Forums before death by AOL, social media and spammers... "We can't have nice things"
|    comp.databases.ms-sqlserver    |    Notorious Rube Goldberg contraption    |    19,505 messages    |
[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]
|    Message 17,666 of 19,505    |
|    Erland Sommarskog to Tim    |
|    Re: how can i figure out what is causing    |
|    08 Oct 09 21:56:08    |
      16a57409       From: esquel@sommarskog.se              Tim (t.hanson@faz.de) writes:       > recently we installed some software updates and noticed that on the       > average the values of the performance counter       >       > PerfCounter = [\SQLServer:General Statistics()\User Connections]       >       > increased roughly 30%. i'm trying to figure out how to measure what       > might have changed (and whether i need to worry). nothing else we       > measure with performance counters and graph with nagios seems to have       > changed significantly (cpu, memory, full scans, batch requests, ...).              In that case, I would not worry that much. It could just be a change       in application behaviour.              > to get a handle on CPU, reads, or writes, for example, i take brief       > profiler snapshots (sometimes restricted in some way, e.g. to queries       > running longer than half a second), then store these to a table,       > calculate a "factor" i can group by from the "TextData" of the       > operations and then look at statistics such as       >       > select top 25       > [Factor], sum([CPU]) sum_cpu       > FROM [   << oldest   |   < older   |   list   |   newer >   |   newest >>   ] |
(c) 1994, bbs@darkrealms.ca