Please start any new threads on our new
site at https://forums.sqlteam.com. We've got lots of great SQL Server
experts to answer whatever question you can come up with.
| Author |
Topic |
|
graz
Chief SQLTeam Crack Dealer
4149 Posts |
 Posted - 2004-01-22 : 10:35:27
|
| I was curious if anyone has any experience using this command. I've been running it and seeing some weird results.The two main columns it returns are Requests and Wait Time. My first thought was that Wait Time is in milliseconds. However the numbers I'm seeing are really, really large. I've got a number that's greater than 3 billion for a 600 second sample on an 8 CPU system. That seems to imply that SQL Server waited over 400,000 seconds per CPU during the 10 minutes (400,000 seconds ~ 5 days). And that can't be right.Has anyone used this successfully before? Am I making some simple mistake?I'm running DBCC sqlperf(waitstats, clear) to reset the counters, delyaing for 10 minutes and then running DBCC sqlperf(waitstats).Further notes: If I run this script:DBCC sqlperf(waitstats, clear)waitfor delay '00:00:01'DBCC sqlperf(waitstats)I get a "waitfor" wait time of 2,000 (although I'd swear it was 1,000 the first time I ran it). That seems to imply a 2:1 ratio. However a 5 second delay gives a value around 9,000. What gives?===============================================Creating tomorrow's legacy systems today.One crisis at a time. |
|
|
|
|
|
DBCC sqlperf(waitstats)