supported
supported
n/a
supported
OS9800 CMM
902492
n/a
supported
OS9-GNI-C24, ASIC A1
902367
supported
supported
n/a
supported
OS9-GNI-U24, ASIC A1
902370
supported
supported
n/a
supported
OS9-XNI-U2, ASIC A1
902379
supported
supported
n/a
supported
OS9-GNI-C24, ASIC B2
902394
not supported
supported
n/a
supported
OS9-GNI-U24, ASIC B2
902396
not supported
supported
n/a
supported
OS9-XNI-U2, ASIC B2
902397
not supported
supported
n/a
supported
OS9-GNI-P24, ASIC B2
902395
not supported not supported
n/a
supported
OS9-XNI-U6, ASIC B2
902398
not supported not supported
n/a
supported
OS6850-24
902457
n/a
n/a
supported
supported
OS6850-48
902495
n/a
n/a
supported
supported
OS6850-24X
902458
n/a
n/a
supported
supported
OS6850-48X
902462
n/a
n/a
supported
supported
OS6850-P24
902459
n/a
n/a
supported
supported
OS6850-P48
902463
n/a
n/a
supported
supported
OS6850-P24X
902460
n/a
n/a
supported
supported
OS6850-P48X
902464
n/a
n/a
supported
supported
OS6850-U24X
902418
n/a
n/a
not supported
supported
OS6850-24L
902487
n/a
n/a
not supported
supported
OS6850-48L
902489
n/a
n/a
not supported
supported
OS6850-P24L
902488
n/a
n/a
not supported
supported
OS6850-P48L
902490
n/a
n/a
not supported
supported
OS6800-24
902349
n/a
n/a
supported
supported
OS6800-48
902350
n/a
n/a
supported
supported
OS6800-24L
902377
n/a
n/a
supported
supported
OS6800-48L
902378
n/a
n/a
supported
supported
OS6800-U24
902351
n/a
n/a
supported
supported
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
not supported not supported
page 9
Supported Hardware/Software Combinations
To determine the ASIC revision for a specific NI, use the show ni command. For example, the following
show ni output display shows a B2 revision level for NI 1:
DC-Core ->> show ni 1
Module in slot 1
Model Name:
Description:
Part Number:
Hardware Revision:
Serial Number:
Manufacture Date:
Firmware Version:
Admin Status:
Operational Status:
Power Consumption:
Power Control Checksum:
MAC Address:
ASIC - Physical 1:
CPLD - Physical 1:
UBOOT Version :
UBOOT-miniboot Version :
POE SW Version :
OS9-GNI-C24,
10-1000 RJ45,
902394-40,
C13,
G1511279,
MAY 03 2006,
,
POWER ON,
UP,
51,
0x0,
00:d0:95:e6:54:80,
BCM56504_B2
0005/00
6.1.1.167.R02
No Miniboot
n/a
To determine the CMM board revision, use the show cmm command. For example, the following show
cmm output display shows a C revision level for the CMM board:
DC-Core ->> show cmm
Module in slot CMM-A-1
Model Name:
Description:
Part Number:
Hardware Revision:
Serial Number:
Manufacture Date:
Firmware Version:
Admin Status:
Operational Status:
Power Consumption:
Power Control Checksum:
MAC Address:
page 10
OS9700-CFM,
FABRIC BOARD,
902444-10,
C11,
G1810128,
MAY 08 2006,
2,
POWER ON,
UP,
27,
0x0,
00:d0:95:e0:6c:ac,
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
New Software Features
New Software Features
The following software features and/or the platforms they are supported on are new with the 6.1.3.R01
release, subject to the feature exceptions and problem reports described later in these release notes:
Feature Summary
Feature
Platform
Software Package
802.1Q 2005 (MSTP)
all
base
802.1W (RSTP) Default
all
base
802.1x Device Classification
OS9000
base
Access Control Lists (ACLs) for IPv6
OS6850/OS9000
base
ACL Manager (ACLMAN)
OS9000
base
Authenticated Switch Access - TACACS+
all
base
BGP Graceful Restart
all
base
DHCP Option-82
OS9000
base
DHCP Snooping
OS9000
base
Generic UDP Relay
all
base
IP DoS Enhancements
OS6850/OS9000
base
IP Multicast Switching (IPMS) - Proxying
all
base
IPv6 Multicast Switching (IPMS) - Proxying OS6850/OS9000
base
IP Route Map Redistribution
all
base
L2 DHCP Snooping
all
base
L2 Static Multicast Addresses
all
base
L2 MAC Address Table Size Enhancement
OS9000
base
OSPFv3
OS6850/OS9000
base
advanced routing
PIM
PIM-SSM (Source-Specific Multicast)
all
base
advanced routing
Policy Based Routing (Permanent Mode)
OS6850
base
Port Mapping
OS9000
base
Port Mirroring (1:128)
OS6850/OS9000
base
Power over Ethernet (PoE)
OS9000
base
Redirection Policies
(Port and Link Aggregate)
OS6850
base
Secure Copy (SCP)
OS9000
base
Server Load Balancing (SLB)
OS6850/OS9000
base
SSH Public Key Authentication
OS9000
base
Syslog to Multiple Hosts
all
base
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 11
New Software Features
Feature
Platform
Software Package
VLAN Range Support
all
base
VLAN Stacking and Translation
OS6850/OS9000
base
VRRPv3
OS6850/OS9000
base
page 12
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Software Supported
Software Supported
In addition to the new software features introduced with the 6.1.3.R01 release, the following software
features are also supported in 6.1.3.R01, subject to the feature exceptions and problem reports described
later in these release notes:
Feature Summary
Feature
Platform
Software Package
802.1Q
all
base
802.1d/1w Spanning Tree
all
base
802.1x Multiple Client Support
all
base
802.1x Device Classification
(Access Guardian)
all
base
Access Control Lists (ACLs)
all
base
ACL & Layer 3 Security
all
base
ACL Manager (ACLMAN)
OS6850/OS9000
base
Authenticated Switch Access
all
base
Authenticated VLANs
all
base
Automatic VLAN Containment (AVC)
all
base
Basic IPv4 Routing
all
base
Basic IPv6 Routing (static, RIPng)
all
base
BGP4
all
base
BPDU Shutdown Ports
OS6800/OS9000
base
Command Line Interface (CLI)
all
base
DHCP Relay
all
base
DHCP Option-82
all
base
DHCP Snooping
all
base
DNS Client
all
base
Dynamic VLAN Assignment (Mobility)
all
base
DVMRP
all
base
End User Partitioning
all
base
Ethernet Interfaces
all
base
Flood/Storm Control
all
base
Health Statistics
all
base
HTTP/HTTPS Port Configuration
all
base
Interswitch Protocols (AMAP)
all
base
IP Multicast Switching (IPMS)
all
base
IPv6 Multicast Switching (MLD)
OS6850/OS9000
base
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 13
Software Supported
Feature
Platform
Software Package
IPv6 (NPD)
all
base
IPX Routing
all
base
Learned Port Security (LPS)
all
base
Link Aggregation (static & 802.3ad)
all
base
Multicast Routing
all
base
Multinetting
all
base
NTP Client
all
base
OSPFv2
all
base
Partitioned Switch Management
all
base
Per-VLAN DHCP Relay
all
base
PIM-SM
PIM-SSM (Source-Specific Multicast)
all
base
Policy Server Management
all
base
Policy Based Routing (Permanent Mode)
OS6850/OS9000
base
Port Mapping
all
Port Mirroring (1:24)
OS6800
base
Port Monitoring
all
base
Power over Ethernet (PoE)
OS6850/OS9000
base
Quality of Service (QoS)
all
base
Redirection Policies
(Port and Link Aggregate)
OS6850/OS9000
base
RMON
all
base
Router Discovery Protocol (RDP)
all
base
Routing Protocol Preference
all
base
Secure Copy (SCP)
all
base
Secure Shell (SSH)
all
base
SSH Public Key Authentication
all
base
sFlow
OS6850/OS9000
base
Smart Continuous Switching
Hot Swap
Management Module Failover
Power Monitoring
Redundancy
all
base
SNMP
all
base
Source Learning
all
base
Software Rollback
all
base
Spanning Tree
all
base
Switch Logging
all
base
page 14
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Software Supported
Feature
Platform
Software Package
Text File Configuration
all
base
User Definable Loopback Interface
all
base
VLANs
all
base
VRRP
all
base
Web-Based Management (WebView)
all
base
Feature Descriptions
802.1Q
Alcatel-Lucent’s 802.1Q is an IEEE standard for sending frames through the network tagged with VLAN
identification. 802.1Q tagging is the IEEE version of VLANs. It is a method of segregating areas of a
network into distinct VLANs. By attaching a label, or tag, to a packet, it can be identified as being from a
specific area or identified as being destined for a specific area.
When a port is enabled to accept tagged traffic, by default both 802.1Q tagged and untagged traffic is
automatically accepted on the port. Configuring the port to accept only tagged traffic is also supported.
802.1Q 2005 (MSTP)
802.1Q 2005 (Q2005) is a new version of Multiple Spanning Tree Protocol (MSTP) that is a combination
of the 802.1D 2004 and 802.1S protocols. This implementation of Q2005 also includes improvements to
edge port configuration and provides administrative control to restrict port role assignment and the propagation of topology change information through bridge ports.
The 6.1.3.R01 release provides support for this feature on the OmniSwitch 6800 Series, OmniSwitch 6850
Series, and OmniSwitch 9000 Series.
802.1W (RSTP) Default
The Rapid Spanning Tree Protocol (RSTP) is now the default Spanning Tree protocol for the OmniSwitch
6800/6850/9000 regardless of which mode (flat or 1x1) is active.
802.1x Device Classification (Access Guardian)
In addition to the authentication and VLAN classification of 802.1x clients (supplicants), this implementation of 802.1x secure port access extends this type of functionality to non-802.1x clients (non-supplicants). To this end device classification policies are introduced to handle both supplicant and nonsupplicant access to 802.1x ports.
Supplicant policies use 802.1x authentication via a remote RADIUS server and provide alternative methods for classifying supplicants if the authentication process either fails or does not return a VLAN ID.
Non-supplicant policies use MAC authentication via a remote RADIUS server or can bypass authentication and only allow strict assignment to specific VLANs. MAC authentication verifies the source MAC
address of a non-supplicant device via a remote RADIUS server. Similar to 802.1x authentication, the
switch sends RADIUS frames to the server with the source MAC address embedded in the username and
password attributes.
Device Classification is supported on the OmniSwitch 6800 Series and OmniSwitch 6850 Series. The
6.1.3.R01 release provides support for this feature on the OmniSwitch 9000 Series.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 15
Software Supported
Access Control Lists (ACLs)
Access Control Lists (ACLs) are Quality of Service (QoS) policies used to control whether or not packets
are allowed or denied at the switch or router interface. ACLs are sometimes referred to as filtering lists.
ACLs are distinguished by the kind of traffic they filter. In a QoS policy rule, the type of traffic is specified in the policy condition. The policy action determines whether the traffic is allowed or denied.
In general, the types of ACLs include:
• Layer 2 ACLs—for filtering traffic at the MAC layer. Usually uses MAC addresses or MAC groups for
filtering.
• Layer 3/4 ACLs—for filtering traffic at the network layer. Typically uses IP addresses or IP ports for
filtering; note that IPX filtering is not supported.
• Multicast ACLs—for filtering IGMP traffic.
Access Control Lists (ACLs) for IPv6
The 6.1.3.R01 release provides support for IPv6 ACLs on the OmniSwitch 6850 Series and OmniSwitch
9000 Series. The following QoS policy conditions are now available for configuring ACLs to filter IPv6
traffic:
source ipv6
destination ipv6
ipv6
nh (next header)
flow-label
Note the following when using IPv6 ACLs:
• Trusted/untrusted behavior is the same for IPv6 traffic as it is for IPv4 traffic.
• IPv6 policies do not support the use of network groups, service groups, map groups, or MAC groups.
• IPv6 multicast policies are not supported.
• Anti-spoofing and other UserPorts profiles/filters do not support IPv6.
• The default (built-in) network group, “Switch”, only applies to IPv4 interfaces. There is no such group
for IPv6 interfaces.
Note. IPv6 ACLs are not supported on A1 NI modules. Use the show ni command to verify the version of
the NI module. Contact your Alcatel-Lucent support representative if you are using A1 boards.
ACL & Layer 3 Security
The following additional ACL features are available for improving network security and preventing malicious activity on the network:
• ICMP drop rules—Allows condition combinations in policies that will prevent user pings, thus reduc-
ing DoS exposure from pings. Two condition parameters are also available to provide more granular
filtering of ICMP packets: icmptype and icmpcode.
page 16
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Software Supported
• TCP connection rules—Allows the determination of an established TCP connection by examining
TCP flags found in the TCP header of the packet. Two condition parameters are available for defining
a TCP connection ACL: established and tcpflags.
• Early ARP discard—ARP packets destined for other hosts are discarded to reduce processing over-
head and exposure to ARP DoS attacks. No configuration is required to use this feature, it is always
available and active on the switch. Note that ARPs intended for use by a local subnet, AVLAN, and
VRRP are not discarded.
• UserPorts—A port group that identifies its members as user ports to prevent spoofed IP traffic. When
a port is configured as a member of this group, packets received on the port are dropped if they contain
a source IP network address that does not match the IP subnet for the port.
• UserPorts Profile—In addition to spoofed traffic, it is also possible to configure a global UserPorts
profile to specify additional types of traffic, such as BPDU, RIP, OSPF, and/or BGP, to monitor on
user ports. The UserPorts profile also determines whether user ports will filter the unwanted traffic or
will administratively shutdown when the traffic is received. Note that this profile only applies to those
ports that are designated as members of the UserPorts port group.
• DropServices—A service group that improves the performance of ACLs that are intended to deny
packets destined for specific TCP/UDP ports. This group only applies to ports that are members of the
UserPorts group. Using the DropServices group for this function minimizes processing overhead,
which otherwise could lead to a DoS condition for other applications trying to use the switch.
ACL Manager
The Access Control List Manager (ACLMAN) is a function of the Quality of Service (QoS) application
that provides an interactive shell for using common industry syntax to create ACLs. Commands entered
using the ACLMAN shell are interpreted and converted to Alcatel CLI syntax that is used for creating
QoS filtering policies.
This implementation of ACLMAN also provides the following features:
• Importing of text files that contain common industry ACL syntax.
• Support for both standard and extended ACLs.
• Creating ACLs on a single command line.
• The ability to assign a name, instead of a number, to an ACL or a group of ACL entries.
• Sequence numbers for named ACL statements.
• Modifying specific ACL entries without having to enter the entire ACL each time to make a change.
• The ability to add and display ACL comments.
• ACL logging extensions to display Layer 2 through 4 packet information associated with an ACL.
ACLMAN is supported on the OmniSwitch 6850 Series. The 6.1.3.R01 release provides support for this
feature on the OmniSwitch 9000 Series.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 17
Software Supported
Authenticated Switch Access
Authenticated Switch Access (ASA) is a way of authenticating users who want to manage the switch. With
authenticated access, all switch login attempts using the console or modem port, Telnet, FTP, SNMP, or
HTTP require authentication via the local user database or via a third-party server. The type of server may
be an authentication-only mechanism or an authentication, authorization, and accounting (AAA) mechanism.
AAA servers are able to provide authorization for switch management users as well as authentication.
(They also may be used for accounting.) User login information and user privileges may be stored on the
servers. In addition to the Remote Authentication Dial-In User Service (RADIUS) and Lightweight Directory Access Protocol (LDAP) servers, using a Terminal Access Controller Access Control System
(TACACS+) server is now supported with the 6.1.3.R01 release.
Authentication-only servers are able to authenticate users for switch management access, but authorization (or what privileges the user has after authenticating) are determined by the switch. Authenticationonly servers cannot return user privileges to the switch. The authentication-only server supported by the
switch is ACE/Server, which is a part of RSA Security’s SecurID product suite. RSA Security’s ACE/
Agent is embedded in the switch.
By default, switch management users may be authenticated through the console port via the local user
database. If external servers are configured for other management interfaces but the servers become
unavailable, the switch will poll the local user database for login information if the switch is configured
for local checking of the user database. The database includes information about whether or not a user is
able to log into the switch and what kinds of privileges or rights the user has for managing the switch.
Authenticated VLANs
Authenticated VLANs control user access to network resources based on VLAN assignment and a user
log-in process; the process is sometimes called user authentication or Layer 2 Authentication. (Another
type of security is device authentication, which is set up through the use of port-binding VLAN policies or
static port assignment.) The number of possible AVLAN users is 1048.
Layer 2 Authentication is different from Authenticated Switch Access, which is used to grant individual
users access to manage the switch.
The Mac OS X 10.3.x is supported for AVLAN web authentication using JVM-v1.4.2.
Automatic VLAN Containment (AVC)
In an 802.1s Multiple Spanning Tree (MST) configuration, it is possible for a port that belongs to a
VLAN, which is not a member of an instance, to become the root port for that instance. This can cause a
topology change that could lead to a loss of connectivity between VLANs/switches. Enabling Automatic
VLAN Containment (AVC) helps to prevent this from happening by making such a port an undesirable
choice for the root.
