Showing results for 
Search instead for 
Do you mean 

vmstat displays page faults

Occasional Visitor

vmstat displays page faults

2 clusterered Es80 running Tru64unix5.1B and
trucluster5.1B. The cluster system is based around Oracle 8.1.7 dbase with 2Gb SGA.
One node runs the applic and the other is in
standby mode ,doing nothing . BUT :
using vmstat 1 displays many page faults on the
idle system . ( 300 pf/s)
using vmstat -R displays a huge page faults on
the active system ( 38M RAD0 ..54G RAD1).
Although the system is running ,fixing this issue (PageFault) may increase performance .
Any suggestion is welcome .Is it Unix ,Oracle..? tuning issue?
Thanks


2 REPLIES
Honored Contributor

Re: vmstat displays page faults

Hi,

How many instances of oracle are running, as the instances number increase, swap must also increase.
Increase the swap size (add the additional swap)and check the vmstat
never give up
Honored Contributor

Re: vmstat displays page faults

Sam,

how long is the member up and running?
Do you refer to the summary right?
What are the numbers per second for that member?

greetings,

Michael
//Add this to "OnDomLoad" event