The indexing was bad. What little maintenance plans were in place had no re-indexing and such in place. The backups weren't being done in any scale to log growth for full recovery model. The logs were outsized for what was needed. The tempdb didn't have log files for the CPU's.

And then any SP I caught that was written poorly I sent in a change request to the dev's to fix. Not all of them were taken up, but I saw some of them changed in the next version.

But just the day-to-day fixing the broken issues for customers kept me busy so I couldn't truly afford to do a systematic top to bottom code review. :crying: