Well, we're running a hardware, three disk RAID5, on an
sym53c896-0-<2,*>: FAST-20 WIDE SCSI 40.0 MB/s interface. This is a
Compaq 3500 system with a CR3500 raid controller. An md5sum of a 1.2Gig
file takes less than two minutes.
We tried rebuilding the indices but that didn't seem to help. We had an
outside consultant do the rebuild and he's not available now so I don't
know what command he ran to do it.
I've never used 'sar'. If you can believe it, I've been poking around
with Unix for the last 20 years and I've never even seen the 'man' page
for 'sar'. I probably should look into it. What flags would give me
the most information to help figger out what's going on here?
Of course the troubleshooting increment is going to be a month or more
so this will probably take some time to resolve :-)
Cheers!
Bob
On Tue, 2002-07-23 at 14:08, Naomi Walker wrote:
>
> >Nightly, we're doing a 'vacuumdb -a -z' after stopping and restarting
> >the database. The performance will get so bad after a month that we
> >start to see load spikes in excess of 30. Normally, we don't see load
> >over 2.5 during the heaviest activity and generally less than 1.0 most
> >of the time.
> Typically, performance is linked to your I/O, but my first guess in this
> case has to do with your indices. As a test, next time performance gets
> really rotten, drop your indicies and rebuild them. It cannot hurt, and
> might just help.
>
> The trick here is to see what is happening while it is tanking. What does
> your disk configuration look like? Is it a raid or stripe where reads are
> spread out among more than one controller? Do sar reports point to
> anything in particular?
>
> ----------------------------------------------------------------------------
> ----------------------------------
> Naomi Walker
> Eldorado Computing, Inc
> Chief Information Officer
> nwalker@eldocomp.com
> 602-604-3100 x242
>
>
--
Robert M. Meyer
Sr. Network Administrator
DigiVision Satellite Services
14 Lafayette Sq, Ste 410
Buffalo, NY 14203-1904
(716)332-1451
