Thanks Tony. Yes, this is a development test server where they use our DB
installer often to recreate the DB for the next round of testing. So,
that's why you see 1-38.
The real purpose of "*restore archivelog from time 'SYSDATE-14' validate;*"
is just for a validation step. Ahhh, I see what you are saying though.
They are not there and the catalog is sync'ed, but the script is
specifically telling it to restore/validate 14 days worth? Hence, the
error?
On Wed, Oct 7, 2015 at 8:08 PM, De DBA <dedba@xxxxxxxxxx> wrote:
Hi Woody,
"Crosscheck archivelog all" does sync the catalog. You will probably see
that they are deleted when you do a list. RMAN is asking for them, because
you are specifically asking them to be restored. By the way, 1-38 are
really low sequence numbers for archived logs. Is this a new database, or
is it recreated often?
I'd suggest that you in the meantime try to manually restore the archived
logs with decreasing periods. At least you should be able to sucessfully
restore archivelog from time 'sysdate - 6';
Cheers,
Tony
On 08/10/15 10:01, Woody McKay wrote:
Thanks Tony. You've mentioned the same as Andrew. I'll for sure create a
ticket to get that changed for the next release.
This is a development server and there is nothing for restoring archived
logs. I thought the restore archive validate was just to validate that the
backed up archived logs were good?
I'd like to just somehow tell the rman catalog that the archive logs are
gone (seq 1-38) and start with the next backup. I was thinking that
"crosscheck
archivelog all" would sync the catalog, but it did not.
Is there another crosscheck command?
On Wed, Oct 7, 2015 at 7:55 PM, De DBA <dedba@xxxxxxxxxx> wrote:
First thing that strikes me is that you configure a recovery period of 7
days, then backup and delete obsolete, then hope to restore archived logs
from 14 days back. RMAN retains only as many backups as necessary to
restore the *database* to up to 7 days ago, and in your script you
create a new level 0 backup. The backups of more than 14 days ago will have
been deleted, as you run this weekly, so the oldest archived logs in a 14 *
24 hour period could be gone - it looks a bit like a race condition really.
...
configure retention policy to recovery window of *7 days*;
...
*backup* as compressed backupset incremental *level 0* cumulative tag
'L0'database
plus archivelog delete all input;
...
delete noprompt *obsolete*;
...
* restore archivelog from time 'SYSDATE-14' validate; *
Did you try to restore archived logs of less than 14 days ago, e.g. 13.5?
Cheers,
Tony
On 08/10/15 09:27, Woody McKay wrote:
Hi, ( note, cross posted in Oracle community Database Backup and Recovery
)
I'm not super-versed with RMAN but am trying to resolve an rman issue on
a development server. The DB environment can be rebuilt if needed, but I'd
like to try and resolve this with a crosscheck command, etc.
The weekly rman backup script looks like:
run {
configure controlfile autobackup on;
configure controlfile autobackup format for device type disk to '%F';
configure retention policy to recovery window of 7 days;
allocate channel d1 type disk maxpiecesize=5G;
backup as compressed backupset incremental level 0 cumulative tag
'L0'database
plus archivelog delete all input;
restore database validate;
release channel d1;
delete noprompt obsolete;
restore database check logical validate;
backup validate database;
*restore archivelog from time 'SYSDATE-14' validate;*
host 'copy C:\oracle\product\12.1.0\dbhome_1\NETWORK\ADMIN\tnsnames.ora
C:\oradata\fast_recovery_area\';
host 'copy C:\oracle\product\12.1.0\dbhome_1\NETWORK\ADMIN\sqlnet.ora
C:\oradata\fast_recovery_area\';
host 'copy C:\oracle\product\12.1.0\dbhome_1\NETWORK\ADMIN\listener.ora
C:\oradata\fast_recovery_area\';
host 'copy C:\oracle\product\12.1.0\dbhome_1\database\pwd*.ora
C:\oradata\fast_recovery_area\';
}
list incarnation of database;
list backup summary;
list backup by file;
list recoverable backup of database;
report schema;
report need backup redundancy=3;
report need backup recovery window of 3 days;
The command "restore archivelog from time 'SYSDATE-14' validate;" is
throwing the below error:
RMAN> restore archivelog from time 'SYSDATE-14' validate;
Starting restore at 07-OCT-15
using channel ORA_DISK_1
RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03002: failure of restore command at 10/07/2015 15:15:49
RMAN-06026: some targets not found - aborting restore
RMAN-06025: no backup of archived log for thread 1 with sequence 38 and
starting SCN of 4330735 found to restore
RMAN-06025: no backup of archived log for thread 1 with sequence 37 and
starting SCN of 4324802 found to restore
RMAN-06025: no backup of archived log for thread 1 with sequence 36 and
starting SCN of 4311903 found to restore
RMAN-06025: no backup of archived log for thread 1 with sequence 35 and
starting SCN of 4230358 found to restore
...
Note, that "RMAN> restore archivelog from time 'SYSDATE-1' validate;"
does work successfully.
I did find and removed on obsolete backup set, but that didn't resolve
the 06026 error.
I tried "crosscheck archivelog all;", but that does not resolve the 06026
error during the restore archivelog validate.
Is there a crosscheck command or other command that would sync the
catalog and disk to resolve the 06026?
Thanks for all replies.
--
Sincerely,
WoodyMcKay
--
Sincerely,
WoodyMcKay
