Web Operations

Database Tactics

Now that we've covered a variety of strategic approaches to create a robust database architecture and seen some things to avoid, let's talk about database tactics. These are day-to-day operational tasks you'll perform to keep that database infrastructure reliable.

This section will focus on getting things done. In contrast to the previous discussion of architecture, where almost everyone reaches intuitively for a handful of obvious and sometimes wrong solutions, there are lots of different ways to do things—so many that I won't spend any time urging you away from long lists of particular tactics. I'll concentrate instead on a few specific things that I think are good to do.

Taking Backups on a Slave

I haven't talked much about backups, so I hope you'll let me lecture just a bit before I go on to the real topic:

• Stop procrastinating on backups. They're not that hard.

• Don't build something awesome; just build something recoverable.

• Document at least the following: acceptable data loss, acceptable downtime, retention policy, and security requirements.

• Practice and document the recovery process. Recovery is more important than backing up!

• You need external verification of a backup job's success or failure. Don't rely on the job itself to notify you.

With the formalities out of the way, let's look at how to use a replication slave for backups.

The first and most obvious thing is to use the slave itself as a backup. Unfortunately, this isn't a real backup. A real backup protects ...