1 Reply Latest reply on Apr 15, 2008 10:09 AM by joshe

    Performance hit - disk activity

    wharfrat Novice

       

      I'm trying to figure out why certain processes take so long.  Unzipping an archive (~100 MB) on one guest took a really long time (over an hour,) and checking the size of folder (in Linux) by running du -sh went on for about 4 hours before I killed the process.  I later determined the size of the folder to be around 7 GB.  This was on a different guest vm.

       

       

       

       

       

      Our host is a dual processor Xeon (3.00 ghz) with 2 GB of DDR2-400 RAM running Archlinux 2.6.23.  Server 2 beta 2 is installed on the main drive, which is a Sata drive using ext3, but our datastores are on hardware raid 5 array running xfs.  The size of the array is 3 terabytes.  All virtual drives have been allocated statically.  Logging/debugging has been turned off.  There are only 3 virtual machines running at any one time, all running Linux without desktops, and nothing is going on that is cpu or memory intensive, just a couple of apache installations and one machine running nagios.  Load averages on the guest machines is below 1.  I'm wondering if the Raid Array might be part of the problem, or could it possibly be xfs?

       

       

       

       

       

      Thanks for any help/tips.

       

       

       

       

       

       

       

       

        • 1. Re: Performance hit - disk activity
          joshe Lurker

           

          I think debug is still on looking at the event log i see:

           

           

          Message from *****: You are running VMware Server with the DEBUG option. Please be advised that the additional logging and error checking enabled by this option result in substantially slower execution. This option cannot be disabled on this build of VMware Server.