8 Replies Latest reply on Jan 19, 2010 2:23 PM by NasaGeek

    9.5 and SQL Performance

      We recently upgraded our 9.1 SLX to 9.5. Since the upgrade, we have seen a major increase in the disk IO on our SQL server. THE Average Disk Que Length will get as high as 60 and will not return to normal. We have to start and stop the SQL service in order to get it go back to normal. It seems to be fine after that, but at some point it will get extremely high again (hours, sometimes days). We are not sure at what point this happens, or of a particular trigger.

      We did not make any changes before the upgrade and the DB server was handling the load just fine. We have two polling engines with ~7500 elements each. A dedicated SQL 2005 server, Dell 2950 III, 4GB RAM, with a 5 (+ hot spare) 10K disks in a RAID 5 used for the database, 2 disk RAID 1 for the OS and SQL server.

      We had just the wireless module, but we moved to a Lightweight solution and we really only monitor PING on the APs now. Just for fun, we unmanaged all light weight APs. This did not help.

      We had a ticket we closed, as we can't get any resolution. They are telling me to change to RAID 10, but I doubt the increase in performance will be enough to hide this problem. It really seems like there is a bug we are seeing.

      Any thoughts?