Today is Meme Monday and is a little thing that Thomas LaRock (Blog | Twitter) has started. Today he has started things off with a list of things that could cause a Database to go “wrong” that are not always disk related.
He has tagged a few people to add to the list in an effort to get up to 99 things that are unique that could cause failures in the database. I haven’t been tagged, but wanted to throw out a few of my thoughts to add to the list.
- invoices – missing, unpaid, or wrong altogether
- vendors (misinformed, ship wrong items)
- undefined requirements
- bad documentation
- rampant use of sa
- no backups at all (full, log, or otherwise)
- very wide clustered indexes (i.e. three guids make up the PK and Clustered Index)
- P2V of a VLDB (not always bad but can be disastrous)
- Catch-all queries
These are some of the things I have seen cause some database problems. Sure some of these things are more encompassing than just the database (like a vendor shipping the wrong part), but they can have a huge impact on the database.