Re: LGWR: terminating instance due to error 340 - What's the cause?

Here's a followup on this: Two of the shutdowns generated trace files. Both of those trace files contained the following lines:

ksedmp: internal or fatal error
ORA-00340: IO error processing online log of thread

One of these listed a current SQL statement, while the other did not.

I hope this extra details helps with any suggestions anyone may have.

John Haskins <76054.334SPAMBEGONE_at_compuserve.com> wrote in message news:twzv4.566$1i.124560_at_dfiatx1-snr1.gtei.net...
> Gentlepersons:
>
> One of my Oracle instances has started shutting itself down, and I haven't
> been able to determine the reason. I'm hoping someone here will recognize
> the symptoms and point me in the right direction. Here are extracts from
 the
> alert log for the four times this has happened:
>
> Tue Jan 11 11:53:06 2000
> LGWR: terminating instance due to error 340
> Instance terminated by LGWR, pid = 471
>
> Fri Jan 21 10:05:53 2000
> Thread 1 advanced to log sequence 16165
> Current log# 1 seq# 16165 mem# 0: D:\ORANT\DATABASE\LOG2ORCL.ORA
> LGWR: terminating instance due to error 340
> Instance terminated by LGWR, pid = 379
>
> Wed Jan 26 12:55:51 2000
> LGWR: terminating instance due to error 340
> Wed Jan 26 12:55:52 2000
> Errors in file D:\ORANT\RDBMS80\trace\ORA00612.TRC:
> ORA-00340: IO error processing online log of thread
> Instance terminated by LGWR, pid = 514
>
> Wed Mar 01 14:46:56 2000
> LGWR: terminating instance due to error 340
> Wed Mar 01 14:46:57 2000
> Errors in file D:\ORANT\RDBMS80\trace\ORA00514.TRC:
> ORA-00340: IO error processing online log of thread
> Instance terminated by LGWR, pid = 462
>
> The Oracle documentation on this is very vague:
> Cause: An IO error occurred on the named online log.
> Action: Restore accessibility to file, or restore file from backup.
>
> There's no evidence that the log files are presenting any accessibility
> problems. The database uses four log files; the first two are 200K, while
> the last two are 10MB. This is an 8.0.5 installation running on NT. I've
> examined the server's event log and did not find any events that
> systematically happen just before the database is brought down.
>
> Anyone have an idea what's causing this?
>
> Thanks.
>
>
Received on Thu Mar 02 2000 - 00:00:00 CST