It can affect the other processes, yes. No process you run is free of cost. Doing a scanned update of statistics is fairly low cost, and usually you can run...

For quick information on what's running slow, use sys.dm_exec_query_stats to pull out the most frequently called queries or the longest running queries. From there though, you need to know what...

Nope, not a single issue at all. Great way to do the upgrade. I've done... oh, a whole bunch that way.

However, you should run the Upgrade Advisor first because there...

The fact that you need a DISTINCT operation in there suggests you might have a structural or data problem you need to concentrate on first.

Also, why are you returning 50000...

Yeah, totally different. When I've run into that one it's usually been a configuration issue. Either I'm not running the right SP on the server, or the developers are running...

GSquared (2/6/2012)

---

Beyond what Grant says, MS recommends setting maxdop to 8 on servers with more than 8 CPUs, because the cost of spreading the job out will be more than...

I'd put them inside the statement too.

I don't remember how to do it, but it's possible to lose comments that are outside the statement (can't remember how that happens for...

I'd do a more thorough comparison between this box and the old one to detrmine what all the differences are.

As far as MAXDOP goes, I prefer to leave it set...

Something over the weekend took more resources than you have available. You just have to figure out what to start trying to troubleshoot the issue.

Is SQL Server alone on the...

Both occur during that period. To see if it's the exec plan, remove it from cache and then just ask for an estimated plan. However long that takes is your...

Understood.

I'm just nervous. We've had a...

Go easy on him. He's trying. Gentle guidance, not all up in his grill.

Changing it from non-clustered to clustered won't affect the data in other tables, no.

There are a ton of other affects you're going to see, yeah, but no, it shouldn't change...

No real way to give you solid advice based on generalities.

But, in general, test your queries. Concentrate on set-based solutions. Avoid row-by-row processing. Make sure the queries are avoiding all...

The problem with disk queue length on a san is that you don't have any idea how many disks are involved, so the queue count isn't a number that allows...