-
folks,
I don't know why there is 'Failed' message in my alert file below on primary site, but the alert file on standby site is OK and the managed rocovery worked well. You can find the archived log both located on directory of primary site and standby site. If you could explain it for me, I'd appreciate it.
a1netALRT.log locatd on pirimary site
Mon May 21 11:11:58 2001
Thread 1 advanced to log sequence 1401
Current log# 3 seq# 1401 mem# 0: E:\ORACLE\ORADATA\A1NET\REDO03.LOG
Mon May 21 11:11:58 2001
ARCH: Beginning to archive log# 2 seq# 1400
Mon May 21 11:11:58 2001
ARC0: Beginning to archive log# 2 seq# 1400
ARC0: Failed to archive log# 2 seq# 1400
Mon May 21 11:11:59 2001
ARCH: Completed archiving log# 2 seq# 1400
archive parameters in the initialation file: init.ora
log_archive_dest string
log_archive_dest_1 location=E:\Oracle\A1NET\archive
log_archive_dest_2 SERVICE=STBY MANDATORY REOPEN=60
a1netALRT.log locatd on standby site
Media Recovery Waiting for thread 1 seq# 1400
Mon May 21 11:11:35 2001
Media Recovery Log E:\ORACLE\ORADATA\A1NET\ARCHIVE\STBYT001S01400.ARC
Media Recovery Waiting for thread 1 seq# 1401
TIA
mike
-
This message isn't a bad message, it's just a warning.
Check your alert log to see how often it happens. It it's only occaisonal then it's not too bad, but could be avoided.
Use the times in the alert log to see how often your logs are switching, if they switch too often then increasing the size should help.
-
investigating ?
Hi, 21st May 2001 13:34 hrs chennai
Normally this is the way error occurs as per the oracle
http://www.oracle.com/oramag/oracle/...ml?39or8i.html
see what this tells on the same issue
http://www.fors.com/eoug97/papers/0142.htm
its really interesting lets have some more time to break the ice from some other angle.
Cheers
Padmam
[Edited by padmam on 05-21-2001 at 05:14 AM]
-
There is a serious error occured after automating recovery.
The following messages are from alter file on standby site:
Mon May 21 11:11:35 2001
Media Recovery Log E:\ORADATA\A1NET\ARCHIVE\STBYT001S01400.ARC
Media Recovery Waiting for thread 1 seq# 1401
Mon May 21 14:17:26 2001
Media Recovery failed with error 3113
ORA-283 signalled during: ALTER DATABASE RECOVER managed standby database ...
And on standby site I found the seq#1401 & 1402 archived logs transfered from the primary.
SVRMGR> select max(sequence#) from v$log_history;
MAX(SEQUEN
----------
1400
From above messages, I identified that the recovery process suspend by the number of log sequence of 1400. So I manually recovered it as the display bellow and it worked out.
Could you figure out why it can't be recovered automately but manually?
SVRMGR> recover standby database;
SVRMGR> recover standby database
ORA-00279: ?? 756950 (? 05/21/2001 11:11:58 ??) ???? 1 ????
ORA-00289: ??: E:\ORADATA\A1NET\ARCHIVE\STBYT001S01401.ARC
ORA-00280: ?? 756950 ???? 1 ???? # 1401 ???
archivelog: {=suggested | filename | AUTO | CANCEL}
record:
ORA-00279: ?? 758685 (? 05/21/2001 14:56:09 ??) ???? 1 ????
ORA-00289: ??: E:\ORADATA\A1NET\ARCHIVE\STBYT001S01402.ARC
ORA-00280: ?? 758685 ???? 1 ???? # 1402 ???
ORA-00278: ??????????? 'E:\ORACLE\ORADATA\A1NET\ARCHIVE\STBYT001S01401.ARC'
archived log: {=suggested | filename | AUTO | CANCEL}
BTW, environment which is Oracle 8.1.6 EE / WIN2K remains identical on both site.
Thanks pipo and padman. Could I have your email?
my email is [email protected].
-mike
-
**************
ARC0: Beginning to archive log# 2 seq# 1400
ARC0: Failed to archive log# 2 seq# 1400
Mon May 21 11:11:59 2001
ARCH: Completed archiving log# 2 seq# 1400
**************
There are many reasons why Archive process failed to archive log file:
1 REDO LOG may be small, so that frequent switch occurs. This leads to ARCH works slowly. Increase the size.
2 If the REDO files are mirrored (duplexed) and they are kept in RAID, then also ARCH will fail because when one of the REDO Log is filled, Oracle instructs ARCH to start archive. Then ARCH process tries to read the 2nd member of the REDO LOG file, which is not yet written by the DISK controller. Hence, ARCH process fails. To over come, do not keep the redo log files in RAID, keep them non-mirrored /striped disk.
3.Avoid SW level and HW level mirroring of redo log files
