IndicatorsPool events causing memory pressure?

  • Hi,

    We are running a system on SQLServer 2017 Enterprise (RTM-CU22) with 839GB of memory.

    I think we appear to be experiencing some memory pressure as we seem to have plans dropping from the cache and I can see various clock-hands running (including CACHESTORE_OBJCP – EXTERNAL and CACHESTORE_SQLCP – EXTERNAL).

    When I check the Memory Clerks the bulk of memory is taken up by MEMORYCLERK_SQLBUFFERPOOL around 750GB, with about 26GB for CACHESTORE_OBJCP and CACHESTORE_SQLCP of around 8GB.  Typically no other clerks are over 1 GB.

    When I look in the RING_BUFFER_RESOURCE_MONITOR the only RESOURCE_MEMPHYSICAL_LOW events we see are for IndicatorsPool.

    I’m a bit at a loss as to working out what is causing the pressure and what the RESOURCE_MEMPHYSICAL_LOW for IndcatorsPool is telling me (and if this is related).

    Could anyone give me any pointers on where to look next?

    Thanks

  • Thanks for posting your issue and hopefully someone will answer soon.

    This is an automated bump to increase visibility of your question.

  • Is max memory for the sql server instance set to lower than  839GB, so other applications like the operating system can run?

    Is SSAS installed on the server? There was a blogpost that the default configuration reserves 20% of ram for SSAS. Which lives 80% for OS + SQL + other programs (backup, ssis, ...)

  • Thanks for the response Jo.

    The server has 960GB with the max memory for SQLServer set to 839GB.   That along with no indications in the RING_BUFFER_RESOURCE_MONITOR of IndicatorsSystem with RESOURCE_MEMPHYSICAL_LOW make me think it isn't external to SQLServer.

    SSAS is installed but disabled and not running.

     

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

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