Re: Hot backups : Catching the last Archived redo log at the end of the backup cycle.

After backing up all the tablespaces, one could temporarily stop the archiving, report the archived redo log, restore the archiving, then just backup the reported files.

on the sh side

sqldba lmode=y << EOF
connect internal
archive log stop;
exit
EOF REDOS= 'ls/whatever path/arch*.dbf' ; export REDOS

sqldba lmode=y << EOF
connect internal
archive log start;
exit
EOF tar -rvf /whereever $REDOS

Hope this could help
Renauld

joseph Li wrote in message <6gnnsv$r82$1_at_otis.netspace.net.au>...
>I usually do not copy the archived redo logs as part of the Hot backup,
>I just leave say a week's worth of archived redo logs on disks, so that
>when I do need to restore I only have to restore the database files from
>tape. Having said that, the archived redo logs on disks can also be
>backed up to tape as part of Operating System backup. This is assuming
>you have enough disk space to store a couple of days worth of archived
>redo logs.
>
>Regards
>
>In article <01bd623d$e8c602e0$6e6bcd8e_at_forrer>, "Rob Forster"
<robertf_at_inforamp.net> says:
>>
>>After one completes backing up their last tablespace, how can we be sure
>>that when copying the last redo log we do not catch ARCH when its in the
>>middle of archiving the last redo log?
>>
>>alter tablespace begin backup;
>>! cp etc.
>>alter tablespace end backup;
>>alter system switch logfile; -- arch starts archiving the latest redo log
>>that was forced by this command
>>! cp ...
>>-- are we going to be copying an incomplete archived redo log here?
>>
>>TIA,
>>Rob.
>>
Received on Thu Apr 16 1998 - 06:16:16 CDT