Re: Redo Log Question

"Bruno Jargot" <see_at_reply.to.invalid> wrote in message news:s3bg0v42ab06s5q38i1rtmquki70nv8t0j_at_4ax.com...
> On 23 Dec 2002 00:57:51 -0800, Burt wrote:
>
> >Uhh ? Why wouldn't Oracle put the same damn corruption in both
> >copies? It did it to me once in version 7.3.2.3 and I spent the next
> >27 hours recovering. Corruption got in the datafile and REDO logs and
> >was also archived. We (myself and an Oracle analyst at Oracle support)
> >didn't know for sure when the corruption occurred. Got lucky doing a
> >point-in-time recovery with a guess.
>
> I don't think multiplexing redolog help against corrupted archivelog.
>

Oh dear. When are people actually going to read about these things and do some tests?

For the record, ARCH reads the first member of a log group and if it encounters corruption there, it switches to the second member. And if it encounters corruption there, it switches to the third member and so on.

Net result: it is perfectly possible to have two corrupt members of a group and yet achieve a non-corrupted archivelog copy. Provided the corruptions aren't in the same place of the same online log.

It's not 50-50 or 33-33-33 or any other 'take a chance' permutation.

> When archiving redolog (single destination), only one member will by
> archived. If a corruption by LGWR occured, there is 1/2 chance that
> the corrupt redolog is archived and as there are 2 writes instead of
> one (2 more chances to have a corrupted redo log), the risk of having
> corrupted archivelog with multiplexed redolog or with no multiplexed
> redolog is the same.
>
> So multiplexing helps against corruption in active redo logs but not
> in archived logs.

'Fraid not. Write back when you actually know what you're talking about.

Multiplexing gives you a chance of a good archive log when your online logs are completely stuffed.

Regards
HJR Received on Tue Dec 24 2002 - 04:46:13 CST