FAQ
Greetings

2 CPU HP-UX 11.11 Oracle 9.2.0.7 STANDBY Database

Applying 450 Archive Logs of 20 MB each took 2.30 hrs,

Is there a way to tune this to cut this time?

Search Discussions

  • Riyaj Shamsudeen at Apr 27, 2006 at 8:56 pm
    ('binary' encoding is not supported, stored as-is)

    I have seen *some* improvement in using parallel clause. Sorry, I didn't
    measure the performance improvement though.

    alter database recover managed standby database parallel 16;

    As I remember, one process reads the log file, but the application of
    those change vectors were performed by slaves.. Depending upon your
    bottleneck, this may or may not help you though.

    BN wrote:
    Greetings

    2 CPU HP-UX 11.11 Oracle 9.2.0.7 <http://9.2.0.7> STANDBY Database

    Applying 450 Archive Logs of 20 MB each took 2.30 hrs,

    Is there a way to tune this to cut this time?
    The information transmitted is intended only for the person or entity to
    which it is addressed and may contain confidential and/or privileged
    material. If the reader of this message is not the intended recipient,
    you are hereby notified that your access is unauthorized, and any review,
    dissemination, distribution or copying of this message including any
    attachments is strictly prohibited. If you are not the intended
    recipient, please contact the sender and delete the material from any
    computer.
  • Jeremiah Wilton at Apr 27, 2006 at 9:09 pm
    Here are some ideas:



    Try a variety of degrees of parallel recovery (recovery_parallelism
    parameter)
    Tune standby by examining wait events and addressing highest waits
    (if addressable)
    Check for buffer cache efficiency and if necessary, increase size on
    standby
    Determine top redo generating jobs and sessions on primary (using
    v$sesstat/redo size) and determine if they can be reduced or eliminated.

    Jeremiah Wilton

    ORA-600 Consulting

    http://www.ora-600.net



    From: BN


    2 CPU HP-UX 11.11 Oracle 9.2.0.7 STANDBY Database



    Applying 450 Archive Logs of 20 MB each took 2.30 hrs,



    Is there a way to tune this to cut this time?
  • Andrey Kriushin at Apr 27, 2006 at 10:07 pm
    Hi,

    Well, 450 "interval" checkpoints in 150 (or you mean 138?) minutes might
    me too much. Not only the server should write down all the dirty blocks,
    but also update all the datafile headers and the controlfiles. And until
    then it can't proceed with the next archived log.

    Suggestions for next time:
    1. trace MRP process to see what it is actually waiting for. Or you can
    monitor V$SESSION_EVENT
    2. agree - applying redo in parallel is rather good suggestion. Don't
    forget to increase PARALLEL_EXECUTION_MESSAGE_SIZE up its maximum
    3. add more DBWRs if you don't use AIO.
    4. increase the size of your online redo logs on primary DB in order to
    decrease the frequency of standby DB checkpoints.

    HTH

    - Andrey

    On Fri, April 28, 2006 0:32, BN said:
    Greetings

    2 CPU HP-UX 11.11 Oracle 9.2.0.7 STANDBY Database

    Applying 450 Archive Logs of 20 MB each took 2.30 hrs,

    Is there a way to tune this to cut this time?

    --
    Regards & Thanks
    BN
    --
    http://www.freelists.org/webpage/oracle-l
  • Joseph Amalraj at Apr 28, 2006 at 2:14 pm
    You can query on v$session_event or v$system_event and find out which event is taking the most time; and then try to reduce the time spent on that event.


    regards


    Joseph

    BN wrote:

    Greetings


    2 CPU HP-UX 11.11 Oracle 9.2.0.7 STANDBY Database


    Applying 450 Archive Logs of 20 MB each took 2.30 hrs,


    Is there a way to tune this to cut this time?
  • Alex Gorbachev at Apr 28, 2006 at 8:14 pm
    You didn't mention is this a logical or physical standby. If physical
    than tuning is the same as for database recovery - you can find plenty
    in the net as well as some sugestions already given in this thread.
    Loagical standby is a different beast.
    Anyway, sampling v$session_wait can give you good information on where
    your sessions (if parallel) are spending their time.

    2006/4/27, BN :
    Greetings

    2 CPU HP-UX 11.11 Oracle 9.2.0.7 STANDBY Database

    Applying 450 Archive Logs of 20 MB each took 2.30 hrs,

    Is there a way to tune this to cut this time?

    --
    Regards & Thanks
    BN
    --
    Best regards,
    Alex Gorbachev
    --
    http://www.freelists.org/webpage/oracle-l
  • Mogens Nørrgaard at Apr 29, 2006 at 7:51 am
    Could it simply be your IO-capacity that sucks?

    Mogens

    BN wrote:
    Greetings

    2 CPU HP-UX 11.11 Oracle 9.2.0.7 <http://9.2.0.7> STANDBY Database

    Applying 450 Archive Logs of 20 MB each took 2.30 hrs,

    Is there a way to tune this to cut this time?

    --
    Regards & Thanks
    BN
    --
    http://www.freelists.org/webpage/oracle-l

Related Discussions

Discussion Navigation
viewthread | post
Discussion Overview
grouporacle-l @
categoriesoracle
postedApr 27, '06 at 8:32p
activeApr 29, '06 at 7:51a
posts7
users7
websiteoracle.com

People

Translate

site design / logo © 2022 Grokbase