When AVC is enabled, it identifies undesirable ports and automatically configures them with an infinite
path cost value.
Balancing VLANs across links according to their Multiple Spanning Tree Instance (MSTI) grouping is
highly recommended to ensure that there is not a loss of connectivity during any possible topology
changes. Enabling AVC on the switch is another way to prevent undesirable ports from becoming the root
for an MSTI.
page 18
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Software Supported
Basic IPv4 Routing
Internet Protocol (IP) is a network-layer (Layer 3) protocol that contains addressing and control information that allow packets to be forwarded on a network. IP is the primary network-layer protocol in the Internet protocol suite. Along with the Transmission Control Protocol (TCP), IP represents the heart of the
Internet protocols. IP is associated with several Layer 3 and Layer 4 protocols. These protocols are built
into the base code loaded on the switch and they include:
• Transmission Control Protocol (TCP)
• User Datagram Protocol (UDP)
• Bootstrap Protocol (BOOTP)/Dynamic Host Configuration Protocol (DHCP)
• Simple Network Management Protocol (SNMP)
• Telnet
• File Transfer Protocol (FTP)
• Address Resolution Protocol (ARP)
• Internet Control Message Protocol (ICMP)
• RIP I / RIP II
The base IP software allows one to configure an IP router interface, static routes, a default route, the
Address Resolution Protocol (ARP), the router primary address, the router ID, the Time-to-Live (TTL)
Value, IP-directed broadcasts, and the Internet Control Message Protocol (ICMP). In addition, this software allows one to trace an IP route, display Transmission Control Protocol (TCP) information, and
display User Datagram Protocol (UDP) information.
OmniSwitch 9000 supports hardware routing/flooding to static arp with multicast MAC address.
Note. The switch operates only in single MAC router mode. In this mode, each router VLAN is assigned
the same MAC address, which is the base chassis MAC address for the switch.
BGP4
The Border Gateway Protocol (BGP) is an exterior routing protocol that guarantees the loop-free exchange
of routing information between autonomous systems. There are three versions of the BGP protocol—
versions 2, 3, and 4. The Alcatel-Lucent implementation supports BGP version 4 as defined in RFC 1771.
The Alcatel-Lucent implementation of BGP is designed for enterprise networks, specifically for border
routers handling a public network connection, such as the organization’s Internet Service Provider (ISP)
link. Up to 65,000 route table entries and next hop routes can be supported by BGP.
BGP Graceful Restart
BGP Graceful Restart is now supported and is enabled by default. On OmniSwitch devices in a redundant
CMM configuration, during a CMM takeover/failover, interdomain routing is disrupted. Alcatel-Lucent
Operating System BGP needs to retain forwarding information and also help a peering router performing a
BGP restart to support continuous forwarding for inter-domain traffic flows by following the BGP graceful restart mechanism.
The 6.1.3.R01 release provides support for this feature on the OmniSwitch 6800 Series, OmniSwitch 6850
Series, and OmniSwitch 9000 Series.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 19
Software Supported
BPDU Shutdown Ports
It is possible to configure a global UserPorts profile, as described in “ACL & Layer 3 Security”, to monitor BPDU on user ports. Such a profile also determines whether user ports will filter BPDU or will administratively shutdown when BPDU are received on the port. Note that this functionality only applies to ports
that are designated as members of the UserPorts port group.
A port configured to administratively shutdown when BPDU are detected will generate an inferior BPDU
every 5 seconds. This will prevent loops in the network if two BPDU shutdown ports are accidentally
bridged together either through an external loop or through a hub, since both ports would be receiving
inferior BPDUs.
Command Line Interface (CLI)
Alcatel-Lucent’s command line interface (CLI) is a text-based configuration interface that allows you to
configure switch applications and to view switch statistics. Each CLI command applicable to the switch is
defined in the CLI Reference guide. All command descriptions listed in the Reference Guide include
command syntax definitions, defaults, usage guidelines, example screen output, and release history.
The CLI uses single-line text commands that are similar to other industry standard switch interfaces.
DHCP Relay
DHCP Relay allows you to forward DHCP broadcast requests to configurable DHCP server IP address in
a routing environment.
DHCP Relay is configured using the IP helper set of commands.
DHCP Option-82 (Relay Agent Information Option)
The DHCP Option-82 feature enables the relay agent to insert identifying information into client-originated DHCP packets before the packets are forwarded to the DHCP server. The implementation of this
feature is based on the functionality defined in RFC 3046.
When DHCP Option-82 is enabled, communications between a DHCP client and a DHCP server are
authenticated by the relay agent. To accomplish this task, the agent adds Option-82 data to the end of the
options field in DHCP packets sent from a client to a DHCP server.
If the relay agent receives a DHCP packet from a client that already contains Option-82 data, the packet is
dropped by default. However, it is possible to configure a DHCP Option-82 policy that directs the relay
agent to drop, keep, or replace the existing Option-82 data and then forward the packet to the server.
DHCP Option-82 is supported on the OmniSwitch 6800 Series and OmniSwitch 6850 Series. The
6.1.3.R01 release provides support for this feature on the OmniSwitch 9000 Series.
DHCP Snooping
DHCP Snooping improves network security by filtering DHCP packets received from devices outside the
network and building and maintaining a binding table (database) to log DHCP client access information.
There are two levels of operation available for the DHCP Snooping feature: switch level or VLAN level.
To identify DHCP traffic that originates from outside the network, DHCP Snooping categorizes ports as
either trusted or untrusted. A port is trusted if it is connected to a device inside the network, such as a
DHCP server. A port is untrusted if it is connected to a device outside the network, such as a customer
switch or workstation. The port trust mode is also configurable through the CLI.
Additional DHCP Snooping functionality includes the following:
page 20
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Software Supported
• Traffic Suppression—Prevents the flooding of DHCP packets on the default VLAN for a DHCP
Snooping port. Note that enabling traffic suppression on a port will prevent DHCP traffic between a
DHCP server and client that belong to the same VLAN domain.
• IP Source Filtering—Restricts DHCP Snooping port traffic to only packets that contain the client
source MAC address and IP address obtained from the DHCP lease information. The DHCP Snooping
binding table is used to verify the client lease information for the port that is enabled for IP source
filtering.
• Rate Limiting—Limits the number of DHCP packets on a port. This functionality is provided using
the QoS application to configure ACLs for the port.
DHCP Snooping is supported on the OmniSwitch 6800 Series and OmniSwitch 6850 Series. The
6.1.3.R01 release provides support for this feature on the OmniSwitch 9000 Series.
DNS Client
A Domain Name System (DNS) resolver is an internet service that translates host names into IP addresses.
Every time you enter a host name, a DNS service must look up the name on a server and resolve the name
to an IP address. You can configure up to three domain name servers that will be queried in turn to resolve
the host name. If all servers are queried and none can resolve the host name to an IP address, the DNS
fails. If the DNS fails, you must either enter an IP address in place of the host name or specify the necessary lookup tables on one of the specified servers.
Dynamic VLAN Assignment (Mobility)
Dynamic assignment applies only to mobile ports and requires the additional configuration of VLAN
rules. When traffic is received on a mobile port, the packets are examined to determine if their content
matches any VLAN rules configured on the switch. Rules are defined by specifying a port, MAC address,
protocol, network address, binding, or DHCP criteria to capture certain types of network device traffic. It
is also possible to define multiple rules for the same VLAN. A mobile port is assigned to a VLAN if its
traffic matches any one VLAN rule.
DVMRP
Distance Vector Multicast Routing Protocol (DVMRP) is a dense-mode multicast routing protocol.
DVMRP—which is essentially a “broadcast and prune” routing protocol—is designed to assist routers in
propagating IP multicast traffic through a network. DVMRP works by building per-source broadcast trees
based on routing exchanges, then dynamically creating per-source, group multicast delivery trees by pruning the source’s truncated broadcast tree.
End User Partitioning (EUPM)
EUPM is used for customer login accounts that are configured with end-user profiles (rather than functional privileges specified by partitioned management). Profiles specify command areas as well as VLAN
and/or port ranges to which the user has access. These profiles are typically used for end users rather than
network administrators.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 21
Software Supported
Ethernet Interfaces
Ethernet and Gigabit Ethernet port software is responsible for a variety of functions that support Ethernet,
Gigabit, and 10 Gigabit Ethernet ports. These functions include initialization of ports, notifying other software modules when a port goes down, configuration of basic line parameters, gathering of statistics for
Ethernet and Gigabit Ethernet ports, and responding to administrative enable/disable requests.
Configurable parameters include: autonegotiation (copper ports 10/100/1000), trap port link messages,
flood control, line speed, duplex mode, inter-frame gap, resetting statistics counters, and maximum and
peak flood rates.
Flood control is configurable on ingress interfaces (flood rate and including/excluding multicast).
Generic UDP Relay
In addition to BOOTP/DHCP relay, generic UDP relay is available. Using generic UDP relay, traffic
destined for well-known service ports (e.g., NBNS/NBDD, DNS, TFTP, and TACACS) or destined for a
user-defined service port can be forwarded to a maximum of 256 VLANs on the switch.
The 6.1.3.R01 release provides support for this feature on the OmniSwitch 6800 Series, OmniSwitch 6850
Series, and OmniSwitch 9000 Series.
Health Statistics
To monitor resource availability, the NMS (Network Management System) needs to collect significant
amounts of data from each switch. As the number of ports per switch (and the number of switches)
increases, the volume of data can become overwhelming. The Health Monitoring feature can identify and
monitor a switch’s resource utilization levels and thresholds, improving the efficiency in data collection.
Health Monitoring provides the following data to the NMS:
• Switch-level input/output, memory and CPU utilization levels
• Module-level and port-level input/output utilization levels
For each monitored resource, the following variables are defined:
• Most recent utilization level (percentage)
• Average utilization level over the last minute (percentage)
• Average utilization level over the last hour (percentage)
• Maximum utilization level over the last hour (percentage)
• Threshold level
Additionally, Health Monitoring provides the capacity to specify thresholds for the resource utilization
levels it monitors, and generates traps based on the specified threshold criteria.
HTTP/HTTPS Port Configuration
The default HTTP port and the default Secure HTTP (HTTPS) port can be configured for the embedded
Web server in the switch.
page 22
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Software Supported
Interswitch Protocol (AMAP)
Alcatel-Lucent Interswitch Protocols (AIP) are used to discover adjacent switches and retain mobile port
information across switches. By default, AMAP is enabled.
Alcatel-Lucent Mapping Adjacency Protocol (AMAP) is used to discover the network topology of Alcatel-Lucent switches in a particular installation. Using this protocol, each switch determines which switches
are adjacent to it by sending and responding to Hello update packets. For the purposes of AMAP, adjacent
switches are those that:
• Have a Spanning Tree path between them
• Do not have any switch between them on the Spanning Tree path that has AMAP enabled
IP DoS Enhancements
By default, the switch filters the following denial of service (DoS) attacks, which are security attacks
aimed at devices that are available on a private network or the Internet:
• ARP Flood Attack - OS6800/OS6850/OS9000
• Invalid IP Attack - OS6850/OS9000
• Multicast IP and MAC Address Mismatch - OS6850/OS9000
• Ping Overload - OS6850/OS9000
• Packets with loopback source IP address - OS6850/OS9000
The 6.1.3.R01 release provides support for these enhancements on the OmniSwitch 6800 Series,
OmniSwitch 6850 Series, and OmniSwitch 9000 Series as noted above.
IP Multicast Switching (IPMS)
IP Multicast Switching is a one-to-many communication technique employed by emerging applications
such as video distribution, news feeds, conferencing, netcasting, and resource discovery (OSPF, RIP2, and
BOOTP). Unlike unicast, which sends one packet per destination, multicast sends one packet to all devices
in any subnetwork that has at least one device requesting the multicast traffic. Multicast switching also
requires much less bandwidth than unicast techniques and broadcast techniques since the source hosts only
send one data stream to the ports on which destination hosts that request it are attached.
Destination hosts signal their intent to receive a specific multicast stream by sending a request to do so to
a nearby switch using Internet Group Management Protocol (IGMP). The switch then learns on which
ports multicast group subscribers are attached and can intelligently deliver traffic only to the respective
ports. This mechanism is often referred to as IGMP snooping (or IGMP gleaning). Alcatel-Lucent’s
implementation of IGMP snooping is called IP Multicast Switching (IPMS). IPMS allows OmniSwitch
9000 Series switches to efficiently deliver multicast traffic in hardware at wire speed.
Both IGMP version 3 (IGMPv3), which handles forwarding by source IP address and IP multicast destination, and IGMP version 2 (IGMPv2), which handles forwarding by IP multicast destination address only,
are supported. IPMS is supported on IPv4 and IPv6 (MLD) on the OmniSwitch 6850 Series and
OmniSwitch 9000 Series. The OmniSwitch 6800 Series only supports IPMS for IPv4.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 23
Software Supported
IP Multicast Switching (IPMS) - Proxying
IP multicast proxying and configuring the IGMP and MLD unsolicited report interval are now available
with this implementation of IPMS. Proxying enables the aggregation of IGMP and MLD group membership information and the reduction in reporting queriers. The unsolicited report interval refers to the time
period in which to proxy any changed IGMP membership state.
The 6.1.3.R01 release provides support for this feature on the OmniSwitch 6800 Series, OmniSwitch 6850
Series, and OmniSwitch 9000 Series. IPv6 IPMS Proxying is not supported on the OmniSwitch 6800
Series.
IP Route Map Redistribution
Route map redistribution provides the ability to control which routes from a source protocol are learned
and distributed into the network of a destination protocol. A route map consists of one or more userdefined statements that can determine which routes are allowed or denied access to the network. In addition, a route map may also contain statements that modify route parameters before they are redistributed.
Redistribution is configured by specifying a source and destination protocol and the name of an existing
route map. Criteria specified in the route map is applied to routes received from the source protocol.
The 6.1.3.R01 release provides support for this feature on the OmniSwitch 6800 Series, OmniSwitch 6850
Series, and OmniSwitch 9000 Series.
IPv6 (NPD)
IPv6 (documented in RFC 2460) is designed as a successor to IPv4. The changes from IPv4 to IPv6 fall
primarily into the following categories:
• Address size increased from 32 bits (IPv4) to 128 bits (IPv6)
• Dual Stack IPv4/IPv6
• ICMPv6
• Neighbor Discovery
• Stateless Autoconfiguration
• RIPng
• Static Routes
• Tunneling: Configured and 6-to-4 dynamic tunneling
• Ping, traceroute
• FTP and Telnet servers
• DNS client using Authority records
OmniSwitch 9000 switches support hardware-based IPv6 routing.
Note. The switch operates only in single MAC router mode. In this mode, each router VLAN is assigned
the same MAC address, which is the base chassis MAC address for the switch
page 24
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Software Supported
IPX Routing
The Internet Packet Exchange (IPX) protocol, developed by Novell for NetWare, is a Layer 3 protocol
used to route packets through IPX networks. (NetWare is Novell’s network server operating system.) This
implementation of IPX routing is software based with limited performance.
IPX specifies a connectionless datagram similar to the IP packet of TCP/IP networks. An IPX network
address consists of two parts: a network number and a node number. The IPX network number is assigned
by the network administrator. The node number is the Media Access Control (MAC) address for a
network interface in the end node.
L2 DHCP Snooping
By default, DHCP broadcasts are flooded on the default VLAN for the client/server port. If the DHCP
client and server are both members of the same VLAN domain, the broadcast packets from these sources
are bridged as Layer 2 traffic and not processed by the relay agent. As a result, any active relay agent
features (e.g., information Option-82, DHCP Snooping) are not applied to this type of DHCP traffic. The
DHCP Relay application has a traffic suppression feature that suppresses the broadcast of DHCP packets
and forwards these packets to the relay agent for processing.
The 6.1.3.R01 release provides support for this feature on the OmniSwitch 6800 Series, OmniSwitch 6850
Series, and OmniSwitch 9000 Series.
L2 MAC Address Table Size Enhancement
There are now two source learning modes available for the OmniSwitch 9000 Series switches: synchronized and distributed. By default the switch runs in the synchronized mode, which allows a total MAC
address tables size of 16K per chassis. Enabling the distributed mode for the switch increases the table size
to 16K per module and up to 64K per OmniSwitch 9000 chassis.
The 6.1.3.R01 release provides support for this feature on the OmniSwitch 9000 Series; increasing the
MAC address table size is not supported on the OmniSwitch 6800 Series and OmniSwitch 6850 Series.
L2 Static Multicast Addresses
Static multicast MAC addresses are used to send traffic intended for a single destination multicast MAC
address to multiple switch ports within a given VLAN. A static multicast address is assigned to one or
more switch ports for a given VLAN. The ports associated with the multicast address are then identified as
egress ports. When traffic received on ports within the same VLAN is destined for the multicast address,
the traffic is forwarded on the egress ports that are associated with the multicast address.
