Re: ora-00354

On Aug 28, 10:06 am, Faraz <sarfarazla..._at_gmail.com> wrote:
> On Aug 28, 10:05 am, "fitzjarr..._at_cox.net" <fitzjarr..._at_cox.net>
> wrote:
>
>
>
>
>
> > On Aug 28, 8:26 am, Faraz <sarfarazla..._at_gmail.com> wrote:
>
> > > On Aug 27, 5:38 pm, sybra..._at_hccnet.nl wrote:
>
> > > > On Mon, 27 Aug 2007 10:30:49 -0700, Faraz <sarfarazla..._at_gmail.com>
> > > > wrote:
>
> > > > >Hi All,
>
> > > > >I have a produciton environment with data guard configuration which
> > > > >for little while pops out ORA-00354 - corrupt redo log block.
> > > > >upuntil today it was resolving itself with out any problems from the
> > > > >multiplex copy. but today after the error thre is a redo log missing
> > > > >on the dr system.
> > > > >Does copying redo log to standby database archive destination manually
> > > > >and try to resolve the gap in applied logs is a good idea??
>
> > > > >thanks for the help
>
> > > > IMO this is a rhetorical question.
> > > > The obvious answer is NO.
> > > > Especially as you state this is a *production* environment.
> > > > *Manual* Dataguard (as you seem to discuss) on a *production*
> > > > environment are two incompatible things, unless you have a trained
> > > > monkey who for a few bananas ships your archives and applies them.
>
> > > > --
> > > > Sybrand Bakker
> > > > Senior Oracle DBA
>
> > > Thanks Bakker,
>
> > > Well NO part is obvious as you say. How about some insight to resolve
> > > the issue?
> > > Although, moving log from production to standby database *MANUALLY*
> > > and then regestering it on standby database seems logical.
> > > - stop the recover process
> > > -copy and regester the redo log manulay
> > > start the redo apply. IT should apply the log and resolve the gap.
> > > in my openion it shold work but as i have this issue in production i
> > > was looking for your input / experties who possibly encountered this/
> > > similar situation and resolved it.
>
> > > again thanks for the post.- Hide quoted text -
>
> > > - Show quoted text -
>
> > Not if you copy the redo log. The automated process involves the
> > archivelogs and if you have corrupted redo logs then you need to
> > investigate why this is occurring as it's a source problem, not a
> > destination problem. Apparently you're having disk issues and
> > possibly need to consider replacing the errant device. Or you're
> > having memory issues and this is corrupting every copy of your redo
> > logs and thus affecting your standby.
>
> > I'd be seriously investigating this as you could lose your production
> > database, and you'd be left with a standby which may be unusable.
>
> > David Fitzjarrell- Hide quoted text -
>
> > - Show quoted text -
>
> Thanks David,
> since i have multiplexed logs so for now it looks like i am ok. what
> has been happening is i encounter this issue on one location and
> process successfuly reads the log file from the duplexed location. I
> am trying to find out tha root cause but in the mean while trying to
> prepare for the worst case which you described above and how to handel
> it.
> thanks for the help.- Hide quoted text -
>
> - Show quoted text -

Then it appears you have a disk which may be failing and I would seriously consider replacing it. How do you have Data Guard configured? This shouldn't be a problem if the other copies of the redo logs are usable.

David Fitzjarrell Received on Tue Aug 28 2007 - 10:13:54 CDT