Hi all, End up the performance issue has been solved by increasing the memory_target from 2 gb to 7 gb. The batch job running time is 40 min faster than on AIX. So it is still pointed back to buffer cache size, not OS i/o. Thanks, Joan From: Michael McMullen [mailto:ganstadba@xxxxxxxxxxx] Sent: Tuesday, March 26, 2013 1:38 PM To: oracle-l@xxxxxxxxxxxxx; Hsieh, Joan Subject: RE: linux Async io I've been using SLOB as a benchmark for all new server builds. It makes it nice, simple and repeatable so that you can work with sysadmins/storage to troubleshoot problems. > From: Joan.Hsieh@xxxxxxxxx<mailto:Joan.Hsieh@xxxxxxxxx> > To: mwf@xxxxxxxx<mailto:mwf@xxxxxxxx>; > frits.hoogland@xxxxxxxxx<mailto:frits.hoogland@xxxxxxxxx>; > david.barbour1@xxxxxxxxx<mailto:david.barbour1@xxxxxxxxx> > CC: ORACLE-L@xxxxxxxxxxxxx<mailto:ORACLE-L@xxxxxxxxxxxxx> > Subject: RE: linux Async io > Date: Fri, 22 Mar 2013 20:00:57 +0000 > > Thanks Mark, > > I'm curious the sync setting, but we are actually experiencing the > performance issue, 60% of wait time is user io, db_ sequential_read is on top > of it. This is a show stopper for us, and I am trying to find out the cause. > > Joan > -- //www.freelists.org/webpage/oracle-l