One of the benefits of using static multicast addresses is that multicast traffic is switched in hardware and
no longer subject to flood limits on broadcast traffic.
The 6.1.3.R01 release provides support for this feature on the OmniSwitch 6800 Series, OmniSwitch 6850
Series, and OmniSwitch 9000 Series.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 25
Software Supported
Learned Port Security (LPS)
Learned Port Security (LPS) provides a mechanism for authorizing source learning of MAC addresses on
10/100/1000, Gigabit, and Gigabit Ethernet ports. Using LPS to control source MAC address learning
provides the following benefits:
• A configurable source learning time limit that applies to all LPS ports.
• A configurable limit on the number of MAC addresses allowed on an LPS port.
• Dynamic configuration of a list of authorized source MAC addresses.
• Static configuration of a list of authorized source MAC addresses.
• Two methods for handling unauthorized traffic: Shutting down the port or only blocking traffic that
violates LPS criteria.
LPS has the following limitations:
• You cannot configure LPS on 10 Gigabit ports.
• You cannot configure LPS on link aggregate ports.
Link Aggregation (static & 802.3ad)
Alcatel-Lucent’s link aggregation software allows you to combine several physical links into one large
virtual link known as a link aggregation group. Using link aggregation can provide the following benefits:
• Scalability. You can configure up to 32 link aggregation groups that can consist of 2, 4, or 8 Ethernet-
ports.
• Reliability. If one of the physical links in a link aggregate group goes down, the link aggregate group
can still operate.
• Ease of Migration. Link aggregation can ease the transition from a Gigabit Ethernet backbone to a 10
Gigabit Ethernet backbone.
• Interoperability with Legacy Switches. Static link aggregation can interoperate with OmniChannel
on legacy switches.
Alcatel’s link aggregation software allows you to configure the following two different types of link
aggregation groups:
• Static link aggregate groups
• Dynamic (802.3ad) link aggregate groups
Multicast Routing
The OmniSwitch 9000 switches support multicast routing on IPv4 and includes configuration options for
multicast address boundaries, the Distance Vector Multicast Routing Protocol (DVMRP), and ProtocolIndependent Multicast (PIM).
Multicast traffic consists of a data stream that originates from a single source and is sent to hosts that have
subscribed to that stream. Live video broadcasts, video conferencing, corporate communications, distance
learning, and distribution of software, stock quotes, and news services are examples of multicast traffic.
Multicast traffic is distinguished from unicast traffic and broadcast traffic.
page 26
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Software Supported
Multicast boundaries confine scoped multicast addresses to a particular domain. Confining scoped
addresses helps to ensure that multicast traffic passed within a multicast domain does not conflict with
multicast users outside the domain.
PIM
PIM-SSM
Protocol-Independent Multicast (PIM) is an IP multicast routing protocol that uses routing information
provided by unicast routing protocols, such as RIP and OSPF. PIM is “protocol-independent” because it
does not rely on any particular unicast routing protocol. Sparse mode PIM (PIM-SM) contrasts with
flood-and-prune dense mode multicast protocols, such as DVMRP and PIM Dense Mode (PIM-DM) in
that multicast forwarding in PIM-SM is initiated only via specific requests, referred to as Join messages.
PIM-DM for IPv4 is now supported with the 6.1.3.R01 release. PIM-DM packets are transmitted on the
same socket as PIM-SM packets, as both use the same protocol and message format. Unlike PIM-SM, in
PIM-DM there are no periodic joins transmitted; only explicitly triggered prunes and grafts. In addition,
there is no Rendezvous Point (RP) in PIM-DM.
Protocol Independent Multicast Source-Specific Multicast (PIM-SSM) is a highly-efficient extension of
PIM. SSM, using an explicit channel subscription model, allows receivers to receive multicast traffic
directly from the source; an RP tree model is not used. In other words, a Shortest Path Tree (SPT)
between the receiver and the source is created without the use of a Rendezvous Point (RP).
Multinetting
This feature allows IP traffic from multiple subnets to coexist on the same VLAN. A network is said to be
multinetted when multiple IP subnets are brought together within a single broadcast domain (VLAN). It is
possible to assign up to eight different IP interfaces per VLAN.
NTP Client
The Network Time Protocol (NTP) is used to synchronize the time of a computer client or server to
another server or reference time source, such as a radio or satellite receiver. It provides client time accuracies within half a second on LANs and WANs relative to a primary server synchronized to Universal
Coordinated Time (UTC) (via a Global Positioning Service receiver, for example).
OSPFv2/OSPFv3
Open Shortest Path First version 3 (OSPFv3) is now available with the 6.1.3.R01 release on the
OmniSwitch 6850 Series and OmniSwitch 9000 Series. OSPFv3 is an extension of OSPF version 2
(OSPFv2) that provides support for networks using the IPv6 protocol. OSPFv2 is for IPv4 networks.
Both versions of OSPF are shortest path first (SPF), or link-state, protocols for IP networks. Also considered interior gateway protocols (IGP), both versions distribute routing information between routers in a
single Autonomous System (AS). OSPF chooses the least-cost path as the best path. OSPF is suitable for
complex networks with a large number of routers by providing faster convergence, loop free routing, and
equal-cost multi-path routing where packets to a single destination can be sent to more than one interface
simultaneously.
The following features are also supported by both versions of OSPF:
• Graceful (Hitless) Support During Failover, which is the time period between the restart and the rees-
tablishment of adjacencies after a planned (e.g., the users performs the takeover) or unplanned (e.g., the
primary management module unexpectedly fails) failover.
• OSPF adjacencies over non broadcast links.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 27
Software Supported
Partitioned Switch Management
A user account includes a login name, password, and user privileges. The privileges determine whether the
user has read or write access to the switch, and which command domains and command families the user
is authorized to execute on the switch. The privileges are sometimes referred to as authorization; the
designation of particular command families or domains for user access is sometimes referred to as partitioned management.
Available command domains and families are listed in the following table:
Domain
Corresponding Families
domain-admin
file telnet dshell debug
domain-system
system aip snmp rmon web config
domain-physical
chassis module interface pmm health
domain-network
ip rip ospf vrrp ip-routing ipx ipmr ipms
domain-layer2
802.1q
vlan bridge stp
linkagg ip-helper
domain-service
dns
domain-policy
qos policy slb
domain-security
session avlan aaa
Per-VLAN DHCP Relay
The OmniSwitch 9000 allows you to configure multiple DHCP relay (ip helper) addresses on a per-vlan
basis. For the Per-VLAN service, identify the number of the VLAN that makes the relay request. You may
identify one or more server IP addresses to which DHCP packets will be sent from the specified VLAN.
Both standard and per VLAN modes are supported.
Policy Server Management
Policy servers use Lightweight Directory Access Protocol (LDAP) to store policies that are configured
through Alcatel-Lucent’s PolicyView network management application. PolicyView is an OmniVista
application that runs on an attached workstation.
The Lightweight Directory Access Protocol (LDAP) is a standard directory server protocol. The LDAP
policy server client in the switch is based on RFC 2251. Currently, PolicyView is supported for policy
management.
Policy Based Routing (Permanent Mode)
Policy Based Routing may be used to redirect traffic to a particular gateway based on source or destination IP address, source or destination network group, source or destination TCP/UDP port, a service or
service group, IP protocol, or built-in source port group.
Traffic may be redirected to a particular gateway regardless of what routes are listed in the routing table.
Note that the gateway address does not have to be on a directly connected VLAN; the address may be on
any network that is learned by the switch.
Policy Based Routing is supported on the OmniSwitch 9000 Series. The 6.1.3.R01 release provides
support for this feature on the OmniSwitch 6850 Series.
page 28
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Software Supported
Port Mapping
Port Mapping is a security feature that controls peer users from communicating with each other. A Port
Mapping session comprises a session ID and a set of user ports and/or a set of network ports. User ports
within a session cannot communicate with each other and can only communicate via network ports. In a
Port Mapping session with user port set A and network port set B, ports in set A can only communicate
with ports in set B. If set B is empty, ports in set A can communicate with rest of the ports in the system.
A port mapping session can be configured in unidirectional or bidirectional mode. In the unidirectional
mode, the network ports can communicate with each other within the same session. In the bidirectional
mode, the network ports cannot communicate with each other. Network ports of a unidirectional port
mapping session can be shared with other unidirectional sessions, but cannot be shared with any sessions
configured in bidirectional mode. Network Ports of different sessions can communicate with each other.
Port Mapping is supported on the OmniSwitch 6800 Series and OmniSwitch 6850 Series. The 6.1.3.R01
release provides support for this feature on the OmniSwitch 9000 Series.
Port Mirroring
When Port Mirroring is enabled, the active “mirrored” port transmits and receives network traffic
normally, and the “mirroring” port receives a copy of all transmit and receive traffic to the active port.
You can connect an RMON probe or network analysis device to the mirroring port to see an exact duplication of traffic on the mirrored port without disrupting network traffic to and from the mirrored port.
Only one Port Mirroring session is supported. That session can be configured to a “N-to-1” session where
“N” can be a number from 1 to 24 (OS6800) or 1 to 128 (OS6850/OS9000) anywhere on the stack. In
other words, you can configure up to 24 or 128 source ports for a single destination port in a session on a
stack. You cannot configure port mirroring and port monitoring on the same NI module.
Power over Ethernet (PoE)
The Power over Ethernet (PoE) software is supported on the OS6850-P24, OS6850-P24X, OS6850-P48,
and OS6850-P48X stackable switches. The 6.1.3.R01 release provides support for this feature on the
OmniSwitch 9000 Series. PoE provides inline power directly from the switch’s Ethernet ports. From these
RJ-45 ports the devices receive both electrical power and data flow. PoE detects power based on PSE
devices and not on class.
PoE supports both IEEE 802.3af and non-IEEE 802.3af standards. The default inline power allotted for
each port is 15400 Milliwatts. The minimum inline power allotted for a port is 3000 Milliwatts and the
maximum is 16000 Milliwatts (OS6850) and 18000 Milliwatts (OS9000).
The maximum PoE power that a 510w power-supply (OS6850/OS9600) can provide is approximately 380
watts. A 360w power-supply (OS6850/OS9600) can provide approximately 230 watts of PoE power. The
OS-IP-Shelf power supplies (OS9000) can provide approximately 600 watts of PoE power. The OS-IPShelf supports up to four power supplies, so a total of approximately 2400 watts is possible.
The redundant power supply for PoE is only for backup. If the primary power supply fails, then PoE can
switch over seamlessly to the backup power supply.
Quality of Service (QoS)
Alcatel-Lucent’s QoS software provides a way to manipulate flows coming through the switch based on
user-configured policies. The flow manipulation (generally referred to as Quality of Service or QoS) may
be as simple as allowing/denying traffic, or as complicated as remapping 802.1p bits from a Layer 2
network to ToS values in a Layer 3 network. QoS can support up to 2048 policies and it is hardware-based
on the first packet. OmniSwitch 9000 switches truly support 8 queues per port.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 29
Software Supported
QoS is implemented on the switch through the use of policies, created on the switch or stored in PolicyView.
While policies may be used in many different types of network scenarios, there are several typical types:
• Basic QoS—includes traffic prioritization and bandwidth shaping
• 802.1p/ToS/DSCP—includes policies for marking and mapping
• Policy Based Routing (PBR)—includes policies for redirecting routed traffic
• Access Control Lists (ACLs)—ACLs are a specific type of QoS policy used for Layer 2, Layer 3/4,
and multicast filtering.
Note. NAT is not supported.
Redirect Policies (Port and Link Aggregate)
Two policy action commands are available for configuring QoS redirection policies: policy action redirect port and policy action redirect linkagg. A redirection policy sends traffic that matches the policy to
a specific port or link aggregate instead of the originally intended destination. This type of policy may use
any condition; the policy action determines which port or link aggregate to which the traffic is sent.
Redirect policies are supported on the OmniSwitch 9000 Series. The 6.1.3.R01 release provides support
for this feature on the OmniSwitch 6850 Series.
RMON
Remote Network Monitoring (RMON) is an SNMP protocol used to manage networks remotely. RMON
probes can be used to collect, interpret, and forward statistical data about network traffic from designated
active ports in a LAN segment to an NMS (Network Management System) application for monitoring and
analyzing without negatively impacting network performance. RMON software is fully integrated in the
software to acquire statistical information.
This feature supports basic RMON 4 group implementation in compliance with RFC 2819, including the
Ethernet Statistics, History (Control & Statistics), Alarms, and Events groups.
Router Discovery Protocol (RDP)
The Router Discovery Protocol (RDP) is an extension of ICMP that allows end hosts to discover routers
on their networks. The implementation of RDP supports the router requirements as defined in RFC 1256.
Using RDP, hosts attached to multicast or broadcast networks send solicitation messages when they start
up. Routers respond to solicitation messages with an advertisement message that contains the router IP
addresses. In addition, routers send advertisement messages when their RDP interface becomes active and
then subsequently at random intervals.
Routing Protocol Preference
OmniSwitch 9000 switches support provide configurable weight for each routing protocol (including static
routes) to control which entry to prefer when two entries exist from different sources. By default, local
routes always have precedence. In the CLI use the ip route-pref command to configure the routing preference.
page 30
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Software Supported
Secure Copy (SCP)
The scp CLI command is now available for copying files in a secure manner between hosts on the
network. The scp utility performs encrypted data transfers using the Secure Shell (SSH) protocol. In addition, scp uses available SSH authentication and security features, such as prompting for a password if one
is required.
Secure Shell (SSH)
The Secure Shell feature provides a secure mechanism that allows you to log in to a remote switch, to
execute commands on a remote device, and to move files from one device to another. Secure Shell
provides secure, encrypted communications even when your transmission is between two untrusted hosts
or over an unsecure network.
The OmniSwitch includes both client and server components of the Secure Shell interface and the Secure
Shell FTP file transfer protocol. SFTP is a subsystem of the Secure Shell protocol. All Secure Shell FTP
data are encrypted through a Secure Shell channel.
When used as an SSH Server, the following SSH Software is supported on the indicated operating
systems:
SSH Software
Supported Operating Systems
OpenSSH
Sun Solaris, Mac OSX, Linux Red Hat
F-Secure
Sun Solaris, Win 2000, Win XP
SSH-Communication
Sun Solaris, Win 2000, Win XP, Linux Red Hat
PuTTY
Win 2000, Win XP
MAC-SSH
Mac OSX
When used as an SSH Client, the following SSH Software is supported on the indicated operating
systems:
SSH Software
Supported Operating Systems
OpenSSH
Sun Solaris, Linux Red Hat, AOS
F-Secure
Sun Solaris, Win 2000
SSH-Communication
Sun Solaris, Win 2000, Win XP, Linux Red Hat
Secure Shell (SSH) Public Key Authentication
DSA public key authentication is supported when using PuTTY SSH software to generate the private and
public key for the client and to access the switch. It is now possible to enforce the use of public key
authentication only on the switch. By default, both password and public key authentication are allowed.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 31
Software Supported
Server Load Balancing (SLB)
Alcatel’s Server Load Balancing (SLB) software provides a method to logically manage a group of physical servers sharing the same content (known as a server farm) as one large virtual server (known as an SLB
cluster). SLB clusters are identified and accessed at Layer 3 by the use of Virtual IP (VIP) addresses or a
QoS policy condition. OmniSwitch 6850/9000 switches operate at wire speed to process client requests
addressed to the VIP of an SLB cluster and send them to the physical servers within the cluster.
Using SLB clusters can provide cost savings (costly hardware upgrades can be delayed or avoided), scalability (as the demands on your server farm grow you can add additional physical servers), reliability (if
one physical server goes down the remaining servers can handle the remaining workload), and flexibility
(you can tailor workload requirements individually to servers within a cluster).
The 6.1.3.R01 release provides support for this feature on the OmniSwitch 6850 Series and OmniSwitch
9000 Series.
sFlow
sFlow is a network monitoring technology that gives visibility to the activity of the network, by providing
network usage information. It provides the data required to effectively control and manage the network
usage. sFlow is a sampling technology that meets the requirements for a network traffic monitoring solution.
sFlow is a sampling technology embedded within switches/routers. It provides the ability to monitor the
traffic flows. It requires an sFlow agent software process running as part of the switch software and an
sFlow collector, which receives and analyses the monitored data. The sFlow collector makes use of SNMP
to communicate with an sFlow agent in order to configure sFlow monitoring on the device (switch).
Smart Continuous Switching - OmniSwitch 6800/OmniSwitch 6850
In stacked configurations, one switch is designated as the primary “management module” for the stack.
Because the stack can be thought of as a virtual chassis, the role of this primary management switch is to
monitor and manage the functions of the entire stack.
Similar to chassis-based switches, the stack also includes a secondary, or backup, management module. A
stack’s secondary switch immediately takes over management functions in the event of a primary switch
failure.
All switches in the stack, besides the primary and secondary switch, are considered idle or in pass-through.
Idle switches act like Network Interface (NI) modules in chassis-based switches.
