July 17, 2014 at 5:39 am
Hi guys,
Just a quick Q... should the clustered SQL Server instance have a dependency on the snapinfo LUN when using SnapManager for SQL?
I know the drive has to be in the cluster resource group, but can't see why there'd be a dependency too?
Cheers
July 24, 2014 at 3:09 pm
The question is, do you want to bring the SQL server resource online even if there is an issue with the snapinfo LUN?
If YES, then you do not need the dependency.
if NO, then you do need the dependency.
July 25, 2014 at 2:53 am
Cheers, the issue was that the LUN filled, went offline, and took the SQL instance with it!
We've removed the dependency, I just wasn't sure if it was a NetApp recommended practice or not.
July 25, 2014 at 9:02 am
Then remove the dependency.
In general you must be very careful using snapshots.
If your database has a lot of churn and frequent index maintenance you will require a lot of snapshot space.
Why did it fill up? how are snapshots configured? what kind of retention? If the snapshot volume is filling up, why does it not just throw away the oldest snapshots and send a warning?
Why are you using snaphots at all? Do you have a requirement to use them?
July 25, 2014 at 10:04 am
Thanks, we have done.
It filled with streaming backup files as snapshot backups aren't currently being used (system db file in the wrong place), plus the user db starting growing much quicker than anticipated.
Moving those system db files soon so won't be an issue then; in the meantime we've reduced backup rentention.
Viewing 5 posts - 1 through 4 (of 4 total)
You must be logged in to reply to this topic. Login to reply