Number 10 reminded of a recently learned lesson. I came in one morning and one of our major applications wasn't working....users could not log in . Of course, when an application isn't working everyone likes to blame the database. A quick sp_who2 and a check of CPU utilization and a few other indicators all said the database was purring like a kitten. And with pride I responded that this time, it wasn't the database. The application developers rolled up their sleeves and began pouring thru code. The network engineers were checking switches and the deployment team was checking recent changes. And I kept watching the database. Turning on profiler I saw RPC calls and SQL statements executing at a normal pace.

But then I look closer.....they were repeating the same queries over and over again. Hmmmm, wonder what that means. Finally, I tried executing one of the RPC calls directly from Query Analyzer. What the @#$%@! An error message: Time out occurred while waiting for buffer latch type 2 After 3 hours of declaring the database innocent, I was not ready to call it guilty. Fifteen minutes later and the outage was over.

The lesson: Never stop looking until the problem is found. If all indicators say the database is fine....check more indicators.

Fortunately, rather than being chastised for my hubris, I was praised for my perseverance.

Gordon