The stack provides support for all idle switches during primary switch failover. In other words, if the
primary switch in the stack fails or goes offline for any reason, all idle switches will continue data transmission during the secondary switch’s takeover process.
Smart Continuous Switching - OmniSwitch 9000
Each OS9000 CMM module contains hardware and software elements to provide management functions
for the OS9000 system. The OS9000 CMM module also contains the switch fabric for the OS9000 system.
User data flowing from one NI module to another passes through the switch fabric.
The OS9700 will operate with one or two CMM modules installed. The OS9600 operates with one CMM.
If there are two CMM modules in an OS9700, one management processor is considered “primary” and is
actively managing the system. The other management processor is considered “secondary” and remains
ready to quickly take over management in the event of hardware or software failure on the primary. In the
event of a failure, the two processors exchange roles and the secondary takes over as primary.
page 32
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Software Supported
The switch fabric on the CMM operates independently of the management processor. If there are two
CMM modules installed in an OS9700, both fabric modules are normally active. Two CMM modules
must be installed in the OS9700 to provide full fabric capacity. However, note that only the one CMM
module in the OS9600 provides full fabric capacity.
If there is one CMM module installed in an OS9700, then there is a single management feature and performance as a dual CMM system, but there is no “secondary” CMM. Hardware or software failures in the
CMM will result in a system reboot. The System fabric capacity on an OS9700 is one half of the fabric
capacity of a dual CMM system.
SNMP
The Simple Network Management Protocol (SNMP) is an application-layer protocol that allows communication between SNMP managers and SNMP agents on an IP network. Network administrators use
SNMP to monitor network performance and to solve network problems. SNMP provides an industry standard communications model used by network administrators to manage and monitor their network
devices. OmniSwitch 9000 switches support SNMPv1, SNMPv2, and SNMPv3.
Source Learning
Source Learning builds and maintains the MAC address table on each switch. New MAC address table
entries are created in one of two ways: they are dynamically learned or statically assigned. Dynamically
learned MAC addresses are those that are obtained by the switch when source learning examines data
packets and records the source address and the port and VLAN it was learned on. Static MAC addresses
are user defined addresses that are statically assigned to a port and VLAN.
In addition, Source Learning also tracks MAC address age and removes addresses from the MAC address
table that have aged beyond the configurable aging timer value.
Accessing MAC Address Table entries is useful for managing traffic flow and troubleshooting network
device connectivity problems.
There are two types of source learning modes currently available: software and hardware. The software
mode performs all source learning using switch software. The hardware mode takes advantage of hardware resources that are now available to perform source learning tasks. At the present time, it is possible
to select the mode that is active for the chassis and/or a given set of ports.
By default, hardware source learning mode is active for the switch. The exception to this is that hardware
source learning is not supported on mobile or Learned Port Security (LPS) ports. As a result, only software source learning is performed on these types of ports.
Software Rollback
The directory structure inherent in an OmniSwitch switch allows for a switch to return to a previous, more
reliable version of image or configuration files.
Changes made to the configuration file may alter switch functionality. These changes are not saved unless
explicitly done so by the user. If the switch reboots before the configuration file is saved, changes made to
the configuration file prior to the reboot are lost.
Likewise, new image files should be placed in the working (non-certified) directory first. New image or
configuration files can be tested to decide whether they are reliable. Should the configuration or images
files prove to be less reliable than their older counterparts in the certified directory, then the switch can be
rebooted from the certified directory, and “rolled back” to an earlier version.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 33
Software Supported
Once the contents of the working directory are established as good files, then these files can be saved to
the certified directory and used as the most reliable software to which the switch can be rolled back to in
an emergency situation.
Spanning Tree
In addition to the Q2005 version of MSTP, the Alcatel Spanning Tree implementation also provides
support for the 802.1w Rapid Spanning Tree Algorithm and Protocol (RSTP) and the 802.1D Spanning
Tree Algorithm and Protocol (STP). All three supported protocols ensure that there is always only one data
path between any two switches for a given Spanning Tree instance to prevent network loops.
Q2005 (MSTP) is only available when the flat mode is active for the switch. The flat mode applies a single
spanning tree instance across all VLAN port connections on a switch. MSTP allows the configuration of
Multiple Spanning Tree Instances (MSTIs) in addition to the CST instance. Each MSTI is mapped to a set
of VLANs. As a result, flat mode can now support the forwarding of VLAN traffic over separate data
paths.
802.1D STP and 802.1w RSTP are available in both the flat and 1x1 mode. However, when using 802.1D
or 802.1w in the flat mode, the single spanning tree instance per switch algorithm applies. Note that
802.1w is now the default Spanning Tree protocol for the switch regardless of which mode is active. This
default value will apply to future releases as well.
Syslog to Multiple Hosts
You can now send syslog files to multiple hosts, up to a maximum of four servers.
Switch Logging
The Switch Logging feature is designed to provide a high-level event logging mechanism that can be
useful in maintaining and servicing the switch. Switch Logging uses a formatted string mechanism to
process log requests from applications. When a log request is received, Switch Logging verifies whether
the Severity Level included with the request is less than or equal to the Severity Level stored for the appropriate Application ID. If it is, a log message is generated using the formatting specified by the log request
and placed on the Switch Log Queue, and Switch Logging returns control back to the calling application.
Otherwise, the request is discarded. The default output device is the log file located in the Flash File
System. Other output devices can be configured via Command Line Interface. All log records generated
are copied to all configured output devices.
Command Line Interface can be used to display and configure Switch Logging information. Log information can be helpful in resolving configuration or authentication issues, as well as general errors.
Text File Configuration
The text file configuration feature allows you to configure the switch using an ASCII-based text file. You
may type CLI commands directly into a text document to create a configuration file. This file resides in the
switch’s file system. You can create configuration files in the following ways.
• You may create, edit and view a file using a standard text editor (such as Microsoft NotePad) on a
workstation. The resulting configuration file is then uploaded to the switch.
• You can invoke the switch’s CLI snapshot command to capture the switch’s current configuration into
a text file.
• You can use the switch’s text editor to create or make changes to a configuration file.
page 34
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Software Supported
User Definable Loopback Interface
Loopback0 is the name assigned to an IP interface to identify a consistent address for network management purposes. The Loopback0 interface is not bound to any VLAN, therefore it always remains operationally active. This differs from other IP interfaces, such that if there are no active ports in the VLAN, all
IP interfaces associated with that VLAN are not active. In addition, the Loopback0 interface provides a
unique IP address for the switch that is easily identifiable to network management applications.
VLAN Range Support
Up to 4094 VLANs for Flat Spanning Tree mode and 252 VLANs for 1x1 Spanning Tree mode are
supported. In addition, it is now possible on the OmniSwitch 6800/6850/9000 to specify a range of VLAN
IDs when creating or deleting VLANs and/or configuring VLAN parameters, such as Spanning Tree
bridge values.
VLAN Stacking and Translation
VLAN Stacking provides a mechanism for tunneling multiple customer VLANs (CVLAN) through a
service provider network over the Ethernet Metropolitan Area Network (EMAN). The service provider
network uses one or more service provider VLANs (SVLAN) by appending an 802.1Q double tag or
VLAN Translation on a customer port that contains the customer’s assigned tunnel ID. This traffic is then
encapsulated into the tunnel and transmitted through the service provider network. It is received on
another Provider Edge (PE) that has the same tunnel ID. This feature enables service providers to provide
their customers with Transparent LAN Services (TLS). This service is multipoint in nature so as to
support multiple customer sites or networks distributed over the edges of a service provider network.
VRRPv2/VRRPv3
The Virtual Router Redundancy Protocol version 3 (VRRPv3) is now available with the 6.1.3.R01 release.
This implementation is based on the latest Internet-Draft for VRRP for IPv6. VRRP version 2 (VRRPv2)
is based on RFC 2338.
Similar to VRRPv2, VRRPv3 is a standard router redundancy protocol that provides redundancy by eliminating the single point of failure inherent in a default route environment. The VRRPv3 router, which
controls the IPv6 address associated with a virtual router is called the master router, and is responsible for
forwarding virtual router advertisements. If the master router becomes unavailable, the highest priority
backup router will transition to the master state.
Both versions of VRRP allow routers on a LAN to back up a static default route with a virtual router.
VRRP dynamically assigns responsibility for a virtual router to a physical router (VRRP router) on the
LAN. The virtual router is associated with an IP address (or set of IP addresses) on the LAN. A virtual
router master is elected to forward packets for the virtual router’s IP address. If the master router becomes
unavailable, the highest priority backup router will transition to the master state.
Note. Authentication is not supported.
In addition, both versions support VRRP Tracking. A virtual router’s priority may be conditionally modified to prevent another router from taking over as master. Tracking policies are used to conditionally
modify the priority setting whenever an ip interface, slot/port, and/or IP address associated with a virtual
router goes down.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 35
Software Supported
Web-Based Management (WebView)
The switch can be monitored and configured using WebView, Alcatel-Lucent’s web-based device
management tool. The WebView application is embedded in the switch and is accessible via the following
web browsers:
• Internet Explorer 6.0 and later for Windows NT, 2000, XP, 2003
• Netscape 7.1 for Windows NT, 2000, XP
• Netscape 7.0 for Solaris SunOS 5.8
WebView contains modules for configuring all software features in the switch. Configuration and monitoring pages include context-sensitive on-line help.
page 36
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Supported Traps
Supported Traps
The following traps are supported in 6.1.3.R01:
No. Trap Name
Platforms Description
0
coldStart
all
The SNMP agent in the switch is reinitiating and its configuration may
have been altered.
1
warmStart
all
The SNMP agent in the switch is reinitiating itself and its configuration is
unaltered.
2
linkDown
all
The SNMP agent in the switch recognizes a failure in one of the communications links configured for the
switch.
3
linkUp
all
The SNMP agent in the switch recognizes that one of the communications
links configured for the switch has
come up.
4
authenticationFailure
all
The SNMP agent in the switch has
received a protocol message that is
not properly authenticated.
5
entConfigChange
all
An entConfigChange notification is
generated when a conceptual row is
created, modified, or deleted in one of
the entity tables.
6
aipAMAPStatusTrap
all
The status of the Alcatel Mapping
Adjacency Protocol (AMAP) port
changed.
7
aipGMAPConflictTrap
—
This trap is not supported.
8
policyEventNotification
all
The switch notifies the NMS when a
significant event happens that
involves the policy manager.
9
chassisTrapsStr
all
A software trouble report (STR) was
sent by an application encountering a
problem during its execution.
10
chassisTrapsAlert
all
A notification that some change has
occurred in the chassis.
11
chassisTrapsStateChange
all
An NI status change was detected.
12
chassisTrapsMacOverlap
all
A MAC range overlap was found in
the backplane eeprom.
13
vrrpTrapNewMaster
all
The SNMP agent has transferred from
the backup state to the master state.
14
vrrpTrapAuthFailure
—
This trap is not supported.
15
healthMonDeviceTrap
all
Indicates a device-level threshold was
crossed.
16
healthMonModuleTrap
all
Indicates a module-level threshold
was crossed.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 37
Supported Traps
No. Trap Name
Platforms Description
17
healthMonPortTrap
all
Indicates a port-level threshold was
crossed.
18
bgpEstablished
all
The BGP routing protocol has entered
the established state.
19
bgpBackwardTransition
all
This trap is generated when the BGP
router port has moved from a more
active to a less active state.
20
esmDrvTrapDropsLink
all
This trap is sent when the Ethernet
code drops the link because of excessive errors.
21
pimNeighborLoss
all
Signifies the loss of adjacency with a
neighbor device. This trap is generated when the neighbor time expires
and the switch has no other neighbors
on the same interface with a lower IP
address than itself.
22
dvmrpNeighborLoss
all
A 2-way adjacency relationship with
a neighbor has been lost. This trap is
generated when the neighbor state
changes from “active” to “one-way,”
“ignoring” or “down.” The trap is sent
only when the switch has no other
neighbors on the same interface with
a lower IP address than itself.
23
dvmrpNeighborNotPruning
all
A non-pruning neighbor has been
detected in an implementation-dependent manner. This trap is generated at
most once per generation ID of the
neighbor. For example, it should be
generated at the time a neighbor is
first heard from if the prune bit is not
set. It should also be generated if the
local system has the ability to tell that
a neighbor which sets the prune bit is
not pruning any branches over an
extended period of time. The trap
should be generated if the router has
no other neighbors on the same interface with a lower IP address than
itself.
24
risingAlarm
all
An Ethernet statistical variable has
exceeded its rising threshold. The
variable’s rising threshold and
whether it will issue an SNMP trap
for this condition are configured by an
NMS station running RMON.
25
fallingAlarm
all
An Ethernet statistical variable has
dipped below its falling threshold.
The variable’s falling threshold and
whether it will issue an SNMP trap
for this condition are configured by an
NMS station running RMON.
page 38
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Supported Traps
No. Trap Name
Platforms Description
26
stpNewRoot
all
Sent by a bridge that became the new
root of the spanning tree.
27
stpRootPortChange
all
A root port has changed for a spanning tree bridge. The root port is the
port that offers the lowest cost path
from this bridge to the root bridge.
28
mirrorConfigError
all
The mirroring configuration failed on
an NI. This trap is sent when any NI
fails to configure mirroring. Due to
this error, port mirroring session will
be terminated.
29
mirrorUnlikeNi
all
The mirroring configuration is deleted
due to the swapping of different NI
board type. The Port Mirroring session which was active on a slot cannot
continue with the insertion of different NI type in the same slot.
30
slPCAMStatusTrap
all
The trap status of the Layer 2 pesudoCAM for this NI.
31
unused
—
32
unused
—
33
slbTrapOperStatus
—
A change occurred in the operational
status of the server load balancing
entity.
34
ifMauJabberTrap
all
This trap is sent whenever a managed
interface MAU enters the jabber state.
35
sessionAuthenticationTrap
all
An authentication failure trap is sent
each time a user authentication is
refused.
36
trapAbsorptionTrap
all
The absorption trap is sent when a
trap has been absorbed at least once.
37
alaStackMgrDuplicateSlotTrap
—
Two or more slots claim to have the
same slot number.
38
alaStackMgrNeighborChangeTrap
—
Indicates whether or not the stack is in
loop.
39
alaStackMgrRoleChangeTrap
—
Indicates that a new primary or secondary stack is elected.
40
lpsViolationTrap
all
A Learned Port Security (LPS) violation has occurred.
41
alaDoSTrap
all
Indicates that the sending agent has
received a Denial of Service (DoS)
attack.
42
gmBindRuleViolation
all
Occurs whenever a binding rule
which has been configured gets violated.
43
unused
—
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 39
Supported Traps
No. Trap Name
Platforms Description
44
unused
—
45
unused
—
46
unused
—
47
pethPsePortOnOff
—
Indicates if power inline port is or is
not delivering power to the a power
inline device.
48
pethPsePortPowerMaintenanceStatus
—
Indicates the status of the power
maintenance signature for inline
power.
49
pethMainPowerUsageOn
—
Indicates that the power inline usage
is above the threshold.
50
pethMainPowerUsageOff
—
Indicates that the power inline usage
is below the threshold.
51
ospfNbrStateChange
all
Indicates a state change of the neighbor relationship.
52
ospfVirtNbrStateChange
all
Indicates a state change of the virtual
neighbor relationship.
53
httpServerDoSAttackTrap
all
This trap is sent to management station(s) when the HTTP server is under
Denial of Service attack. The HTTP
and HTTPS connections are sampled
at a 15 second interval. This trap is
sent every 1 minute while the HTTP
server detects it is under attack.
54
alaStackMgrDuplicateRoleTrap
—
The element identified by alaStackMgrSlotNINumber detected the presence of two elements with the same
primary or secondary role as specified
by alaStackMgrChasRole on the
stack.
55
alaStackMgrClearedSlotTrap
—
The element identified by alaStackMgrSlotNINumber will enter the pass
through mode because its operational
slot was cleared with immediate
effect.
56
alaStackMgrOutOfSlotsTrap
—
One element of the stack will enter
the pass through mode because there
are no slot numbers available to be
assigned to this element.
57
alaStackMgrOutOfTokensTrap
—
The element identified by alaStackMgrSlotNINumber will enter the pass
through mode because there are no
tokens available to be assigned to this
element.
58
alaStackMgrOutOfPassThruSlotsTrap
—
There are no pass through slots available to be assigned to an element that
is supposed to enter the pass through
mode.
page 40
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Supported Traps
No. Trap Name
Platforms Description
59
gmHwVlanRuleTableOverloadAlert
all
An overload trap occurs whenever a
new entry to the hardware VLAN rule
table gets dropped due to the overload
of the table.
60
lnkaggAggUp
all
Indicates the link aggregate is active.
This trap is sent when any one port of
the link aggregate group goes into the
attached state.
61
lnkaggAggDown
all
Indicates the link aggregate is not
active. This trap is sent when all ports
of the link aggregate group are no
longer in the attached state.
62
lnkaggPortJoin
all
This trap is sent when any given port
of the link aggregate group goes to the
attached state.
