Re: strange 'cannot allocate log, archival required' message

borism_at_ubitrade.com.au (boris milos) wrote in message news:<4e10c565.0211132234.1ae27c35_at_posting.google.com>...
> Hi all,
>
> This morning our oracle cluster failed to the secondary node due
> to some hardware issues. Since then, twice, an oracle error pop-up window,
> informed me that oracle "cannot allocate log, archival required"
>
> Automatic archiving is enabled, and so is the archiver.
> There is plenty of disk space in the archival destination,
> and the archive logs are continuosly being created.
> Database is not hung. And there is nothing in the alert log.
>
> With the first message, I assumed that it was a left-over from previous
> emergency, but the second one made me suspicous.
>
> System seems fine (users are not complaining), but still?
>
> Environment: NT4 server, Oracle 8.1.7.0.0, Oracle FailSafe
>
> Any suggestions are welcome,
>
> cheers,
> Boris

Your online redo logs are too small and/or too few in number. Hence, all your online redo logs need archiving at the same time. This is why you are getting this error.
You need to resize your online redo logs and/or add extra

alter database drop logfile group <n>
alter database add logfile group ('<filename1>','<filename2>') size <n> -- at least 5M, at least 4 groups
If your online log can not be dropped because it is in use alter system switch logfile
will remedy that.

This is a standard issue, so I'm 100 percent sure this will resolve it. Have needed to do this *many* times before.

Regards

Sybrand Bakker
Senior Oracle DBA Received on Mon Nov 18 2002 - 09:40:29 CET