The Ol' PLE

  • Hi Folks

    Setting:

    2 node 4 Instance Cluster

    All SQL Instance on 10.50.4000

    Node 1: SQL1

    Node 2: SQL 2, SQL3, SQL4

    64GB Host memory

    All Instances host User DBs and applications except SQL4 - just a spare Instance with the system, DBs..

    Each instance Max Memory set.

    Story:

    A few days ago, SCOM Started reporting PLE dropping below 300 consistently for SQL4.

    Applying the rule as laid out by Paul Randall ( http://www.sqlskills.com/blogs/paul/page-life-expectancy-isnt-what-you-think/ ), it still reports as anything between 1, yes, 1 and 1765.

    Moving this instance over to Node 1 resulted in nothing different, same results.

    The instance is not maxing out the allowed max memory either, it idles round the 536MB region.

    Every other SQL Instance PLE Reports massive (healthy) figures.

    FAQ Answers:

    No DBCC checks running, no backups running.

    Checking actively what is running, absolutely nothing.

    No processes, no jobs, no nothing.

    Absolutely dormant.

    Question:

    Has anyone ever noticed this behavior or am I missing a critical check to perform?

    As far as I can see nothing is using this instance and the PLE should be unaffected.

    -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
    This thing is addressing problems that dont exist. Its solution-ism at its worst. We are dumbing down machines that are inherently superior. - Gilfoyle

  • Henrico Bekker (5/22/2015)


    The instance is not maxing out the allowed max memory either, it idles round the 536MB region.

    How are you getting that figure?

    Gail Shaw
    Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
    SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability

    We walk in the dark places no others will enter
    We stand on the bridge and no one may pass
  • Hi Gail, just by looking at the service in Task Manager it's obvious it is idle.

    Any ideas?

    -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
    This thing is addressing problems that dont exist. Its solution-ism at its worst. We are dumbing down machines that are inherently superior. - Gilfoyle

  • Henrico Bekker (5/22/2015)


    Hi Gail, just by looking at the service in Task Manager it's obvious it is idle.

    Any ideas?

    Yeah, stop using a tool which mis-reports memory usage. 🙂

    Task manager does not show memory allocated via the AWE mechanisms, which means if you use locked pages in memory (which I'm guessing you do), you will see only the memory which SQL allocates via non-AWE mechanisms, which is usually well under 1 GB, because it's just the thread stacks and other small usages.

    Use perfmon to track SQL's memory usage.

    Gail Shaw
    Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
    SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability

    We walk in the dark places no others will enter
    We stand on the bridge and no one may pass

Viewing 4 posts - 1 through 3 (of 3 total)

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