Re: How to check the corruption in online redo logs and archived logs

ARCH occasionly can create a corrupted archive log from a perfectly good redo log. I have this happen occasionly and only discover it when the archive is applied to a standby DB. There is an option ( I believe un-documented ) to the alter system command that can check for a corrupted archive log:

alter system dump logfile '<archive log name>'

This reads the archive just like in a recover operation but doesn't apply it to anything - it simply reads through the archive. If it's corrupt, you'll get an error.

HTH

--
Jim Gregory
Principal Consultant for Keane, Inc.
Currently assigned to NCR, Dayton, OH

Opinions are my own and do not reflect those
of my employer or clients
"Howard J. Rogers" <howardjr_at_www.com> wrote in message
news:3aaf730c_at_news.iprimus.com.au...


> I am probably living under a misapprehension, but I could have sworn that

 if


> an archive is successfully taken, it is practically guaranteed

> clean -because ARCH doesn't just do an O/S copy of the redo log files, but

> intelligently reads them.  To the extent that if member 1 of group 1 is

> corrupt, ARCH will automatically switch to reading member 2 of group to

> complete the archive.

>

> And if all members of the group are corrupt, ARCH refuses to complete the

> archive, such that the entire database hangs when the log switches come

> round on top of the corrupt group.

>

> To that end, I wouldn't stay awake at night worrying about it.

>

> Regards

> HJR

>

>

> "Chuan Zhang" <chuan_at_asiaonline.net> wrote in message

> news:l_xr6.13$Nm5.65601_at_news.interact.net.au...

> >

> >  Hi, all

> >

> > As I know in Oracle 8.0.6, at least, we can set up log_block_checksum

> > initialiazation parameter to make sure archived log files are clean. But

 in


> > Oracle 8.1.5, this parameter is obsolote.  Are there any ways to make

 sure


> > my archived log files can be used for recovery WITHOUT using RMAN?

> >

> > Thanks for your help.

> >

> > Chuan

> >

> >

> >

>

>