1 Nickel

Need help figuring out a snapshot work flow


I want to test new schema against our production databases, but obviously don't want to run it in production.

My plan is to:

1. Create Snapshot Mount Points for each prod server, and add ALL of them to the "test server" storage group

2. Create the snapshots for each server, then attach them to the SMPs in the "test server" storage group

3. Make sure I can mount each SMP/snapshot's file system, and point the TEST db server to the partition (as its DB directory)

4. Cycle out SMPs as needed to run tests against the different prod server snapshots

5. Update snapshots as needed to run tests against

Although I have something like this up and running, I feel like I might be running into trouble. I'm just basically using naviseccli to create/delete/attach snapshots to the SMPs, which are basically static for now. I run into trouble sometimes when I try to mount the filesystems on the test server, but mostly it works fine.

Is there anything I should watch out for? I looked at snapcli, but I don't want to run anything on the actual production servers, if possible. If there's a problem with synced data, or corruption, the db -should- be able to recover, and if not, I can always just take another snapshot.

I'd really love to hear if anyone is doing anything similar to this, and advise me.



0 Kudos