UCS FI failback

[Varun K S]

Hi All,

I had rebooted FI-A all the traffic got failed over to FI-B but when FI-A came back up after reboot, I lost the connection to almost all esxi and also physical servers. any idea what can be the reason, is there any thing like preempt delay that need to configured on FI ?

Your help and comments are much appreciated. 

Regards

Varun KS

[Walter Dey]

Hi Varun

Did you verify that the FI cluster came back in a healthy state ?

CLI: show cluster status ?

What is the FI Northbound connectivity ?

Walter.

[Varun K S]

Hi Walter,

Custer status was normal didnt find any issue. 

also one thing note was vsm also lost connection to vem once FI-A came back up after reboot. 

Northbound connectivity is over vpc fibre towards 5K.

Regards

Varun KS

[Walter Dey]

Hi Varun

If everything is properly configured, you should not have any single point of failure.

- vpc Northbound

- UCS fabric A/B

- Service profile with vnic's/vhba's connected to fabric A/B

- OS vswitch DVS with loadbalancing / failover

Before rebooting the FI, did you ever verify, eg. disable all uplinks on one FI, and see if everything keeps running ok or not.

Walter.