vNIC Failover and Failback

[Adrian Sutton]

Hi,

We have been testing vNIC Failover from Fabric A to B and ony 1 ping is dropped in this process and everyone is happy.

However when we enable the uplink again and the vNIC Failsback to Fabric A we can see up to 5 pings dropped before normal comms are returned.

This is not really ideal so has anyone seen a similar scenario

We are testing this failover and back with Hyper-V and Windows install on the servers with the same results

Thanks

A

[David.Jarzynka]

This is normal. The failover tends to lose 2 pings and the failback tend to lose 5 or so. The additional lose is proabably due to the MAC tables syncing or reconvergance of the fabric. Both VMware and Hyper-v should never really see a failover. The beauty of UCS

Hope that helps

[george.steward]

This is probably 4 years to late but for anyone else! I ran across this myself and found the solution.

http://www.cisco.com/c/en/us/td/docs/unified_computing/ucs/sw/gui/config/guide/2-0/b_UCSM_GUI_Configuration_Guide_2_0/b_UCSM_GUI_Configuration_Guide_2_0_chapter_010100.html

Failback Timeout field

After a vNIC has started using its secondary interface, this setting controls how long the primary interface must be available before the system resumes using the primary interface for the vNIC. Enter a number of seconds between 0 and 600.

Server tab > Policies > root > Adapter Policy > Eth Adapter Policy Linux/VMware/Windows etc…

The default setting for most of the Eth Adapter policies are 5 seconds. Change it to 30 seconds and you'll experience a 1 ping loss on failback. 

[denniskuan_csa]

Thanks George, your post help a lots!