Post reply

One server 2 identical databases - 1 DB's performance is slower

  • April 26, 2008 at 2:20 am

    #121338

    Hi,

    We have live CRM server running Windows 2003 Enterprise SP1 with 2 x Quad Core processors and 22gb RAM. We have a live DB attached the the server that is constantly being used and we have a backup of this database attached to the server that is very rarely used, the problem I have is we are just implementing a NetApp SAN and are trying to performance test it. I have the rarely used DB attached on the NetApp disk (config below) but it is consistently performing my queries slower than the DB on the Direct Attached disk (details below)...Could there be a simple explanation for this??

    SAN DISK

    Clustered 3040 4gb Cache

    30 144gb 15K Fibre Channel Disks, Configured Raid-DP (NetApp RAID6)

    Connected using 2 x GB NIC's via iSCSI.

    Direct Disk

    8 x 144gb 15k SAS disks RAID 10 connected using Perc 5/i Array controller 256mb RAM - No read Ahead and Write Back.

    Now I would have hoped that the NetApp would have wiped the floor with the Direct Attached disk but it never does, I have had a look at the IOPS and it doesn't seem that there is anything too massive happening......I am pulling my hair out here any help would be greatly appreciated!

    Thanks

    Paul

  • April 29, 2008 at 9:24 am

    #808471

    Difference in RAID configurations may be one of issues.

    http://en.wikipedia.org/wiki/Standard_RAID_levels#RAID_6_performance

  • April 29, 2008 at 11:41 am

    #808580

    Caches may not be loaded on the rarely used DB.

    [font="Times New Roman"]-- RBarryYoung[/font], [font="Times New Roman"] (302)375-0451[/font] blog: MovingSQL.com, Twitter: @RBarryYoung[font="Arial Black"]
    Proactive Performance Solutions, Inc.     [/font]    [font="Verdana"] "Performance is our middle name."[/font]

  • April 30, 2008 at 1:58 am

    #808844

    Thanks for the replies.

    I am not sure that the RAID should be an issue we have 24 more disks, and although I genereally agree RAID6 has overheads NetApp design their filers to use this system (Raid-DP) so I think that should be faster!?

    Paul

  • April 30, 2008 at 2:03 am

    #808848

    Thanks for the reply.

    I am not an SQL expert, so dont really understand what you mean caches not loaded?

    Are the caches stored per database? All I was aware of was maybe Execution plans may be held in cache but thought after running the same query a few times that would negate this problem?

    Thanks

    Paul

  • April 30, 2008 at 2:02 pm

    #809307

    Paul Campbell (4/30/2008)

    Thanks for the reply.

    I am not an SQL expert, so dont really understand what you mean caches not loaded?

    Are the caches stored per database? All I was aware of was maybe Execution plans may be held in cache but thought after running the same query a few times that would negate this problem?

    Thanks

    Paul

    Lots of other things are held in cache also (indexes, tables' metadata, DB metatable info, table data, etc.). However, they also should be loaded after running the query a few times.

    [font="Times New Roman"]-- RBarryYoung[/font], [font="Times New Roman"] (302)375-0451[/font] blog: MovingSQL.com, Twitter: @RBarryYoung[font="Arial Black"]
    Proactive Performance Solutions, Inc.     [/font]    [font="Verdana"] "Performance is our middle name."[/font]

Viewing 6 posts - 1 through 5 (of 5 total)

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