Hi,
I just took over this configuration and I figured Id ask this to ensure we dont run into any issues.
The Back up admin is executing a script on a secondary site to mount filesystems RO to back them up.
Aside from issuing the correct mount command if we have to do a takeover. Will this cause any other affects?
RHEL 6.2/SF 7.2 no HA
Given that VVR provides continuous replication, simply mounting a volume on Secondary without any quiesce/unquiesce wrappers will likely produce just gigabytes of garbage instead of a correct backup, due to files changing while backup is happening.
The correct way to perform a backup through VVR is to create a snapshot at the RVG level using something like vradmin ibc or vxibc, check them out, then split the snapshot volume(s) off, mount them (r/o or r/w your choice) and then join them back again and resynchronise utilising the Fast Mirror Resync feature
Mounting a file system on secondary ver server is not recommended. The file systems on any volumes in RDS are considered as an "application" to the volumes in RDS on secondary, just like a database application. This is why in most of configuration, disk group resources and RVG resources and IP resources are in one service group while applications including file systems are in a parent service group which is kept offline during normal data rep mode. Also mounting a file system in RO mode on secondary changes file system metadata (like access time), this is not desired.
Frank
