DBAsupport.com Forums - Powered by vBulletin
Results 1 to 10 of 44

Thread: Slow OLTP on 14CPU SunFire

Threaded View

  1. #17
    Join Date
    Jun 2005
    Location
    Yorkshire, UK
    Posts
    12
    Currently we have 2 x 140GB drives reserved for REDO directly attached.

    I'm very much considering moving to raw for redo logs. We have 6 independant drives (each 140GB) on a fibre back plane - direct attached, not SAN. I don't think the array controller is a bottleneck - it would not even break into a sweat with 6 drives.

    Q. How would I configure them? My thoughts are:

    Three sets of mirrored drives. Each mirror contains a log member.

    i.e.
    Drive 1 - REDO1A
    Drive 2 - REDO1B

    Drive 3 - REDO2A
    Drive 4 - REDO2B

    Drive 5 - REDO3A
    Drive 6 - REDO3B

    and then Drive 1 & 2 would contain REDO4A/B, drive 3 & 4 REDO05A/B and drive 5 & 6 REDO06A/B to give 6 members - each 256MB in size to keep the 15 minute switch rate.

    Archiving goes to a 470GB 9*72GB RAID5 array (and soon to a Standby server over the WAN)

    Afriad my SUN hardware knowledge is a bit basic - I have experienced engineers to do the physical stuff - but they don't know Oracle very well!


    So - do we think this REDO configuration allow us to do DML during the working day without killing the OLTP users?


    __________________
    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
    Guy Wicks
    Database Manager
    The British Library
    Last edited by hrishy; 06-06-2005 at 06:17 AM.
    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
    Guy Wicks
    Database Manager
    The British Library

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  


Click Here to Expand Forum to Full Width