VMware Cloud Community
Drewan99
Contributor
Contributor
Jump to solution

Recovery after nic failover

I have an esxi 6.5 update 2 standalone server in a test lab.

I can create a vswitch with two nics for failover. When one of the nics goes down, the failover works as it should….yaaaa.

The problem is when the failed network ports becomes active again, the vm’s on that vswitch can no longer communicate with the network.  They can communicate with vm’s on that vswitch but not beyond. The only way to get it to work again is to remove the nic that failed from the vswitch. Once removed communication starts working again. I'm then able to add the restored nic to the vswitch without problem.The problem is reproducible and happens to any vswitch that has nic failover.

Any Suggestions?

Reply
0 Kudos
1 Solution

Accepted Solutions
Drewan99
Contributor
Contributor
Jump to solution

Found the answer. It was the Cisco switch. We had both nics going from the ESXi server into the same Cisco switch which was causing a loop back. The switch guy added a command to fix that and all is good now and things are running a lot faster also.

View solution in original post

Reply
0 Kudos
5 Replies
Dave_the_Wave
Hot Shot
Hot Shot
Jump to solution

When I was new to ESXi, I too was wondering how nic teaming and failover would work.

This thread may help:

https://communities.vmware.com/thread/583631

Reply
0 Kudos
dbalcaraz
Expert
Expert
Jump to solution

Hi,

Would you able to share the vswitch's config?

It really depends in how do you configure your vswitch and also that physical switches' ports are configured correctly.

-------------------------------------------------------- "I greet each challenge with expectation"
Reply
0 Kudos
Dave_the_Wave
Hot Shot
Hot Shot
Jump to solution

Screenshots are all on that thread I started a while ago.

I don't like having Physical Adapters on standby, which is a real waste.

I like to team all available ones at the same time, and should one fail (someone tugging on a patch cord), the remaining one still works.

Reply
0 Kudos
Drewan99
Contributor
Contributor
Jump to solution

Found the answer. It was the Cisco switch. We had both nics going from the ESXi server into the same Cisco switch which was causing a loop back. The switch guy added a command to fix that and all is good now and things are running a lot faster also.

Reply
0 Kudos
Dave_the_Wave
Hot Shot
Hot Shot
Jump to solution

That's really exciting, I think a lot of folks will team their nics onto the same switch and believe there'll be parallel speeds to that, can someone from VM confirm or correct otherwise? What about standalone switches that are not managable?

Reply
0 Kudos