NIC configuraton inconsistency to deploy vswitch

You may have NIC configuration inconsistencies with the vswitch. This happens regularly when you want to make some changes when the cluster is down or when the vswitch is enabled. In my case this showed up on a fresh cluster install, I changed the bond type and added NICs in the bond without disabling the vswitch. (Thanks Alejandro Suarez from Nutanix to helping me and Farhan 😉 ).

Exemples :

Solution :

If you do not have access to the CVMs, you can connect to the hypervisor from the IPMI and connect in SSH via the address 192.168.5.254.

➔ We need to disable the virtual switch. Don’t stop the cluster !

➔ Edit your configuration according to your wishes.

➔  Check if lacp is configured and enabled.

➔ Check the configuration uplinks.

➔  Migrate the bridge br0 to the vswitch vs0. (to enter in edit acropolis : « acli »)