Re: redologs removed

Mark D Powell wrote:
> "Volker Hetzer" <volker.hetzer_at_ieee.org> wrote in message
> news:<bf5vnu$poh$1_at_news.fujitsu-siemens.com>...
>> Knackeback wrote:
>>> During a large update the redolog-files filled up my oracle
>>> partition
>>> completely.
>>> The database came to a standstill.
>>> I rebooted my machine but was still not able to start the database.
>>> Now I did a big big mistake - I removed the redolog files to get
>>> temporary space (my idea was to shrink datafiles after the database
>>> would
>>> be online again).
>>> The database now complains exactly about the missing redolog files.
>>> Is there a way to make the database fit for work without recreate
>>> it ?
>> Yes. Look up carefully what "alter database open resetlogs" means,
>> when to use it, its implications, what you lose and then do it.
>> And put your redologs on a different disks. I know, it's always a
>> fight with
>> the HW admin but one that should be fought.
>>
>> Lots of Greetings!
>> Volker
>
> Knackeback, Volker's advice may work if the 'current' redo log is
> still there, otherwise you would have potential corruption since
> current redo log activity may or may not have been written to disk.
> If this is a production db then your best option might be to perform a
> point in time recovery up throught the last archived redo log before
> the problem.

I had this problem once but I couldn't exactly recall what to do, only that resetlogs was involved in some way. If everything else is fine I think, what I did was a shutdown, possibly shutdown abort, startup and went from there. Typically, the db finds out about the missing logs and then you can start fiddling with alter database recover which you will find in the doc. I think, this is done when the db is mounted.

Hope this helps.
Volker

--
While it is a known fact that programmers
never make mistakes, it is still a good idea
to humor the users by checking for errors at
critical points in your program.
-Robert D. Schneider, "Optimizing INFORMIX
 Applications"