63
lnkaggPortLeave
all
This trap is sent when any given port
detaches from the link aggregate
group.
64
lnkaggPortRemove
all
This trap is sent when any given port
of the link aggregate group is
removed due to an invalid configuration.
65
pktDrop
all
The pktDrop trap indicates that the
sending agent has dropped certain
packets (to blocked IP ports, from
spoofed addresses, etc.).
66
monitorFileWritten
all
A File Written Trap is sent when the
amount of data requested by the user
has been written by the port monitoring instance.
67
alaVrrp3TrapProtoError
all
Indicates that a TTL, checksum, or
version error was encountered upon
receipt of a VRRP advertisement.
68
alaVrrp3TrapNewMaster
all
The SNMP agent has transferred from
the backup state to the master state.
69
gmHwMixModeSubnetRuleTableOverloadAlert
—
This trap is not supported in the current release.
70
pethPwrSupplyConflict
—
This trap is not supported in the current release.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 41
Unsupported Software Features
Unsupported Software Features
CLI commands and Web Management options maybe available in the switch software for the following
features. These features are not supported:
Feature
Platform
Software Package
OSPF Database Overflow (RFC 1765)
all
base
Flow Control 802.3x
all
base
Unsupported CLI Commands
The following CLI commands are not supported in this release of the software:
Software Feature
BGP
Unsupported CLI Commands
ip bgp redist-filter status
ip bgp redist-filter
ip bgp redist-filter community
ip bgp redist-filter local-preference
ip bgp redist-filter metric
ip bgp redist-filter effect
ip bgp redist-filter subnets
Chassis Mac Server
mac-range local
mac-range duplicate-eeprom
mac-range allocate-local-only
show mac-range status
Command Line Interface (CLI) 10 gig slot [slot] phy-a|phy-b
Flow Control
flow
flow wait time
interfaces flow
Ethernet Interfaces
interfaces long
interfaces runt
interfaces runtsize
Hot Swap
reload ni [slot] #
[no] power ni all
NTP
no ntp server all
OSPF
ip ospf redist status
ip ospf redist
ip ospf redist metric
ip ospf redist metric-type
ip ospf redist-filter
ip ospf redist-filter effect
ip ospf redist-filter metric
ip ospf redist-filter route-tag
ip ospf redist-filter redist-control
QoS
qos classify fragments
qos flow timeout
show policy classify destination interface type
show policy classify source interface type
page 42
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Unsupported CLI Commands
Software Feature
Unsupported CLI Commands
RIP
ip rip redist status
ip rip redist
ip rip redist metric
ip rip redist-filter
ip rip redist-filter effect
ip rip redist-filter metric
ip rip redist-filter route-tag
ip rip redist-filter redist-control
VLANs
vlan router mac multiple enable|disable
vlan binding mac-port-protocol
vlan binding mac-ip
vlan binding ip-port
Chassis Supervision
show fabric
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 43
Unsupported MIBs
Unsupported MIBs
The following MIBs are not supported in this release of the software:
MIB
Feature
Quality of Service (QoS)
IETF_P_BRIDGE
Flow Control
AlcatelIND1Port
Unsupported MIB Variables
MIB Name
Unsupported MIB variables
AlcatelIND1AAA
aaauProfile
AlcatelIND1Bgp
alaBgpGlobal
alaBgpPeerTable
alaBgpAggrTable
alaBgpNetworkTable
alaBgpRedistRouteTable
alaBgpRouteTable
alaBgpPathTable
alaBgpDampTable
alaBgpRouteMapTable
alaBgpAspathMatchListTable
alaBgpAspathPriMatchListTable
alaBgpPrefixMatchListTable
alaBgpCommunityMatchListTable
alaBgpCommunityPriMatchListTable
alaBgpDebugTable
AlcatelIND1Dot1Q
qPortVlanForceTagInternal
AlcatelIND1GroupMobility
vPortIpBRuleTable
vMacIpBRuleTable
vMacPortProtoBRuleTable
vCustomRuleTable
AlcatelIND1Health
healthDeviceTemperatureCmmCpuLatest
healthDeviceTemperatureCmmCpu1MinAvg
healthDeviceTemperatureCmmCpu1HrAvg
healthDeviceTemperatureCmmCpu1HrMax
AlcatelIND1Ipms
alaIpmsForwardSrcIpAddr
alaIpmsForwardSrcIfIndex
AlcatelIND1LAG
alclnkaggAggEniActivate
alclnkaggSlotTable
AlcatelIND1Pcam
alcatelIND1PCAMMIBObjects
alaCoroL3HrePerModeTable
alaCoroL3HrePerCoronadoStats
Table
alaCoroL3HreChangeTable
AlcatelIND1Port
esmPortCfgLongEnable
esmPortCfgRuntEnable
esmPortCfgRuntSize
esmPortPauseSlotTime
esmPortCfgFLow
page 44
alcether10GigTable
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Unsupported MIBs
MIB Name
Unsupported MIB variables
AlcatelIND1QoS
alaQoSPortPdiTable
alaQoSSlotPcamTable
alaQoSPortProtocolTable
alaQoSSlotProtocolTable
alaQoSSlotDscpTable
alaQoSRuleReflexive
alaQoSAppliedRuleReflexive
alaQoSActionSourceRewriteIpAddr
alaQoSActionSourceRewriteIpAddrStatus
alaQoSActionSourceRewriteIpMask
alaQoSActionTable alaQoSActionSourceRewriteNetworkGroup
alaQoSActionTable alaQoSActionSourceRewriteNetworkGroupStatus
alaQoSActionTable alaQoSActionDestinationRewriteIpAddr
alaQoSActionTable alaQoSActionDestinationRewriteIpAddrStatus
alaQoSActionTable alaQoSActionDestinationRewriteIpMask
alaQoSActionTable alaQoSActionDestinationRewriteNetworkGroup
alaQoSActionTable alaQoSActionDestinationRewriteNetworkGroupStatus
alaQoSActionTable alaQoSActionLoadBalanceGroup
alaQoSActionTable alaQoSActionLoadBalanceGroupStatus
alaQoSActionTable alaQoSActionPermanentGatewayIpAddr
alaQoSActionTable alaQoSActionPermanentGatewayIpAddrStatus
alaQoSActionTable alaQoSActionAlternateGatewayIpAddr
alaQoSActionAlternateGatewayIpAddrStatus
alaQoSAppliedActionSourceRewriteIpAddr
alaQoSAppliedActionSourceRewriteIpAddrStatus
alaQoSAppliedActionSourceRewriteIpMask
alaQoSAppliedActionSourceRewriteNetworkGroup
alaQoSAppliedActionSourceRewriteNetworkGroupStatus
alaQoSAppliedActionDestinationRewriteIpAddr
alaQoSAppliedActionDestinationRewriteIpAddrStatus
alaQoSAppliedActionDestinationRewriteIpMask
alaQoSAppliedActionDestinationRewriteNetworkGroup
alaQoSAppliedActionDestinationRewriteNetworkGroupStatus
alaQoSAppliedActionLoadBalanceGroup
alaQoSAppliedActionLoadBalanceGroupStatus
alaQoSAppliedActionPermanentGatewayIpAddr
alaQoSAppliedActionPermanentGatewayIpAddrStatus
alaQoSAppliedActionAlternateGatewayIpAddr
alaQoSAppliedActionAlternateGatewayIpAddrStatus
alaQoSPortDefaultQueues
alaQoSPortAppliedDefaultQueues
alaQoSConfigNatTimeout
alaQoSConfigAppliedNatTimeout
alaQoSConfigReflexiveTimeout
alaQoSConfigAppliedReflfexiveTimeout
alaQoSConfigFragmentTimeout
alaQoSConfigAppliedFragmentTimeout
alaQoSConfigClassifyFragments
alaQoSConfigAppliedClassifyFragments
AlcatelIND1Slb
slbFeature
slbClusterTable
slbServerTableg
AlcatelIND1StackManager
alaStackMgrStatsTable
AlcatelIND1SystemService
systemUpdateStatusTable
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 45
Unsupported MIBs
MIB Name
Unsupported MIB variables
AlcatelIND1VlanManager
vlanIpxNet
vlanIpxEncap
vlanIpxRipSapMode
vlanIpxDelayTicks
vlanSetMultiRtrMacStatus
AlcatelIND1WebMgt
alaIND1WebMgtRFSConfigTable
alaIND1WebMgtHttpPort
alaIND1WebMgtHttpsPort
IEEE_802_1X
dot1xAuthDiagTable
dot1xAuthSessionStatsTable
dot1xSuppConfigTable
dot1xSuppStatsTable
IETF_BGP4
bgpRcvdPathAttrTable
bgp
bgpPeerTable
bgp4PathAttrTabl
IETF_BRIDGE
dot1dTpPortTable
dot1dStaticTable
IETF_ENTITY
entLogicalTable
entLPMappingTable
entAliasMappingTable
IETF_ETHERLIKE
dot3CollTable
dot3StatsSQETestErrors
dot3StatsInternalMacTransmitErrors
dot3StatsCarrierSenseErrors
dot3StatsInternalMacReceiveErrors
dot3StatsEtherChipSet
dot3StatsSymbolErrors
dot3ControlInUnknownOpcodes
IETF_IF
ifRcvAddressTable
ifTestTable
IETF_IP_FORWARD_MIB
ipForwardTable
IETF_IPMROUTE_STD
ipMrouteScopeNameTable
IETF_MAU (RFC 2668)
rpMauTable
rpJackTable
broadMauBasicTable
ifMauFalseCarriers
ifMauTypeList
ifMauAutoNegCapability
ifMauAutoNegCapAdvertised
ifMauAutoNegCapReceived
IETF_OSPF (RFC 1850)
ospfAreaRangeTable
IETF_OSPF_TRAP
ospfTrapControl
IETF-PIM
pimRPTable
page 46
vlanIpxStatus
vlanSetIpxRouterCount
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Unsupported MIBs
MIB Name
Unsupported MIB variables
IETF_P_BRIDGE
dot1dExtBase
dot1dPortCapabilitiesTable
dot1dPortPriorityTable
dot1dUserPriorityRegenTable
dot1dTrafficClassTable
dot1dPortOutboundAccessPriorityTable
dot1dPortGarpTable
dot1dPortGmrpTable
dot1dTpHCPortTable
dot1dTpPortOverflowTable
IETF_Q_BRIDGE (RFC 2674)
dot1qTpGroupTable
dot1qForwardAllTable
dot1qForwardUnregisteredTable
dot1qStaticMulticastTable
dot1qPortVlanStatisticsTable
dot1qPortVlanHCStatisticsTable
dot1qLearningConstraintsTable
IETF_RIPv2
rip2IfConfDomain
IETF_RMON
hostControlTable
hostTable
hostTimeTable
hostTopNControlTable
hostTopNTable
matrixControlTable
matrixSDTable
matrixDSTable
filterTable
channelTable
bufferControlTable
captureBufferTable
IETF_RS_232 (RFC 1659)
all synchronous and sdlc objects and tables
rs232SyncPortTable
IETF_SNMPv2
sysORTable
snmpTrap
sysORLastChange
IETF_SNMP_
COMMUNITY (RFC 2576)
snmpTargetAddrExtTable
IETF_SNMP_
NOTIFICATION (RFC 2576)
snmpNotifyTable
snmpNotifyFilterProfileTable
snmpNotifyFilterTable
IETF_SNMP_PROXY (RFC 2573)
snmpProxyTable
IETF_SNMP_TARGET (RFC 2573)
snmpTargetAddrTable
snmpTargetParamsTable
snmpTargetSpinLock
IETF_SNMP_USER_BASED_SM
(RFC 2574)
usmUser
IETF_SNMP_VIEW_BASED_ACM
(RFC 2575)
vasmMIBViews
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 47
Open Problem Reports and Feature Exceptions
Open Problem Reports and Feature Exceptions
The problems listed here include problems known at the time of the product’s release. Any problems not
discussed in this section should be brought to the attention of the Alcatel Technical Support organization
as soon as possible. Please contact customer support for updates on problem reports (PRs) where no
known workaround was available at the time of release.
Switch Management
Command Line Interface (CLI)
Problem Reports
PR 94134
Display of mobile ports by using the CLI “port range” command shows only the information about the
first port in the range on an OS9000.
Workaround: The current CLI output for mobile ports is line based, which does not allow for a practical
output of the port range. CLI command to display the mobile port should not use port range in the
command.
PR 95896
On an OS9000, when user permissions are changed, they do not have an immediate effect.
Workaround: The Session Manager polls the servers (through AAA) every 5 minutes for changes in
permissions. For an immediate impact, the administrator may remove the user from the switch to deprive
the user from certain permissions. This way the new permissions are current when the user logs on again.
By the same token, whenever an administrator decides to remove a user from the switch, he/she should
close all of the sessions to which the user is logged on.
PR 104635
Applying the non-hybrid version of the "interfaces" command to combo ports may result in an incorrect
configuration on an OS6850.
Workaround: Use the hybrid command for combo ports and use the regular (non-hybrid) command for
regular (non-combo) ports. For example, to set speed, use:
"interfaces <slot/port> hybrid fiber speed 100" for combo port.
"interfaces <slot/port> speed 100" for non-combo port.
PR 105168
When using the CLI "show ni" command, the display of XFP and SFP Model name and Description
displays the manufacturer's name. These fields should display the Model Name and device description.
Workaround: There is no known workaround at this time.
page 48
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Open Problem Reports and Feature Exceptions
PR 106811
Show interface slot/port "SFP/XFP" field output for a port having SFP plugged in cannot differentiate
between 100Fx and Bidirectional SFP & between Gig and CWDM SFP on an OS6850.
Workaround: There is no known workaround at this time.
RMON
Problem Reports
PR 87683
On an OS6800/OS9000, the RMON object etherStatsPkts65to127Octs and other similar objects from RFC
1757 contain TX and RX packet counts.
Workaround: There is no known workaround at this time.
PR 106919
On an OS6850, it is not possible to disable all RMON probes with a single SNMP command.
Workaround: When using SNMP, disable each probe individually.
sFlow
Problem Reports
PR 100009
On an OS6850, sFlowCpTable and sFlowFsTable return data of zero when an existing NI is removed or
powered down.
Workaround: There is no known workaround at this time.
PR 106130
The syntax for sflowReceiver and sFlowAgent returned by an OS9000 are not octets, but strings.
Workaround: There is no known workaround at this time.
PR 106878
When replacing an NI with a different type of NI, the sFlow configuration is not deleted correctly for the
ports which do not exist anymore.
Workaround: Delete the sFlow configuration before swapping different NI types.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 49
Open Problem Reports and Feature Exceptions
PR 107462
Sflow Datagrams do not use the EMP port on a OS9000 for IPV4 packets.
Workaround: Use the ports on the slots.
SNMP
Problem Reports
PR 82635
On an OS9000, there is no display of the number or the status of fan modules on an OmniSwitch via
SNMP or WebView.
Workaround: The number and status of fans can be displayed via the CLI show fan command only.
PR 105290
SNMP and the WebView DVMRP --> Routes page does not display Route Flags.
Workaround: Use the CLI show ip dvmrp route command to display Route Flags.
PR 107059
WebView and SNMP queries may show more ports than those displayed by the CLI command "show tcp
ports". The ports shown by the CLI command are valid.
Workaround: There is no known workaround at this time.
Web-Based Management (WebView)
Feature Exceptions
• WebView uses signed applets for the automatic IP reconfiguration. Those applets are signed using
VeriSign Certificates that expire every year. The certificate used for Internet Explorer and Netscape
expires every August. WebView users have to validate a warning indicating that the certificate used by
the applet has expired.
Problem Reports
PR 85135
In WebView > Health > LED page, the XFP1 and XFP2 LEDS are not displayed.
Workaround: XFP1 and XFP2 LEDs are for realtime packet traffic activity indications. For precise
reflection of the Rx/Tx activity, please refer to the corresponding Rx/Tx statistics all through the
WebView > Health pages.
page 50
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Open Problem Reports and Feature Exceptions
PR 89084
For Partition Management, giving a user routing protocols permission, but no IP permission results in
some table data not being displayed in the routing pages.
Workaround: Make sure a user gets permission for IP if the user is allowed to view any sort of IP routing protocol pages.
PR 96146
When viewing XOS adjacencies with multiple (2 or more) XoS devices connected to an OS9800 using
AOS WebView, will return unknown devices with build 4.4.4.188.B or greater which is correct behavior.
With earlier builds, the XOS adjacencies could be returned as an AOS device.
Workaround: There is no known workaround at this time.
PR 96274
On an OS9000, in WebView, Networking > IP Multicast > IPv4 > Switching > Configuration and
Networking > IP Multicast > IPv6 > Switching > Configuration pages show the actual configuration only,
but not the effective configuration of the system in parenthesis of the Status, Querying, Spoofing, and
Zapping.
Workaround: Use CLI to view the effective configuration.
PR 99581
The “scp” command is not supported in WebView.
Workaround: There is no known workaround at this time.
