Re: Need help with redo log writer

Flyfish wrote:

> I have a DSS system on a HP9000 (K 460) running HPUX 11.0, Oracle
> 8.1.6.0.
>
> Redo log write time is "high" per Spotlight on Oracle monitoring.
> Averages between 60-120+ ms.
>
> Redo logs are on a striped logical (3 physical drives, 10K rpm scsi-2)
> volume with low disk I/O according to sar (7-10% utilized)
>
> The redo buffer is set to 1536K.
>
> The database is not in archivelog mode.
>
> What can I do to reduce the redo log write time? Would increasing or
> decreasing the size of the redo log buffer have any impact?
>
> thanks in advance
>
> Dave Bottom
> dave_l_bottom AT keane-nne DOT com
>

My expirience with high volume, high speed (automated) transactions, redolog writing hurts when sar reports ~8% or worse. The (unfortunate) raid 5 - is it striped at the same size as the redolog write size?. To my knowledge HP writes in n*1K chunks. You have a problem if commit rate is high => 1k (small) writes and the stribe size larger than say 1K. Then one write is converted to two reads (to get 'old' data), then merge the 1k into a larger striped block, write this new block and the calculated checksum back. Short, one write gives two reads + two writes. And there must be at least one disk revolution between read and write. A god write cache can help - but not create miracles. A setup like the one outlined here, cant be efficient.

/Svend Received on Fri Jan 18 2002 - 13:04:41 CST