Re: Corrupt redo logs and datafiles

From: Rauf Sarwar <rs_arwar_at_hotmail.com>
Date: 18 Mar 2005 10:45:21 -0800
Message-ID: <1111171521.035110.309520@f14g2000cwb.googlegroups.com>

SG wrote:
> Using ext3 and using IDE hardware raid, 3Ware brand. Never had a
problem
> before with them. No ram module on raid card. Never tinkered with IDE

> parameters. Even with running unpatched kernel, could that becausing
corrupt
> redo log files? Doesn't seem likely, but I guess it's possible.
>
> Another thing of note, I get the following message in the
/var/log/messages
> all the time:
>
> "kernel: application bug: sqlplus(26572) has SIGCHLD set to SIG_IGN
but
> calls wait()."
> "kernel: (see the NOTES section of 'man 2 wait'). Workaround
activated."
>
> Can anyone explain this error message and is it a valid source to the

> corrupt redo logs?
> I guess changing the memory could be an easy thing to do to assess
the
> memory being the problem
>

A quick search on the metalink shows a bug 3662585 (Internal/Unpublished) and docid 531733.996 similar to this error message. You may want to contact Oracle support and open a TAR.

Regards
/Rauf Received on Fri Mar 18 2005 - 12:45:21 CST