Oracle FAQ | Your Portal to the Oracle Knowledge Grid |
![]() |
![]() |
Home -> Community -> Usenet -> c.d.o.server -> Re: Don't Use Shutdown Immediate for Full Backup
There are times where neither a shutdown normal or shutdown immediate will
return in a reasonable amount of time and a shutdown abort is required. If
a shutdown immediate or shutdown abort is required, follow that with a
startup restrict, followed by a shutdown normal. Then do the cold backup.
Robert Prendin wrote in message <34d7c61e.624317_at_news.istar.ca>...
>from personal experience I have not had a problem and have restored
>databases after a shutdown immediate and cold backups afterwards.
>
>Immediate is a much cleaner shutdown that abort....
>
>BTW, why not put the database in archive log mode and do hot
>backups.....
>
>Good Luck, Robert Prendin
>
>catcox_at_aol.com (Catcox) wrote:
>
>>The manual says not to do a SHUTDOWN IMMEDIATE if you are going to do a
full
>>cold backup of the database. However SHUTDOWN NORMAL isn't likely to
bring
>>the database down since invariably there are sessions hanging. Of course
the
>>backup is done in the middle of the night, and I won't be around to kill
any
>>sessions. I'm wondering what other people have done. Should I write a
script
>>to kill all sessions before the shutdown script runs or what?
>>
>>Below is the exact text from the manual.
>>
>>Oracle 7 Server Administrator's Guide (part no 6694-70-1292) from December
1992
>>on page 18-8 says
>>
>>Preparing to Take a Full Backup:
>>
>>To guarantee that a database's data files are consistent, always shut down
the
>>database in normal priority before making a full database backup; never
perform
>>a full database backup after an instance failure or after the database is
>>shutdown with immediate or abort priority (that is using a SHUTDOWN
IMMEDIATE
>>or SHUTDOWN ABORT statement). In both these cases, the data files are
likely
>>not to be consistent with respect to a specific point in time.
>>
>>
>>Thank you,
>>
>>Catherine Cox
>>catcox_at_aol.com
>
Received on Thu Feb 05 1998 - 00:00:00 CST
![]() |
![]() |