PR 101446
WebView > Physical > Adjacencies home page may show an "Applet Not Found" error when the Java
Virtual Machine SDK version is less than 1.2.
Workaround: Upgrade the Java Virtual Machine browser plug-in to a more recent version.
PR 105205
Reloading the switch using in and at times fails in WebView.
Workaround: Use the CLI for timed reloads.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 51
Open Problem Reports and Feature Exceptions
PR 100607
Accessing WebView > Security > Servers using Netscape, a window may appear asking for user name
selection on an OS6850 when adding a new server.
Workaround: Close the window and ignore. This window shown is not a WebView pop-up window but a
Netscape Form Manager window -- perhaps triggered by the "Server Name" label in the server add
windows and being treated as if it was a regular form asking for a user name.
PR 106670
On an OS9000, the ipxServTable returns the name of the IPX SAP. This is a variable length name. The
SNMP standard specifies that variable length strings should be preceded by a length byte. Some SNMP
applications could have problems with this.
Workaround: There is no known workaround at this time.
PR 106869
In WebView > Layer 2 > VLAN Mgmt > VLANs Add page, on an OS9000, an error message appears
whenever user enters VLAN ranges to be created that include more than 63 VLANs at once.
Workaround: Break desired VLAN range to create into smaller ranges (63 or less). (Note: This configuration setting limit occurs all throughout WebView.)
PR 106914
In WebView System > System Mgmt > Install page, on an OS9000, the file transfer fails in a dual NIC
(Network Interface Card) environment.
Workaround: Momentarily disable the network of the NIC not in use for the transfer and re-enable after
the transfer.
PR 107228
In WebView Physical > Health > Port Statistics page after clearing statistics, the last port of the slot may
display values other than zeroes on an OS9000.
Workaround: Please ignore non-zero values. (Any non-zero values are still being cleared when the page
is served.)
PR 107283
On an OS6850, when the supplicant fails the authentication with the external authentication server and if
user configured a "failed" vlan on the device classification policy, then AOS is going to locally authenticate the supplicant. In CLI, the "show" command will show as authenticated-L to indicate that the supplicant failed by the external authentication server but is locally authenticated to be authorized into the user
defined vlan. WebView is not showing this information and will only see as authenticated.
Workaround: Look at which vlan the supplicant is authorized and that can tell if the supplicant learned
on the user defined "failed" VLAN.
page 52
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Open Problem Reports and Feature Exceptions
PR 107284
Supplicant and non-supplicant are not displayed together on the same front page for 802.1x on an OS6850.
Workaround: View the non-supplicants on a different page - 802.1x non supplicant page. Display Issue
Only.
PR 107315
A duplicate slot in a stack of switches will not display the switch picture in the WebView > Chassis Mgmt
> Chassis homepage.
Workaround: Remove the duplicate slot.
PR 107529
WebView Networking > IP > BGP > Policy Route Maps Modify help "local Preference Value" hyperlink
is broken.
Workaround: There is no known workaround at this time.
PR 107530
When using the Netscape browser to access the WebView > Policy > Policy > Ports page after clicking on
either "Previous" or "Next" buttons and then clicking on the "Refresh" button, a "method not allowed"
error is displayed.
Workaround: Click on the "Ports" menu to refresh.
Layer 2
Autonegotiation
Problem Reports
PR 86826
On an OS6800/OS9000, when autonegotiation is disabled and speed is forced to 10 Mbps, copper ports
may confuse with a forced 100Mbps link partner and can detect a false link UP. Traffic may not pass in
this condition.
Workaround: Enable autonegotiation for this copper port and configure desired speed and duplex
settings.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 53
Open Problem Reports and Feature Exceptions
Bridging
Problem Reports
PR 86261
On an OS6800, Ethernet SNAP packets with non-zero Organizationally Unique Identifiers (OUI) will not
be classified using port-protocol rule.
Workaround: If applicable, other rules should be used to classify such packets.
PR 94269
The OmniSwitch 9000 counts all error packets as unicast packets in the packets received and error
counters, regardless of whether the packet is a unicast packet or a multicast packet. An oversized packet is
defined as a packet longer than 9216 bytes. This causes the following behavior in the switch:
1) Received packets longer than 9216 bytes are counted as unicast packets AND as error packets even if
the packet is a broadcast or multicast packet.
2) For ports operating at speeds of 10 Mbps or 100 Mbps, a packet is not counted as an error packet unless
it is longer than 9216 bytes.
Workaround: There is no known workaround at this time.
PR 94866
Invalid Packets with SA or DA set to all 0's continue to get bridged by the system.
Workaround: There is no known workaround at this time.
PR 99331
On an OS6800, when protocol or subnet based mobility rules are deleted and created after a VLAN delete,
the traffic coming in after this is not classified correctly.
Workaround: Delete the rule first before deleting the VLAN.
PR 99899
On an OS6800/OS6850, binding rules that require a port number to be specified as part of the rule can no
longer be deleted if the port is not set to be mobile.
Workaround: Make the port mobile before deleting the rule.
page 54
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Open Problem Reports and Feature Exceptions
Ethernet Interfaces
Problem Reports
PR 93114
It is not possible to broadcast or multicast oversized ingress packets. Therefore, ingressing oversized layer
2 multicast packets are treated as unicast packets.
Workaround: There is no known workaround at this time.
PR 94866
Invalid packets with SA or DA set to all 0s continue to be bridged by system.
Workaround: There is no known workaround at this time.
PR 95125
On an OS9000, the throughput on a 1 Gbps or 10 Gbps port might be 99.998%. This is due to a variance
in the oscillator; the clocking can differ by a few PPM. This variance is accepted by the IEEE standard.
Workaround: There is no known workaround at this time.
PR 95526
On an OS9000, if the admin status of the physical interface is toggled while sending high rates of traffic
on the 10G interface, some CRC errors are reported on the host connected to the interface.
Workaround: There is no known workaround at this time.
PR 105079
On OS6850 combo copper ports, the link will toggle once when the SFP is plugged in the corresponding
fiber cage.
Workaround: There is no known workaround at this time.
PR 105080
If a Gig Copper SFP is plugged into a combo port on an OS6850, then those combo ports can be used in
forced mode only. To use a copper combo port, then set the mode to Forced-Copper. To use the Gig
copper SFP in the corresponding copper port, then set the mode of the port to Forced-Fiber.
Workaround: There is no known workaround at this time.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 55
Open Problem Reports and Feature Exceptions
PR 105789
The ifMauTable returns object with out of range values for 10 Gig ports on an OS9000. A get call to
IETF_MAU MIB ifMauAutoNegSupported object returns a value of 0 (zero) for 10 Gig ports. Value 0
(zero) is out of range because only two values 1 (Autoneg supported) and 2 (Autoneg not supported) are
part of this object.
Workaround: There is no known workaround at this time.
PR 105847
On OS6850-U24x board combo ports with default media preference (i.e. Preferred-Fiber) and with 100M
Fiber SFP (Avago HFBR-57E0PZ), the link on fiber combo ports doesn't come up if copper cable is
plugged in the corresponding copper ports and the copper link is UP.
Workaround: On OS6850-U24x board combo ports, 100M Fiber SFP (Avago HFBR-57E0PZ) can get a
link with fiber cable in Forced-Fiber mode only.
PR 106041
On OS6850-U24x combo ports, if two OS6850-U24x boxes are connected back to back in Preferred-Fiber
mode with copper cable on copper combo ports and no cable on corresponding fiber ports then if the mode
of those ports is changed to Force-Fiber mode then LEDs of copper ports stay ON even though the link on
those copper ports go down. The LED's on copper media are turned off once the fiber side has a link.
Workaround: Once the fiber media gets a link, the copper media LED is turned off.
PR 106783
If fiber combo ports containing dual speed SFPs of two OS6850-U24X units are connected back to back in
Preferred-Fiber mode with no connection on corresponding copper ports, and if you change the modes of
these ports in a sequence to Preferred-Copper first and then to Forced-Copper, the LEDs on those Fiber
ports stay ON even though the link goes down.
Workaround: There is no known workaround at this time.
PR 108631
MIB Walk of HybridPort fails on OS6850 Lite 48port having combo ports (both upgraded and nonupgraded version).
Workaround: There is no known workaround at this time.
page 56
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Open Problem Reports and Feature Exceptions
Group Mobility
Problem Reports
PR 98417
When a MAC is learned on an OS6850 as "Filtered" for one port due to a Group Mobility rule violation,
and if the MAC reappears on another port, it will not be updated. The MAC will not be shown as filtered
for the new port, but will continue to show filtered in the old port.
Workaround: There is no known workaround at this time.
PR 98661
When using MAC based group mobility rules for VLAN classification on an OS6800/OS9000, some
entries are not inserted into the MAC VLAN table. This results in the frame not being classified.
Workaround: If this condition occurs on the mobile port, admin down and up the mobile port.
PR 107326
When the admin state of VLAN 1 is toggled between disable/enable on an OS6850, it is possible to have
mobile ports stuck in a blocking state.
Workaround: If this condition occurs on the mobile port, admin down and up the mobile port.
IP Multicast Switching (IPMS)
Problem Reports
PR 95057
On an OS6850/9000, the IPv6 Multicast Switching Spoofing feature is not replacing the clients' MLD
packets' source IPv6 address with the switch's IPv6 address. The Spoofing feature is replacing the clients'
MLD packets' source MAC address with the switch's MAC address.
Workaround: There is no known workaround at this time.
PR 98869
On an OS6800, egressing IP Multicast traffic onto a mobile VLAN from a 6800's mobile port will not
remove the 802.1Q frame.
Workaround: There is no known workaround at this time.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 57
Open Problem Reports and Feature Exceptions
PR 99269
An IGMPv3 or MLDv2 membership report with an exclude filter containing 0.0.0.0 or :: in the source list
will potentially disrupt IP Multicast Switching and Routing for the particular group membership being
reported.
Workaround: Neither 0.0.0.0 or :: are valid source addresses. They should not be used in source filters.
PR 105039
When Proxying in IPMS is enabled on an OS9000, IGMPv3 and MLDv2 reports generated by the system
on behalf of clients are not aggregated; rather IGMPv3 and MLDv2 reports are generated containing a
single record.
Workaround: There is no known workaround at this time.
Link Aggregation
Problem Reports
PR 90656
The same actor admin key assigned for two different dynamic link aggregates could cause ports to join
wrong aggregates.
Workaround: Assign unique actor parameters for each configured dynamic link aggregate.
PR 100537
If the link aggregation state is changed on an OS6850, the config sync status shows incorrectly as synchronised. This will result in the NI's going down on takeover.
Workaround: Issue the copy working certified flash-synchro command and then do a takeover.
Port Mirroring/Monitoring
Problem Reports
PR 86338
An OS6800/OS9000 preserves the INGRESS tag format of the packet for EGRESS mirroring, which
makes the mirror packet go out tagged though the real egress packet is not tagged.
Workaround: There is no known workaround at this time.
page 58
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Open Problem Reports and Feature Exceptions
Source Learning
Problem Reports
PR 83087
On an OS6800/OS9000, the MAC aging time can take up to twice the configured value.
Workaround: There is no known workaround at this time.
PR 94127
On an OS9000, in the Hardware learning mode, the source MACs from control packets (BPDUs) are
learned.
Workaround: There is no known workaround at this time.
PR 94180
On an OS9000, in the default learning mode, a MAC address will still be learned as a “bridging” entry
instead of the “filtering” entry when it matches the QoS Drop rules. However, the actual packet is being
discarded so the operation is functional.
Workaround: The switch could be configured to use “software” learning mode.
PR 94181
On an OS9000, if a QoS rule is set on a port to drop all traffic and if LPS is configured on the same port,
no MACs are learned as part of LPS, as all the traffic is dropped.
Workaround: There is no known workaround at this time.
PR 94515
On an OS9000, on the port which has learned port security enabled, the first packet is not forwarded. Once
this packet is validated, subsequent packets are forwarded without any problem.
Workaround: There is no known workaround at this time.
PR 94646
On an OS9000, the MAC address of an untagged packet does not get learned in the default VLAN with
"filtering" and with "accept only" tagged frames.
Workaround: There is no known workaround at this time.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 59
Open Problem Reports and Feature Exceptions
PR 95322
On an OS9000, duplicate MAC is not learned as the "filtered" mode, if the entry has already been configured as a permanent entry on a different port. The side effect is that the packet with this MAC from a
different port would still be forwarded or flooded.
Workaround: There is no known workaround at this time.
PR 96264
On an OS9000, in dynamic link aggregation, even when the link aggregation is in admin state “disable”,
the individual ports of the link aggregation exchange LACP packets. Therefore, the MACs from these
packets are learned due to hardware learning.
Workaround: There is no known workaround at this time.
PR 97310
On an OS6850, some MAC addresses get flushed from the CMM when the primary port of a link aggregate moves from one NI to another. The total count on the NI is correct.
Workaround: Set the source learning aging time value to a small value.
PR 97670
When the source learning mode is changed on an OS6850, sometimes the number of MACs learned may
not be displayed accurately although all the MACs are actually learned.
Workaround: The display count should catch up after one aging time has elapsed.
PR 98053
In an OS6800 Learned Port Security trap, only the IP address of the EMP port of the switch is provided (if
the switch has an EMP port) and not the IP address of the offending entity.
Workaround: There is no known workaround at this time.
PR 100761
On an OS9000, disabling a linkAggr port while traffic is running might leave some MAC addresses
learned on the primary member port of the linkAggr.
Workaround: Administratively down all ports that transmit across the linkAggr or link down/up the
primary port after disabling the link aggregate.
page 60
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Open Problem Reports and Feature Exceptions
PR 100932
Removing the last member port of a linkAggr while traffic is running can result in MAC addresses learned
on that port in the default VLAN of the linkAggr.
Workaround: Administratively down all ports that transmit across the linkAgg or link down/up the port
after removing it from the linkAggr.
PR 105399
When the OS9000 chassis is operating under the distributed mode of source learning and the traffic from
slots egressing link aggregation on different slots and the ingress and egress traffic are asymmetric, traffic
is flooded instead of unicast bridged.
Workaround: Configure the source MAC as static or use link aggregation on the same slot.
PR 106194
On OS6800 and OS6850 stacking platforms, a takeover reboots both the old primary CMM and NI. If traffic is coming on this NI, a 'port-security <slot/port> maximum x' command taking affect when it comes
back up could result in different MACs being filtered. The effect is the new secondary has the new MACs
filtered but the new primary has both the old and new set of MACs filtered. This is why the need to flush
out that particular port before the rule takes effect again.
Workaround: If this occurs, then explicitly configure which MACs should be allowed via the
port-security <slot/port> mac command.
PR 106462
When an OS9000 switch has less than 16K MAC address learned and 2 more slots receive data with MAC
addresses simultaneously that lead to more than 16K mac on the switch, the CMM displays all the MAC's
learned on each NI.
Workaround: There is no known workaround at this time.
PR 106533
On OS9000, when STP converges, MACs may not be flushed on ports that go to blocking. This will result
in traffic disruption until MAC ages out.
Workaround: Wait until MAC ages or perform a manual MAC flush.
PR 106781
On OS9000, sometimes max 16k macs do not learn across linkagg in chassis-distributed.
Workaround: Do not use chassis-distributed mode with linkagg and many MACs.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 61
Open Problem Reports and Feature Exceptions
PR 107231
A stack of 3 OS6850's (slots 1 & 2 are 48 ports and slot 3 is 24 ports) with a mobile port of 1/35 (Second
Asic on Slot 1) and a binding mobile rule. A MAC violating is programmed as filtered. Now, this filtered
MAC is programmed on all slots. However, since there is only one ASIC on Slot 3, there is an error and
the address is not programmed as filtered on slot 3.
Workaround: There is no known workaround at this time.
Spanning Tree
Problem Reports
PR 89316
A BPDU packet with the Root BridgeID of 0xffff...is sent out with every link-up to elicit a BPDU reply
from the adjacent switch in the current Auto-Edge Detection mechanism.
Workaround: There is no known workaround at this time.
PR 90297
On an OS6800/OS9000, CST Root convergence in 802.1s may be slow due to the circulation of old ‘good’
spanning tree vectors in the network when a root switch is powered off.
Workaround: 1) Use single MSTP region as much as possible. 2) Tune the performance parameters
maxAge and hop count to optimal values for the network.
PR 95308
Temporary traffic loops could happen under the following scenarios:
1. Reloading of a non-root bridge.
This happens when the bridge is going down and is due to the sequential bringing down of NIs during a
reload process. It is purely temporary in nature and stops when all the NIs eventually get powered off.
2. NI power down
When an NI power down command is executed for an NI and if that NI has the Root port and other NIs
have Alternate ports, it is possible to see some traffic looping back from the newly elected Root port. The
traffic loop back is temporary and will stop once the NI gets powered off.
