What do I monitor to prove the server needs more memory

  • I don't see Page Life Expectancy in my list of objects in PerfMon? I'm on version 2.0 do I have to upgrade to 3.0?

  • I'll agree that Page Life Expectancy is a good measure of memory pressure.  Values consistently below 300 represent a possible memory bottleneck.  I usually look at this along with Lazy Writes/sec.  This is a measure of how often the Lazy Writer process writes dirty data pages to disk in order to make room in the buffer cache.  This should remain as close to zero as possible and levels consistently above 20 represent a possible need for more memory.  Buffer Cache Hit Ratio is another good measure; however, be cautioned that unlike most metrics, BCHR represents a running total since the last time your SQL Service has been cycled.  So if you want to include this in your analysis, make sure you re-start SQL Server so that you are getting an accurate measure.  By the way, this thread was posted to the SQL Server 2005 forum and this is a SQL Server 2000 question. 

    Markus - Memory is cheap.  I would request more than 1 GB.  RDMS database servers thrive on memory - the more the merrier.

    Ron - page life expectancy can usually be found under the SQLServer:Buffer Manager object. 

     

    John Rowan

    ======================================================
    ======================================================
    Forum Etiquette: How to post data/code on a forum to get the best help[/url] - by Jeff Moden

  • I was able to add 1 gig of memory to the server and it does appear to have helped some. We also took a look at a long running process and tweaked it a tad which also helped. It also appeared that on the app server side there was an issue which was mostly the problem in the first place.

Viewing 3 posts - 16 through 17 (of 17 total)

You must be logged in to reply to this topic. Login to reply