Hi
I'm a little rusty with HACMP, but wanted to find out if it is possible to remove a disk heartbeat network from a running HACMP cluster.
Reason is, I need to migrate all the SAN disk, so the current heartbeat disk will be disappearing. Ideally, I'd like to avoid taking the cluster down to reconfig the heartbeats.
I'd also need to add a new disk heartbeat back afterwards - so is that also possible on a running cluster?
np with that, ideally add the new HB network before removing the old one.
Make sure you have PV IDs assigned to the new HB disk and known to both systems, run a discovery from within HACMP and create the new disk HB network. Once you then synchronize the cluster, the new HB network becomes active. You can then go ahead and remove the old disk HB network, synchronize again and you are done.
But, sooner or later I'd schedule some downtime to sort out the LUN numbering, to have it continuous again from hdisk0 up. As long as you work with the system its ok, but imagine someone else (not knowing how big your organization is) works with the system, removes all drives and cfgmgr's them again in, they will be differently named. HACMP is stupid though and does not rely on the PVID, but the hdisk name. The previous HB disks are gone/renamed, you waste time on troubleshooting why the cluster doesn't start...
1 Like
Thanks for this. Should offer some more options to make the required work a little easier. Agree with what you say about the hdisk numbering. If I remove the disk HB networks (temporarily of course), I think that is the only dependency HA has to actual hdisk numbers. The HA resource groups only contain the VG names, so it should be possible to clean up the hdisk numbering on one side of the cluster at a time (while HA is stopped there) - and then add back the HBs with the new numbering in place.