I have a simple cluster 2 Esxi 5U1 hosts with a SAN (HP P2000).
My vSwitch are as follow:
vswitch0: Management network
vswitch1: vmkernel with 2 Portgroup for ISCSI access (192.168.250.0/24 and 192.168.251.0/24) No option like vMotion of Management network checked
vswitch3: Portgroup 1 vmkernel one for HA on a management out-of band network
Portgroup 2: vmkernel with vMotion activated
vSwitch4: Virtual machine LAN
On host1, no warning, but on host2, I have an error message telling me that HA cannot reach a management network. It try to ping 192.168.250.30 who is on vSwitch1. There is no managment network checked ?
I have rebooted the hosts as it is not in production, without success.
Any idea?
technically you can remove these warnings, however if you want to do a clean conclusion, then try to ping ths ips from the console of the 2nd host... here is what I would have done
from Host 1ping all the ips on host 2 and vice versa. Might be HA has taken some of other vmk ips for primary management ip (ips you used for storage for example).
Using the console, I cannot ping the ip, and it is normal as it is not a management network ip but a vmkernel without managemnet network checked used for ISCSI on a separate network.
What is strange is that a management network trie to ping a non-management ip for establishing an HA heartbeat.
this is a very common issue and many times I do get the same problem. However it is non-blocking, here is what I do often,
- I apply some HA customization
- dictate HA which management network to use in vcenter settings (if am not wrong then it should be the 'das.allowNetwork....'
However if I can ping both management ips from each host, I would not worry too much
Setting das.allowNetwork did the job. I added my 2 Management network and reconfigure HA.
No more warning
cool... glad it helped
cheers