A consultant is completing final network failover testing on a recently deployed AHV cluster.
The cluster consists of the following:
1. Three nodes
2. Two uplinks connected to two switches
3. Supported SFPs
The consultant removed one SFP+ out of each node. Two of the nodes (Hostl I Host2) failed over to the backup NIC correctly without any interruption but have lost communication with the third node (Host 3) hypervisor/CVM. Inserting the original SFP+ back into the problematic node allowed the network to be connected and the two healthy nodes could communicate successfully again.
manage_ovs showjnterfaces output
manage_ovs showjinterfaces output:
name mode link speed
cthO 10000 True 10000
ethl 10000 False None
eth2 10000 True 10000
eth3 10000 False None
manage_ovs show.uplinks output:
Bridge: brO
Bond: brO-up
bond_mode: active-backup
interfaces: ethl ethO
lacp: off
lacp-fa I 1 back: false
lacp_speed: slow
What should the consultant validate first?