Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Re: Shutdown Abort and Backup

Re: Shutdown Abort and Backup

From: Matthias Gresz <GreMa_at_t-online.de>
Date: 1998/03/04
Message-ID: <6djsof$1ho$1@news00.btx.dtag.de>#1/1

On Tue, 3 Mar 1998 12:11:19 -0000, "Doug Burns" <dougburns_at_email.msn.com> wrote:

>
>Matthias Gresz wrote in message <6ddopm$l89$1_at_news02.btx.dtag.de>...
>
>But wouldn't you just be using the backups for one of 2 reasons
>
> 1) To recover to the point in time when the backup was taken, in which
>case you restore the online redo log files.
> 2) For complete recovery of all committed transactions, in which case
>you wouldn't.
>
>In any case, Oracle say you must not do this even if you are not in
>ARCHIVELOG mode, in which case you wouldn't be doing any point-in-time
>recovery anyway ;-)
>
>>If you backup this way you'll always had to restore also your online redo
>logs to be able to recover automatically. So you'll never be able to recover
>you db close to point it crashed. Assuming your
>>db is running in archive mode and you've three online redo logs your state
 while backing up will be:
>>
>>archive logs redo logs
>> 230 231
>> 229 232
>> 228 233
>> ...
>
>Surely this would be a bit more like ...
>
>Archived Online
>232 233
>231 232
>230 231
>
>With 233 being the only log that is required to recover from the SHUTDOWN

Wrong! An archived redo log can never be online! What's about a transaction that takes as long that all your three online redo logs are filled?

>ABORT
>
>>
>>CRASH
>>
>>archive logs redo logs
>> 410 411
>> 409 412
>> 408 413
>> ...
>>
>>After your restore
>>
>>archive logs redo logs
>> 410 231
>> 409 232
>> 408 233
>>
>>
>>
>>How to recover now? You'll only be able to recover till log 233. All
 further changes will be lost.
>>
>>For bringing down your db to backup try this:
>>
>>
>>CONNECT <user>/<pwd>
>>SHUTDOWN IMMEDIATE;
>
>
>I have definitely had problems with IMMEDIATE on many occasions and having
>backups that haven't run, along with a database which is in a shutting-down
>state whenI come in the next morning is a non-starter on a production
>database! If I want to avoid using abort, I use STARTUP FORCE RESTRICT
>followed by a STARTUP (which does the instance recovery) and then a SHUTDOWN
>NORMAL.
>
>>STARTUP RESTRICT;
>>ALTER SYSTEM SWITCH LOGFILE; --Archive online redos
>>ALTER SYSTEM SWITCH LOGFILE;
>>ALTER SYSTEM SWITCH LOGFILE;
>>ALTER SYSTEM SWITCH LOGFILE;
>>ALTER DATABASE BACKUP CONTROLFILE TO TRACE;
>>SHUTDOWN;
>
>

--

Regards

Matthias Gresz    :-)

GreMa_at_T-online.de
Received on Wed Mar 04 1998 - 00:00:00 CST

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US