Info
Here's the restore checklist I promised, short version:
- Stop cron jobs and the queue worker first
- Restore the database, then check row counts on a few key tables
- Restore uploads and fix file permissions
- Clear the cache and log in as a regular test member, not as admin
Add your own steps in the comments and I'll fold them in.
Adding one I forgot: check that outgoing email works after the restore. Members won't tell you when notifications stop arriving, they just drift away.
Another one: compare the size of the restored uploads folder with the original. A mismatch is the quickest sign that something was skipped.
And write down how long each step took. The next time something breaks for real, you'll know whether you're looking at ten minutes or a whole afternoon.
Last one, I promise: keep the checklist somewhere other than the server you're restoring.