Home » Infrastructure » Unix » Performance Issue on Sun Sloaris - Maybe Unix or Hardware Option ? (Solasris 8 / 10 Oracle 9.2.0.7)
Performance Issue on Sun Sloaris - Maybe Unix or Hardware Option ? [message #311460] |
Fri, 04 April 2008 09:57  |
Uwe
Messages: 260 Registered: February 2003 Location: Zürich, Switzerland
|
Senior Member |
|
|
Hi Folks,
I have a queation but I do not know in which board it will be the best option to post.
We have migrated an OWB Database from an older SUN Systemboard with 4*1,34 Ghz Processors and 32 GB Ram under Solaris 8 to a newer one with 12 CPUs with 1,8 GHZ and 96 GB Ram. Both were located on a Sun Fire 25k.
Formerly on the old borad we have had runtimes for Some Marts at around 25 Mins. On the new one it runs 5hrs plus (job was killed sfter 5 hrs).
Is there any way to analyze if the new hardware needs other parameter to set for oracle ?
We saw many "nested loops" and a huge work on TEMPSPACE. This seems to be one of the major factors for the runtimes, but why does it uses TEMP instaed of the Memory or fork over Processors ?
I do have the sql-scripts, but first I want to know if we could do something within the Hardware or Oracle configuration.
kind regards
Uwe
|
|
|
|
|
|
|
|
Goto Forum:
Current Time: Mon May 05 12:43:11 CDT 2025
|