Cisco Nexus 5000 Series and Cisco Nexus 2000 Series Release Notes, for Cisco NX-OS Release 5.0(2)N1(1) and NX-OS Release 5.0(2)N2(1) OL-22747-02, December 2010)

Send documentation comments to [e-mail]
40
Cisco Nexus 5000 Series and Cisco Nexus 2000 Series Release Notes, for Cisco NX-OS Release 5.0(2)N1(1) and NX-OS Release 5.0(2)N2(1)
OL-22747-02
Caveats
• CSCtj39025
Symptom:
On a Cisco Nexus 5000 Series switch after upgrading to a 5.0(2)N1(1) release, if the
partner side of the LACP port channel is configured as lacp rate fast and there are more than 10 ports
connected through a Cisco Nexus 2148 Fabric Extender, these ports flap periodically. The partner
side of the LACP port channel is configured as lacp rate fast. The Cisco Nexus 2148 image is
upgraded through ISSU.
Workaround: Reload Cisco Nexus 2148 Fabric Extender or configure the 10 or less ports at a time
on the partner.
• CSCtj40413
Symptom:
When checkpoints are deleted using the clear checkpoint database command,
subsequent verify/commit operation from switch-profile fails with the following error:
Failed: Rollback in progress.
Workaround: Delete every checkpoint file individually using the no checkpoint <name>
command and create another checkpoint after using the clear checkpoint database command.
• CSCtj46045
Symptom:
A Cisco Nexus 5000 Series may crash when the Fabric Extender (chassis) IDs for 2
Fabric Extenders are exchanged.
Workaround: Perform a shut on the fabric ports of the Fabric Extenders before exchanging the
Fabric Extender IDs and perform a no shut after the Fabric Extender IDs are exchanged.
• CSCtj46688
Symptom: Very short power cuts / fluctuations on the power grid in the DC may trigger a reload of
the FEX. However, in certain situations the FEX does not reboot. It remains "stuck" and the blue
locater led stays on. This occurs when the reload is triggered by very short power cut / fluctuations.
(typically seen when no UPS is used
Workaround: Turn off the power to the FEX for five seconds and power it back up.
• CSCtj56177
Symptom:
After downgrade from the Cisco NX-OS Release 5.0(2)N1(1) to Cisco NX-OS 4.1(3)N
based releases, class-ip-multicast or class-all-flood is lost from type qos policymaps. This causes
all traffic mapped to such classes to be classified to other classes in the type qos policymap. This
happens only when the Cisco Nexus 5000 Series switch is downgraded from the Cisco NX-OS
5.0(2)N1(1) release to Cisco NX-OS 4.1(3)N based releases.
Workaround: Reconfigure class-ip-multicast and class-all-flood classes in the type qos
policy-maps after downgrading to the Cisco NX-OS 4.1(3)N based releases
• CSCtj65732
Symptom: Peer keep-alive syslog messages are missing.
Workaround: None
• CSCtj71171
Symptom:
Upgrade from 4.2(1)N2(1) to 5.0(2)N1(1) failing. N5K stuck in a boot loop. The service
"port-profile" is causing the reload. This occurs when the Found Banner motd not being left
justified.
Workaround:
1) Ensure that the banner motd text is left aligned with no preceding spaces.