If you are using Nexus 5000 VPC as the method for host connection to Nexus 1000v, you may want to watch out for a current situation. Last week at Networkers, best practice was described as using mode active on Nexus 5000, and mode passive on Nexus 1000v, thus allowing Nexus 5000 to establish the LACP port channels.
However, there is a reported error condition, when triggered by something like a VSM reload, which will effectively put certain ports into "suspended" mode, as shown below:
vsm-1(config)# sh port-channel summary
...
5 Po5(SU) Eth LACP Eth5/5(s) Eth5/7(P)
Another symptom is "show CDP" does not match between 1000v and 5000, with only one side sending hellos.
Using a Cisco internal command, shows more detailed event information:
vsm-1(config)# sh port-c internal event-history errors
1) Event:E_DEBUG, length:162, at 710664 usecs after Fri Jul 9 20:38:39 2010
[102] pcm_proc_response(373): Ethernet5/5 (0x1a040400): Setting response status to 0x402b000c (port not compatible) for MTS_OPC_ETHPM_PORT_BRINGUP (61442) response
2) Event:E_DEBUG, length:84, at 710191 usecs after Fri Jul 9 20:38:39 2010
[102] pcm_eth_port_ac_drop_all_txns(798): Interface Ethernet5/5 suspended by protocol
It will probably be fixed in a future N1000v release, current workaround? Set LACP mode to "active" on Nexus 1000v:
port-profile type ethernet systemuplink_profile
vmware port-group
switchport mode trunk
switchport trunk allowed vlan all
channel-group auto mode active
no shutdown
system vlan ...
state enabled
No comments:
Post a Comment