What’s the right way to stop/ start cluster:
As per RH, we can bring down cluster services in reverse order BUT service group might failover to 2nd node which I think can be stop by freezing cluster (please correct me if I am wrong)
- , if you using high-availability (HA) services (
service rgmanager stop).rgmanager - , if you are using Red Hat GFS2
service gfs2 stop - , if you are using Red Hat GFS2 in conjunction with
umount -at gfs2, to ensure that any GFS2 files mounted duringrgmanagerstartup (but not unmounted during shutdown) were also unmounted.rgmanager - , if CLVM has been used to create clustered volumes
service clvmd stop service cman stop
OR can I simply stop by running “clusvcadm -s servicegroup” on both nodes (A&B) & restarts with “clusvcadm -e servicegroup -m nodename”.
Appreciate if anyone share their experience or best practice.
Thanks
Responses