Tom LaRock's (Blog | @SQLRockstar) latest Meme Monday challenge asks: "how many things can go wrong with a database server that are NOT related to a disk drive?" and invites SQL bloggers to list 9 problems in an attempt to come up with 99 total. Here are some things I've seen over the years at different places I've been:

1. Access rights - too much, too few, & everywhere in between
2. Not understanding that replication turns one update statement affecting 500K rows into 500K update statements that affect one row each
3. Not understanding why 500K update statements affecting one row each don't replicate immediately
4. WITH (NOLOCK)
5. Ambiguous schema design
6. Not paying attention to alerts (if they're even set up to begin with)
7. Default database growth settings
8. Not setting max server memory (especially on x64 instances)
9. Dynamic SQL

Why stop at 9? Here's a few more fun ones…

1. ORM tools…they're getting better, but jeez some of the first iterations of these generated some nasty, horrible SQL (see #9)
2. Not taking backups
3. Not testing restores when you actually do take backups
4. Index hints on every table referenced in a select statement (do you really think you're that much better than the query optimizer?)
5. SELECT * FROM…
6. Error messages displayed on public facing web applications which include the connection string (bonus: showing the username and password for the SQL login in plain text)
7. Public facing web based "tools" which let you enter T-SQL in a text box and execute it against any one of the servers shown in a dropdown list…(yes, unfortunately I really did see this once)
8. Public facing access to SQL Server instances
9. SQL injection

Tag, you're it! What kinds of bad stuff have you seen before?