3. New Root bridge selection
Temporary loops could occur during the process of electing a new Root bridge, if this election process is
triggered by the assignment a worse priority for the existing root bridge or a root bridge failure. This
happens due to the inconsistent spanning tree topology during the convergence and stops entirely once the
network converges.
page 62
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Open Problem Reports and Feature Exceptions
Workaround: For items 1 and 2 above, there is no known workaround at this time. For item 3, the
following workarounds could be applied:
1. Tune the max age (and or max hops in the case of MSTP) parameter to a lower value that is optimal for
the network. This will reduce the convergence time and thereby the duration of temporary loops.
2. To select a new root bridge, consider assigning better priority for that bridge instead of assigning worse
priority for the existing root bridge.
PR 96358
On an OS9000, during a Spanning Tree reconvergence triggered by link up/down or board power up/
down, the DVMRP subsystem may print the following error message: "dvmrpRecvProbe, Looping back
our probes".
This happens only if the Spanning Tree protocol selected is RSTP and is caused by rapid transitions of
port states. It has been verified that the problem does not happen for switched VLAN traffic. So there is no
chance of a real loop in the network. Packets handled in software might experience the problem due to
larger transit delays, but does not cause any malfunction.
Workaround: There is no known workaround at this time.
PR 100356
During boot-up, a Link aggregate port can momentarily become forwarding before blocking, causing
BPDU and dynamic MACs to be learned on that port. The dynamic MACs could potentially remain
learned on the Link aggregate port for some time, resulting in traffic disruption.
Workaround: Manually flush the MACs learned on the blocked port or admin down/up the blocked port.
PR 101214
Toggling the edge port of an OS9000 the very first time after boot-up may cause a TCN to occur in the
STP (1d) network.
Workaround: Configure the port as an edge port instead of an autoEdge port or configure the switch to
run in RSTP (1w) protocol.
PR 105493
Enabling Spanning Tree on an OS9000 in flat mode after disabling does not work when VLAN 1 is
disabled.
Workaround: Enable VLAN 1, enable STP and disable VLAN 1.
Step1: disable Spanning Tree -> vlan 1 stp disable
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 63
Open Problem Reports and Feature Exceptions
PR 105788
Some STP entries in the 802.1D standard mib may return out of range or undefined values on an OS9000.
This is because we are returning the true values of a newer version of STP (802.1Q 2005), not 802.1D
1998 that the MIB is based on. When a new standard MIB is defined, we can obsolete the old version.
Workaround: There is no known workaround at this time.
PR 106106
STP may return a hello time of 0 on SNMP queries on disabled instances after an OS9000 has been
running for some time.
Workaround: There is no known workaround at this time.
PR 106513
On an OS9000, STP can display the wrong port as the next best port. The next best port is not actually
used in topology calculation and has no effect on the network.
Workaround: There is no known workaround at this time.
VLAN Stacking
Problem Reports
PR 102958
On an OS9000, setting SVLAN priority using "vlan svlan priority" command may not work properly.
Workaround: Use the QoS command to trust all VLAN Stacking ports, and use QoS policy to configure
SVLAN priority.
PR 106514
Deleting a large range of SVLANs will sometimes result in an STP error message about the SVLAN status
on an OS9000.
Workaround: Delete the SVLANs in a smaller range of VLANs.
PR 107054
An error message is received on an OS6850 console after assigning a range of tagged vlans to an aggregate from CLI.
Workaround: This message can be ignored since it does not affect the actual assignment of the VLANs in
the range.
page 64
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Open Problem Reports and Feature Exceptions
Layer 3
Basic IP Routing
Problem Reports
PR 94621
In most cases, routed packets needing fragmentation due to a smaller IP MTU on the egress network will
not be fragmented and will be forwarded as-is.
Workaround: There is no known workaround at this time.
PR 106913
On OS6800 and OS6850, exceeding 20K routes in the FIB can cause system instability, especially if they
are all on the same VLAN.
Workaround: Reconfigure the switch such that the number of routes in the FIB do not exceed 20K.
PR 106931
Some objects in alaIpNetToMediaTable return wrong data type (counter, instead of integer).
Workaround: There is no known workaround at this time.
DHCP Snooping
Problem Reports
PR 100435
On an OS6850, a mobile port with incoming DHCP traffic is able to have a DHCP binding created for it
but it is not shown in the Binding Port Table.
Workaround: There is no known workaround at this time.
PR 106730
After DHCP Snooping is enabled against the VLAN on OS6850, adding the q-tag port to the VLAN, the
newly added q-tag port is not shown in the DHCP Snooping port list.
Workaround: Added the q-tag port to the VLAN first, then enable DHCP Snooping against the VLAN.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 65
Open Problem Reports and Feature Exceptions
PR 106814
When DHCP Snooping is enabled against mobile port, sometimes the mobile port's VLAN ownership is
not reflected in the DHCP Snooping port table.
Workaround: There is no known workaround at this time.
PR 106977
When DHCP Snooping is enabled, binding entry is not created against the new root port when STP topology is changed.
Workaround: There is no known workaround at this time.
PR 107186
For DHCP Snooping feature, the CLI should not allow user to configure ip-source-filtering on a trusted
port. However, if the port is a member of a link aggregate, the CLI does not display an error and allows the
configuration.
Workaround: Make sure the port is not a member of a link aggregate before configuring the ip-sourcefiltering.
IPv6
Problem Reports
PR 86669
On an OS6850/OS9000, IPv6 Router Advertisement decrementing timers are not supported for prefix
valid lifetimes or prefix preferred lifetimes.
Workaround: Use IPv6 Router Advertisement fixed timers.
PR 91228
An OS6850 does not detect IPv6 port scanning or other IPv6 denial of service attacks.
Workaround: There is no known workaround at this time.
PR 94546
An OS9000 does not make use of the MTU interface configured via the IPv6 interface command and only
supports the port MTU. Therefore, even if the configured IPv6 interface MTU is smaller, packets will still
be forwarded instead of being dropped and “a packet too big” message returned to the sender. This can
cause problems with path MTU discovery.
Workaround: There is no known workaround at this time.
page 66
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Open Problem Reports and Feature Exceptions
PR 96061
On an OS6850/OS9000, the IPv6 implementation does not follow ICMPv6 RFC 2463 section 2.2 Message
Source Address Determination in regard to anycast addresses. Instead IPv6 uses any unicast address
configured on the switch.
Workaround: A node that sends an ICMPv6 message has to determine both the source and destination
IPv6 addresses in the IPv6 header before calculating the checksum. If the node has more than one unicast
address, it must choose the source address of the message as follows:
"If the message is a response to a message sent to a multicast or anycast group in which the node is a
member, the source address of the reply must be a unicast address belonging to the interface on which the
multicast or anycast packet was received.
PR 99374
On an OS6850, the SNMP alaIPv6NeighborState MIB object has been deprecated. It is replaced by
alaIPv6NeighborReachability.
Workaround: All SNMP management previously done using the alaIPv6NeighborState object should
switch to the new alaIPv6NeighborReachability object.
PR 105893
A ping6 initiated on an OS9000 to one of its own addresses will succeed, even if the interface on which
the address is configured is disabled or inactive.
Workaround: There is no known workaround at this time.
PR 106826
In certain circumstances, on an OS9000, the RIPng routing table (as shown by the "show ipv6 rip routes"
CLI command) may display an inactive gateway for redistributed routes (i.e. routes redistributed "into"
RIPng).
In the RIPng routing table, the gateway shown is the gateway which was active when the route was originally redistributed into RIPng. If the original gateway becomes inactive at a time when other active gateways for the route exist, the RIPng routing table will not be updated to display the new gateway. As a
result, the RIPng routing table may display a gateway which is not currently active even though the route
is still active via other gateways.
Workaround: The active gateway for all routes (redistributed and otherwise) is available in the IP routemanager routing table. Use the show ipv6 router database CLI command to display the IP route-manager
routing table.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 67
Open Problem Reports and Feature Exceptions
PR 106213
When using "configuration apply" to apply routing configuration to an OS6850 system that has completed
its boot up, parts of the configuration may fail with a "system busy" error.
Workaround: 1) Ensure that the protocol is manually loaded prior to applying the configuration file. For
example,
> ipv6 load ospf
> configuration apply ospf3.asc.1
2) (Optional) After configuring the protocol, use a "write memory" to commit the changes to the boot up
configuration file.
Server Load Balancing (SLB)
Problem Reports
PR 105700
For SLB clusters on OS9000, there are no statistics or flow distribution metrics on a server basis. The only
server statistic is the number of packets passed to a cluster.
Workaround: There is no known workaround at this time.
PR 107490
When executing a “show ip slb cluster <clustername>” CLI command on a non-existing SLB cluster name
more than 23 characters long, an “slbCtrl” STR is generated. However, no adverse conditions are seen
with the switch.
Workaround: There is no known workaround at this time.
UDP Relay
Problem Reports
PR 97666
TCP does not send an RST after receiving an unacceptable ACK in SYN-RCVD state on OS6800 and
OS6850.
Workaround: There is no known workaround at this time.
page 68
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Open Problem Reports and Feature Exceptions
PR 106942
On OS9000, statistic display of ip-helper (DHCP) and Generic UDP relay are mixed together, but
complete statistical information is present.
Workaround: There is no known workaround at this time.
PR 107204
If the user defines a UDP service and does not assign a VLAN to that service, the service does not show
up on the Webview display but does appear when using the CLI to display services.
Workaround: There is no known workaround at this time.
VRRP
Problem Reports
PR 106943
Due to a coding incompatibility, standard SNMP managers will not be able to perform Get, GetNext, Test,
or Set Operations on objects in the alaVrrp3AssoIpAddrTable found in the proprietary
AlcatelIND1VRRP3.mib.
Workaround: For VRRPv2, the parameters controlled through these objects can be managed through the
vrrpAssoIpAddrTable in the IETF_VRRP.mib (RFC 2787), as well as through WebView and the CLI. For
VRRPv3 the parameters controlled through these objects can be managed via WebView and the CLI.
Advanced Routing
BGP4
Problem Reports
PR 103524
If there are more than 185 BGP redistribution filters configured in a boot.cfg file from an earlier release on
an OS6850 or OS9000, they are not consistently translated into IPRM route-maps that handle the redistribution.
Workaround: Do not attempt to restore more than 185 redistribution filters from a boot.cfg file generated from an earlier release.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 69
Open Problem Reports and Feature Exceptions
DVMRP
Problem Reports
PR 100130
On an OS6800, when interfaces from two different routers that are physically attached to the same
network are changed from native DVMRP interfaces to DVMRP tunnel interface on-the-fly or vice versa,
the multicast traffic will stop being routed across this path.
Workaround: Disable DVMRP before making the change, and then re-enable it again.
OSPF
Problem Reports
PR 106790
On OS9000, the forwarding address of AS-External LSA (corresponding to gateway of redistributed route)
is not updated when there is a change in the OSPF route to the forwarding address, due to deleting OSPF
interface configuration.
Workaround: 1) Disable and Enable OSPF admin status after deleting OSPF interface configuration, if
redistributed route's gateway was reachable on that OSPF subnet. 2) Disable and Enable IPRM route-map
that redistributes routes into OSPF, after deletion of OSPF interface configuration.
OSPFv3
Problem Reports
PR 104078
The route table in OSPFv3 cannot be retrieved via SNMP on an OS9000. There is no support for retrieving the OSPFv3 route table in the official IETF draft MIB for OSPFv3. Displaying the OSPFv3 border
router table is also not supported in the IETF draft MIB. This affects all hardware platforms that support
OSPFv3.
Workaround: The CLI can display the routing table (show ipv6 ospf routes) and border router table
(show ipv6 ospf border-routers) for OSPFv3.
page 70
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Open Problem Reports and Feature Exceptions
PR 105491
Existing inter-area-prefix and inter-area-router LSAs are not originated into areas that are created after
initial convergence of the network on an OS9000. For example, the router is running and it originates a set
of inter-area-prefix or inter-area-router LSAs into all known areas. Later, another area is added, the new
area will not have any of the inter-area-prefix or inter-area-router LSAs that exist in the other areas prior
to the creation of the new area.
Workaround: When adding an area to OSPFv3 after it has been running for a while, it should be globally
disabled and then re-enabled using the "ipv6 ospf status disable/enable" command.
PR 105770
Point-to-point interfaces are not supported in OSPFv3 at this time; therefore, 6to4 tunnels cannot be used
with OSPFv3.
Workaround: There is no known workaround at this time.
PR 107353
OSPFv3 will leak memory in either of the following two cases on an OS9000: 1) During SPF runs. 2)
During adjacency formation.
Workaround: There is no workaround that will eliminate this leak. However, it's effect can be somewhat
mitigated by trying to minimize the causal events (namely SPF calculations and new adjacency formations) between reboots.
PIM
Problem Reports
PR 102501
Currently, there is no support for an "ip unload" command. This can be a problem for customers who have
loaded a DRC loadable module and executed a "write memory" command to update the boot.cfg file. Then
decided that they wanted the module permanently removed.
Workaround: To permanently disable a module, execute the module's "disable" command followed by
another "write memory" command and a reboot. To also free up any memory possibly used by the
disabled module, edit the boot.cfg file using the "vi" command to delete the "ip load <module>" command
along with any non-default configuration lines associated with the module.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 71
Open Problem Reports and Feature Exceptions
PR 106343
High multicast traffic rate on an OS6850 running PIM-SM where the RP is on a remote switch may cause
high CPU utilization. This is due to the packets being PIM register-encapsulated which are software
routed. The high data rate is causing the Register Stop packets to be dropped. Once the Register Stop
packets are received, the register-encapsulation will stop and native forwarding will take over.
Workaround: Avoid register-encapsulation by configuring the RP to be on the same switch as the source.
Quality of Service (includes ACLs)
General
Problem Reports
PR 105380
If there is no arp entry for the destination, routed traffic matching a drop policy gets dropped in software
by the CPU instead of being dropped by the OS6800.
Workaround: There is no known workaround at this time.
PR 105496
Rules that match multicast traffic do not get logged properly on an OS9000.
Workaround: There is no known workaround at this time.
PR 105764
Due to the method QoS handles condition groups (network, services, etc.), the flushing of conditions
corrupts the linkages between the condition and the groups. The condition remain attached to the SLB
cluster. However, the group pointers in the condition are invalid. Because of this, the group parameters are
restricted.
Workaround: There is no known workaround at this time.
PR 107091
On OS9000, QoS currently will not accept names of 23 character from SLB.
Workaround: Use names of 22 characters or less.
PR 107190
SLB will return an error if the configured ping period is out of range on an OS9000. However, the ping
period will still be updated.
Workaround: Change the ping period to a correct value.
page 72
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Open Problem Reports and Feature Exceptions
PR 107232
On OS9000, if a condition is associated with an SLB cluster and a policy rule, it can be deleted in a flush.
Workaround: Create separate conditions for SLB and policy rules.
Policy Manager
Problem Reports
PR 94083
Policies, which specify a destination slot/port or destination port group will not be applied to traffic which
is routed by the switch, these policies will match only bridged traffic.
Workaround: There is no known workaround at this time.
PR 94125
The OS9000 and OS6850 does not support QoS or ACL rules containing destination port, destination
VLAN, or destination MAC on traffic that is routed by the OS9000 or OS6850.
Workaround: There is no known workaround at this time.
PR 95249
On an OS9000, the 802.1p value of IP packets is set to 0. On trusted ports, the 802.1p is not altered for
non-ip packets. For IP packets, the prioritization is per the TOS value. The 802.1p is also restamped to
reflect the ingress TOS precedence value. If the TOS value is 0, the 802.1p is set to 0.
Workaround: Use QoS 802.1p stamping policies, which match on ingress the 802.1p value, to retain the
ingress 802.1p. For example:
policy condition c 802.1p 1
policy action a 802.1p 1
policy rule r condition c action a
qos apply
PR 99336
On an OS6800, QoS policies which specify ethertype do not match SNAP or 802.3 Raw packets. QoS
policies only support Ethernet-II packet format.
Workaround: The policy will match traffic which matches the policy criteria even if the packet is not
Ethernet-II if the policy specifies only one of the following: source slot/port, destination slot/port, source
MAC or destination MAC.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 73
Open Problem Reports and Feature Exceptions
PR 99931
When tagging a link aggregate on an OS6850, QoS does not trust the individual ports of the link aggregate.
Workaround: Manually set the trust bit of the underlying ports through QoS (qos port <slot/port>
trusted), or set the port default to trusted (qos trust ports).
PR 99983
The OS6850 switch cannot boot up properly with a boot.cfg that exceeds the QoS limitation. It is not
recommended to manually edit the boot.cfg to configure your QoS. Booting up with a boot.cfg obtained
from a "write memory" is fine.
The hardware allocation checking is not done during boot up causing QoS configurations to be out of sync
with the hardware capability.
Workaround: To prevent the boot.cfg from going beyond the QoS limitations on a large QoS configuration, proceed as follows: edit a text file with your desired qos configuration, apply the configuration using
"configuration apply text_file", and save the boot.cfg using "write memory".
PR 101223
On an OS6800, if a policy rule specifies the keyword "log" or "log interval", then the policy is rejected.
Workaround: Logging is not supported by the OS6800. The keyword "log" and "log interval" has to be
removed from the policy rule definition.
