Re: last latch ratio (redo copy) and general performance advice

In article <967321165.21325.0.nnrp-10.9e984b29_at_news.demon.co.uk>,   "Jonathan Lewis" <jonathan_at_jlcomp.demon.co.uk> wrote:
>
> You are worried because you have 3 redo
> copy latch gets which missed - but in the
> same report you have
> 184,000 redo allocation latch gets
> and
> 143,000 'nowait' redo copy latch gets.
>
> How much of a difference to performance is
> it going to make that 3 redo-related latch
> gets out of over 300,000 missed ?

I agree, Jonathan, but I'm worried because it's a consistent behaviour - while the db goes on, the redo-related latches increases, too, but remains in small numbers. So, as I said before, I applied the common answers (log size, log files, etc, )but it remais the same. To me, the problem resides on the hardware : because I have only a RAID-5 array (due to IBM hardware restrictions) and a local disk, the only choice for the redo files is in the local disk, aside the OS. Some loss in IO times and some latches/waits for will be inevitable, I think, while I fight for another disk controller hardware capable of RAID0+1 or RAID10.

Many thanks for ur attention.

[]s

      Chiappa

Sent via Deja.com http://www.deja.com/
Before you buy. Received on Mon Aug 28 2000 - 06:59:35 CDT