Why does my MySQL instance crash for no good reason?
Posted on 2014-02-13
I've got a lot of ETL reading and writing to my DB all day, every day. It runs quickly but about 1x per day, MySQL just crashes. I just restart it when I notice, but then my ETL is backlogged. I've added a scheduled 'net start mysql' that repeats every 5 minutes on Task Scheduler, but that's a bandaid, not a fix.
If I check the *.err file at C:\Program Files\MySQL\MySQL Server 5.5\data, I don't see much of use in there. It'll say "InnoDB: Database was not shut down normally!" but that's it.
My guess is that these crashes happen because of highly concurrent reads/writes and also some nasty Select queries that are called by a script on Windows scheduler every now and then.
How can I go about getting a better clue of what causes the mysql service to just stop? Are there any settings in the options file can look at / adjust to help avoid crashes?