• Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 714
  • Last Modified:

Can Increasing SGA size cause paging !

The SGA size was 5GB and database was running very fine with no performance issues. we changed to the new server with 16 GB of RAM on AIX 5.3 and we allocated 12 GB for SGA MAX SIZE. and allocated 10GB to shared pool. we are experiencing high paging when running RMAN backup's and even a simpleselect statement takes 2-3 minutes(like select name from v$database). As I allocated this memory to the new server and is in production I am  not sure that sga can be a problem.

when we run backup's the disk IO is more and the server is paging.
0
ballioballi
Asked:
ballioballi
  • 2
1 Solution
 
Franck PachotCommented:
Hi,

First I don't understand why you increased the SGA as it was running very fine with no performance issues...
10GB for the shared pool is so huge ! Do you expect to have 1GB of sql statements (not data, statements) frequently used ?

An oracle server should not swap. You need to have all memory structures (SGA but also PGA) in physical memory.

Regards,
Franck.
0
 
woolmilkporcCommented:
Hi,
if it's AIX, use 'vmo -a'  to examine the minperm%, maxperm% and maxclient% settings.
From IBM -

minperm% specifies the point below which the page-stealer will steal file or computational pages regardless of repaging rates.  
Can be useful to decrease this parameter if large number of file pages in memory is causing working storage pages to be replaced
.  
maxperm% specifies the point above which the page-stealing algorithm steals only file pages. Reducing this value may reduce or eliminate page replacement of working storage pages caused by high number of file page accesses.
maxclient% specifies maximum percentage of RAM that can be used for caching client pages. Same as maxperm.
In other words - decreasing minperm% and maxperm% leads to working storage pages ('COMP' in topas. e.g. SGA) being left in memory for a longer time, because primarily file pages will be paged out. Thus paging to swap will be reduced, as file cache shrinks.
You can experiment with minperm%/maxperm%/maxclient% by beginning with e.g  'vmo -o minperm%=15 -o maxperm%=25 -o maxclient%=25' and watch the outcome. Reduce it further, e.g. to 10%/20%/20% if needed. Leave maxperm%/maxclient% roundabout 5%-10% higher than minperm%
To keep the new values permanently across reboots, issue e.g. 'vmo -p -o minperm%=10 -o maxperm%=20 -o maxclient%=20'
I have systems here running at a minperm% value of as low as 3% with no huge adverse effect on file I/O performance.
wmp
 
 
0
 
ballioballiAuthor Commented:
did not get the exact answer
0
 
woolmilkporcCommented:
What would have been the exact answer then?
0

Featured Post

Free Tool: IP Lookup

Get more info about an IP address or domain name, such as organization, abuse contacts and geolocation.

One of a set of tools we are providing to everyone as a way of saying thank you for being a part of the community.

  • 2
Tackle projects and never again get stuck behind a technical roadblock.
Join Now