This is a quick after action report on restoring backups. This is mostly for future me to know what I did.
The backup scripts on the main server stopped working. The IT person at that company had already figured out this was because of the database not booting on the machines which were supposed to receive the backups, but figuring out the underlying cause was beyond his capability. So I was called in. Yes. I’m slowly turning into a database administrator at this rate.
So xampp1 told us to look at the logs, those said:
error: trying to access page number 123 in space 0
Stack overflow said that we need to recover from backup if
innodb_force_recovery doesn’t work. It doesn’t say how. Remember, we can’t even start the database! What I did instead was deleting the
innodb0 file, and the
innodb.log files. Remember, I actually don’t care about the data, it’s just a backup machine. This made the database boot again, making it ready to receive backups.
One more peculiarity occurred (of course). The script would timeout saying it’s a connection error. However this error happened every time I executed the script, and I’d assume that at an office the connection is pretty reliable. This made me think it was just mislabeling the error. Once again stack overflow agreed. This database has lots of blob data, so setting
64M indeed fixed the connection error.
I’m starting to think I should move this entire stack over to postgres, at least that doesn’t randomly start failing, or has mislabeled errors. But I know that’s next to impossible with this clients’ resources. So for now, Jappie is here for all your database concerns.
Why are they using xampp? Legacy choices. Remember I don’t get to chose their tech stack, that has been decided more then a decade ago. I’m just keeping it operational. ↩