Security
General
Problem Reports
PR 89262
NESSUS reports bogus “Vulnerabilities”. Basically, NESSUS collects all those known attacks/vulnerabilities into their test suites.
For example, NESSUS sends: http://<switch-address>/cgi/bin/guestbook.cgi
WebView/HTTP-Server’s response: Prompts user for the default switch login page (which is the normal
operation for our embedded server).
Since our HTTP server replies with some form of an HTTP response, NESSUS mistakenly concludes that
the HTTP server is vulnerable to this attack.
Workaround: There is no known workaround at this time.
page 74
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Open Problem Reports and Feature Exceptions
PR 91681
On an OS9000, the following is noted when running a test called "alya.cgi (Backdoors)" in the NESSUS
test suite:
Security Note: Web mirroring - http (80/tcp)
The following CGI have been discovered:
Syntax: cginame (arguments [default value])
/web/content/login.html (userName [] password [] B1 [Login])
Workaround: This is the expected behavior for the login pages for WebView and Web-AVLAN authentication. NESSUS is known to provide those bogus reports.
PR 95642
On an OS9000, the Denial of Service testing tool (NESSUS) generates bogus reports.
Workaround: There is no known workaround at this time.
PR 107176
OS6800 may incorrectly show the router-mac address as the source of a DoS attack.
Workaround: There is no known workaround at this time.
802.1x
Problem Reports
PR 98375
On an OS6850, DHCP rules are not being used to classify traffic on a regular group mobility port. For this
reason, with a matching DHCP rule, Device Classification policy will not consider matching a DHCP rule
as having a matching GM rule when the policy is applied.
Workaround: There is no known workaround at this time.
PR 99658
On an OS6850, not all MAC addresses will be learned when testing with a traffic generator to simulate
traffic with incremental MAC addresses.
Workaround: There is no known workaround at this time.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 75
Open Problem Reports and Feature Exceptions
PR 100189
On an OS6850, when the MAC address table is full, source learning will not learn MAC addresses dynamically and the non-supplicant table will show more entries because the tables are not synchronized.
Workaround: There is no known workaround at this time.
PR 100614
On an OS6850, when a device is moved from one 802.1x port to another 802.1x port, the device is not
classified according to the device classification policy that applies to the new port but is learned on the
default VLAN for the new port.
Workaround: Reconfigure the new port by disabling and enabling 802.1x on the port. May also have to
reconfigure the device classification policy for the new port as well.
PR 106463
The CLI command “802.1x initialize <slot>/<port>” only applies to the supplicants on the specified port.
All the supplicants are forced to authenticate again. Non-supplicants on the same port are not affected; no
re-classification for non-supplicants is required when this CLI command is used.
Workaround: There is no known workaround at this time.
Authenticated Switch Access
Problem Reports
PR 91812
On an OS9000, the server information displayed with the show configuration snapshot aaa command or
saved with the configuration snapshot aaa <file_name> command contains hashed (encrypted) password/key information. In order for a file created with the latter command to be used for configuring servers, password/key information needs to be edited. AAA expects this information encrypted only at boot-up
time while at run time the information should be in plain text. In this particular case, the servers created
with configuration apply command could not be used because password/key information is wrong.
Workaround: Always edit password/key information before applying a snapshot file.
PR 106557
SNMP MIB walk on aaaAcctCmdTable returns empty even when the table is set.
Workaround: There is no known workaround at this time.
PR 107085
Accounting log for scp-sftp displays user IP address as 0.0.0.0 on an OS6850.
Workaround: There is no known workaround at this time.
page 76
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Open Problem Reports and Feature Exceptions
Authenticated VLANs
Problem Reports
PR 87642
On an OS6800, the CLI command to specifically disable 802.1x or AVLAN authentication on a port will
disable either of the authentication options configured on the port.
Workaround: There is no known workaround at this time.
PR 98369
DHCP is not supported with port-binding AVLANs on OS6800/OS6850. When DHCP packets are used to
trigger the port binding rules, none of the rules work.
Workaround: There is no known workaround at this time.
PR 106976
When DHCP Snooping's IP Source Filtering is enabled on the Authenticated VLAN port of an OS6850,
the authentication (via Telnet or HTTP) will fail.
Workaround: Cannot enable IP Source Filtering on AVLAN ports, since IP Source Filtering (work as
expected) is blocking the IP traffic.
Policy Server Management
Problem Reports
PR 103324
An OS6850 will not change the IP address automatically even if the supplicant client is running that can
automatically do the ipconfigure release and renew when dynamically changing classification policy when
an IP net rule is configured. Depending on what traffic is running, some packets may satisfy the IP net
rule and the supplicant will be classified according to the IP net rule.
Workaround: User has to be aware that when the IP net rule is configured and when dynamically changing the classification policy that as group mobility as one of the classification option, traffic from supplicant may still have the old IP address on the vlan that the supplicant was classified before the policy is
changed. The IP net rule will cause the client to be learned on the vlan that it was previously learned on.
E.g. supplicant is learned on vlan x and has an IP address with vlan x's subnet. There is also an IP net rule
for vlan x's IP to be classified on vlan x. When user dynamically changes the classification policy, the
supplicant may still be learned on vlan x because the PC has traffic coming out with VLAN x's subnet and
thus device classification task will classify the supplicant on VLAN x again.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 77
Open Problem Reports and Feature Exceptions
PR 107086
User with readwrite-scp-sftp privileges is initially queried for authentication and authorization on an
OS6850. After login to the scp/sftp shell, only accounting requests are sent to TACACS+ server (if
enabled), commands are not queried for authorization.
Workaround: There is no known workaround at this time.
PR 107278
SNMP variable policyStatsQueryCount always show as zero.
Workaround: See policyStatsAccessCount instead.
PR 107543
If Loopback0 is defined by the user and there is not a physical route for that IP subnet, a RADIUS client
will not be able to communicate with a RADIUS server. As a result, RADIUS authentication will fail as
the server is unreachable.
Workaround: There is no known workaround at this time.
System
General
Problem Reports
PR 86853
A stack of 9 or greater is not supported on an OS6800 or OS6850.
Workaround: The maximum stack elements is 8.
PR 93114
If an ingress packet is oversized, it is not for broadcast or multicast anymore. As a result, ingressing oversized layer 2 multicast packets are counted as unicast packets.
Workaround: There is no known workaround at this time.
PR 97213
On an OS9000, in all CPU exceptions, a Trace Buffer dump is offered at the beginning of the exception.
Workaround: There is no known workaround at this time.
page 78
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Open Problem Reports and Feature Exceptions
PR 99862
When an over stressed test caused 100% CPU usage on an OS6850, the "Unable to send running checksum" message was displayed, and the switch was rebooted.
Workaround: Avoid 100% CPU usage.
PR 100127
By default, an OS6850 switch is configured to run in strict-priority. If over-subscription is done on priorities 6 & 7, it will bring down the switch.
Workaround: Over-subscription on priorities 6 & 7 is not yet supported on the switch. Do not configure
the over subscription with priority 6 or 7.
PR 105381
The flash file system on an OS6850 shows no space but allows copy anyway.
Workaround: Remove any unneeded files before performing copy.
PR 105646
The entPhysicalModelName MIB variable returns vendor name of SFP instead of model name for an
SNMP get/getNext call to this object on an OS6850.
Workaround: There is no known workaround at this time.
PR 106559
Output to STDOUT is re-directed to the PMD file on an OS9000. A printf is occurring that causes the
“show log” command to fail.
Workaround: The contents of the pmd file can still be viewed with vim, gvim, etc. as most of the
contents are ascii strings.
PR 106812
The "update default miniboot" is still supported on OS6850 even though there is no default or backup
miniboot.
Workaround: This command will still cause the miniboot to be updated properly.
PR 106649
The “update miniboot” command is not working on OS-9000.
Workaround: Use the “update uboot-miniboot” command to update the miniboot.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 79
Open Problem Reports and Feature Exceptions
PR 107139
In a stacked system on an OS6850, if the redundant cable is absent, the primary module could be one of
the elements with a higher slot ID on a reboot of the system.
Workaround: Use the stacking cables and provide complete redundancy.
PR 107276
Swlog does not allow the user to select the TCP destination port. By default, the fixed port of 514 is
selected on both OmniVista and the AOS switch. On some setups, the OmniVista port needs to be changed
and the switch can not send to this new port.
Workaround: Use the default TCP port setting of OmniVista or set a third party syslog server to port 514.
PR 107033
When port 1/1 on the system has software learning enabled...either by the port being mobile or authenticated and if there is a linkagg configured on the switch, the internal mac 0x0020da999999 is seen in the
output of show mac-address table learned as being learnt on port 1/1.
Workaround: There is no known workaround at this time.
Chassis Supervision
Problem Reports
PR 94404
Mixing OS7000 and OS9000 CMMs in any chassis is not supported.
Workaround: There is no known workaround at this time.
PR 95320
On an OS9000, the Unix mv command does not update a file's time stamp. Therefore, the check sum will
not detect the change.
Workaround: Use the cp command instead.
PR 96225
On an OS9000, the "+++ i2cReadRemoteCMM: Error writing starting address!" error message is seen on
the secondary after the primary crashes and a takeover is in progress. However, the system functions
normally and takeover still occurs.
Workaround: There is no known workaround at this time.
page 80
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Open Problem Reports and Feature Exceptions
PR 96327
Extra display character in the swlog. No effect on the switch.
Workaround: There is no known workaround at this time.
PR 96584
On an OS9000, during a reload, when the fabric LED is turned off and on, there is no effect on the switch.
Workaround: There is no known workaround at this time.
PR 98768
If hardware configuration changes are introduced to an OS6800/6850 stack without first ensuring that the
Primary software configuration is certified, it is possible to create an endless synchronization cycle.
Workaround: Before making any hardware configuration changes ensure that the Primary is certified.
Run the copy working certified flash-syncro command before making hardware configuration changes.
PR 98956
An OS6850 supports only single point of failure. Removing multiple stacking cables will introduce multiple points of failure from which the system may not recover.
Workaround: Do not remove multiple stacking cables at the same time.
PR 100244
Installing 6.1.3.R01 images on an OmniSwitch 6800 that is running 5.3.1.R02 code returns an error.
Workaround: The code base has changed significantly from 5.3.1 to 6.1.2. This is especially true of
Jos.img. So the install from 5.3.1 will not be able to support them. Reboot the 6.1.2 images and then do
the install.
PR 100825
On an OS9000, if both CMMs are removed from the chassis, Layer 2 local (local to the NI) traffic with
learned MACs is switched.
Workaround: Insert at least one CMM into the switch to reset the NIs.
PR 103625
Pulling all the cable simultaneously on a stack of 8 OS6850's causes problem.
Workaround: When pulling the cable, the system will start its topology change. Pull the cable one by one
during the topology change will lose the topology information. Try to pull the cable slowly to avoid this
problem.
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
page 81
Open Problem Reports and Feature Exceptions
PR 104874
If CPU is at 100% on an OS6800, then the commands "copy working certified" or "copy working certified
flash-synchro" can make the system timeout and the primary will crash with pmd. The timeouts are printed
on the screen. So, the user can know when it timeouts. User has to be aware of the copy from working to
certified process.
Workaround: Prevent doing a copy working certified or copy working certified flash-synchro if high
CPU or remove CPU load.
PR 106941
Processing for CMM MIB variable chasTempRange not correct for non-primary slots on OS6800.
Workaround: There is no known workaround at this time.
PR 106975
If a WINTEK compact flash is installed on an OS6850, WINTEK is not displayed when you use the show
hardware info command.
Workaround: There is no known workaround at this time.
PR 107619
NI 1 becomes non-primary in a OS6850 stack after “reload stack” if NI 1 is an OS6850 Lite.
Workaround: There is no known workaround at this time.
Power Over Ethernet
Problem Reports
PR 99583
OS6850 POE units support either 510 or 360 Watt power supplies. If unlike power supplies are mixed, or
if an unsupported power supply, such as a 120-Watt power supply are used, a console message and a trap
are generated.
Workaround: There is no known workaround at this time.
page 82
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Open Problem Reports and Feature Exceptions
PR 106065
During the execution of the "update lanpower all" command, it is possible that the update of an individual
port on the OS9800 may fail with a message similar to the following example for slot 14:
FRI OCT 06 11:51:55 : LANPOWER (108) info message:
+++ Shutting down daughter module in slot 14!
call failed: RPC: Success
PD63000 Programming Failed Slot #14!!!.
Workaround: Execute a lanpower update for the specific slot which failed.
PR 106121
Currently the test portion of the OS9000 lanpower start command checks available power before any
attempt is made to activate.
Workaround: There is no known workaround at this time.
PR 106791
The non-default value of 230 watts is not saved for the 510 watt power supply when performing "write
memory" or "configuration snapshot..." on an OS6850.
Workaround: Manually modify resulting configuration file.
PR 106906
Users may over allocate power from the available power supply in Power Shelf powered applications on
OS9000. PoE Power Shelf and PoE
Port Guidelines clearly state the user procedures for calculation of appropriate steps for calculation of
blade power requirements based on port power requirements, and total power requirements for the system.
Redundancy and power supply calculation is also provided.
Workaround: There is no known workaround at this time.
PR 106978
Currently PoE configurations are not deleted after a P24 blade is removed from OS9000 chassis.
Workaround: If a P24 blade is to be permanently removed from a chassis, then no action is necessary. If
a P24 blade is removed, a different blade is substituted and then at some later time a P24 blade is again
placed in the same slot position, either the original configurations for the P24 blade must be changed or
deleted (this is the preferred method) or the configuration file may be edited (this should only be
performed by an individual extremely familiar with configuration files and PoE configuration as the
ENTIRE unit configuration may be changed, or damaged).
page 83
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Open Problem Reports and Feature Exceptions
PR 107287
During "update lanpower all" on a stack of 8 OS6850-P24/48, you might see the following error message:
THU OCT 12 14:06:27 : LANPOWER (108) error message:
+++ Unable to Read S19 Response!
Reset Daughter Module!
Done
THU OCT 12 14:48:59 : LANPOWER (108) error message:
+++ Unable to Read BOOT_SECTION_RESPONSE Response!
+++ General Programming Error!
Workaround: Do a "lanpower start" on the NI which failed the firmware update. Once lanpower failed to
start which is expected, you then try to "update lanpower <slot>" again on that specific NI.
Redundancy / Hot Swap
Hot Swap Time Limitations for OmniSwitch 9000
• All removals of NI modules must have a 30 second interval before initiating another hot swap activity.
• All insertions of NI modules must have a 3 minute interval before initiating another hot swap activity.
• All hot swaps of CMM modules must have a 10 minute interval before initiating another hot swap,
reload or takeover activity.
• All takeovers must have a 10 minute interval before following with another hot swap, reload or take-
over activity.Problem Reports
Problem Reports
PR 91287
After takeover, the new primary CMM does not keep the DOS statistics held by the previous primary
CMM.
Workaround: There is no known workaround at this time.
PR 95840
The whole chassis will reload if more than half of NIs are hot swapped in and out of the chassis at roughly
the same time.
Workaround: If hot swap of NIs is required, a user may have to wait until a previous NI has been booted
first, and then hot swap the next NI.
page 84
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Open Problem Reports and Feature Exceptions
PR 96011
On an OS9000, NIs will not reset on an unsynchronized takeover if those NIs are manually reset between
configuration change and takeover.
Workaround: There is no known workaround at this time.
PR 98992
The copy working certified flash-synchro command display takes a long time on an OS6800 switch.
OS6800 does not support tffs (true flash file system) but OS6850 does. So an OS6800 switch takes much
longer whenever there is an operation related to file operation.
Workaround: There is no known workaround at this time.
PR 106872
Upon unsynchronized takeover in Fuji 16 slot chassis, some NIs don't come up gracefully after resetting.
Workaround: Power cycle the NIs that are not UP and ready.
page 85
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Technical Support
Technical Support
Alcatel technical support is committed to resolving our customer’s technical issues in a timely manner.
Customers with inquiries should contact us at:
Region
Phone Number
North America
800-995-2696
Latin America
877-919-9526
Europe
+33-388-55-69-04
Asia Pacific
+65-6586-1555
Other International
818-878-4507
Email: support@ind.alcatel.com
Internet: Customers with Alcatel service agreements may open cases 24 hours a day via Alcatel’s support
web page at: http://eservice.ind.alcatel.com.
Upon opening a case, customers will receive a case number and may review, update, or escalate support
cases on-line. Please specify the severity level of the issue per the definitions below. For fastest resolution, please have telnet or dial-in access, hardware configuration—module type and revision by slot, software revision, and configuration file available for each switch.
Severity 1 Production network is down resulting in critical impact on business—no workaround available.
Severity 2 Segment or Ring is down or intermittent loss of connectivity across network.
Severity 3 Network performance is slow or impaired—no loss of connectivity or data.
Severity 4 Information or assistance on product feature, functionality, configuration, or installation.
page 86
OmniSwitch 6800/6850/9000—Release 6.1.3.R01
Download PDF

advertising