Database corruption is something that you hope you never have to deal with, but is eventually going to happen to everyone if you deal with databases long enough. In this article from Brent Ozar, we get his take on the steps to resolve the issue, as well as when and how to communicate the status of the issue to your team.
- Query the msdb.dbo.suspect_pages table.
- If it’s just one database, send an email to your manager and your team.
- If you’re using a SAN, alert the storage team.
- Alert the application owners and stakeholders.
- Turn off your backup-delete jobs: you’re gonna need ’em.
- Review the history of your CHECKDB and backup jobs.
- If you have the full CHECKDB output, look for an easy fix.
- Fork your recovery efforts into multiple teams.
- Send out a status update as work starts.
- Keep working the plan.
You should read the entire article for all the details.