Have you tried it? Is it not working or are you getting an error?

I was just trying to type Page Life Expectancy but Craig beat me to it. It's one of the best indicators for memory pressure.

Also, you should monitor the wait...

In addition to roles, also use schemas. That way you can grant execute permissions for a role to a schema. It makes things much easier to manage as Steve says.

Excellent selection on the feedback, Brad. I also like your advice on filegroups and files. It's not something I had considered before, but certainly will going forward. I can't wait...

There are a couple of ways you could try to fix this. Probably the easiest is to look at the column order on the primary key of the first...

Teh Gila Monster has teh big scares on me and makes me afeared fer me woefully inadequate sexual equipment.

I'M SPARTACUS!

Uh, I mean, I'm a Gila Monster too. Really.

One other note, you can't create a nonclustered materialized view. The very construct is always a cluster. No other way to do it.

Dale Cunningham (4/12/2011)

---

Thanks Grant for the reply. Yes the profiler can definately impact production. But I'm using the default trace in Profiler. It captures both the RPC statement begin and...

The thing to do is restore to a moment in time before the event, but do the restore on a separate database. But then, you have to figure out how...

What event are you capturing in the trace? When I run a trace on a database with forced parameterization set, I still see values passed in. If you're seeing that,...

You're asking for more control than is possible through SQL Server. DBCC is how you determine where data other than tables are stored. Tables storage is defined when you create...

Craig Farrell (4/12/2011)

---

Grant Fritchey (4/12/2011)

---

It really sounds like you've got a non-standard design, so I doubt any standard set of checklists on performance tuning is going to help. All the...

You've got a couple of options. My favorite is to use Dynamic Management Objects such as sys.dm_exec_requests or sys.dm_exec_sessions to see what's happening on the server. You can combine these...

A key lookup or a RID lookup, operation is caused when the nonclustered index that the optimizer uses doesn't have all the columns to satisify the query. The common method...