AlertLog message

Hi

I have paste below a copy of the output of one of Alert.log
On that server I have 3 databases.
Each database has an ALERT.LOG file displaying the same message.
As you can see,
1/ Try to archive
2/ Fail to archive
3/ Archive successful

Anyone has a clue ???

Thanks in advance

Latyr


Wed Feb 25 21:02:32 2004
Thread 1 advanced to log sequence 82
Wed Feb 25 21:02:32 2004
ARCH: Beginning to archive log# 3 seq# 81
Wed Feb 25 21:02:32 2004
Current log# 2 seq# 82 mem# 0: R:\ORACLE\ORADATA\ABEL\REDO\REDO02A.RDO
Wed Feb 25 21:02:32 2004
ARC0: Beginning to archive log# 3 seq# 81
ARC0: Failed to archive log# 3 seq# 81
Wed Feb 25 21:02:33 2004
ARCH: Completed archiving log# 3 seq# 81
Thu Feb 26 21:02:33 2004
Thread 1 advanced to log sequence 83
Thu Feb 26 21:02:33 2004
ARCH: Beginning to archive log# 2 seq# 82
Thu Feb 26 21:02:33 2004
Current log# 1 seq# 83 mem# 0: R:\ORACLE\ORADATA\ABEL\REDO\REDO01A.RDO
Thu Feb 26 21:02:33 2004
ARC0: Beginning to archive log# 2 seq# 82
ARC0: Failed to archive log# 2 seq# 82
Thu Feb 26 21:02:33 2004
ARCH: Completed archiving log# 2 seq# 82
Fri Feb 27 21:02:32 2004
Thread 1 advanced to log sequence 84
Fri Feb 27 21:02:32 2004
ARCH: Beginning to archive log# 1 seq# 83
Fri Feb 27 21:02:32 2004
Current log# 3 seq# 84 mem# 0: R:\ORACLE\ORADATA\ABEL\REDO\REDO03A.RDO
Fri Feb 27 21:02:32 2004
ARC0: Beginning to archive log# 1 seq# 83
ARC0: Failed to archive log# 1 seq# 83
Fri Feb 27 21:02:32 2004
ARCH: Completed archiving log# 1 seq# 83
Sat Feb 28 21:02:37 2004
Thread 1 advanced to log sequence 85
Sat Feb 28 21:02:37 2004
ARCH: Beginning to archive log# 3 seq# 84
Sat Feb 28 21:02:37 2004
Current log# 2 seq# 85 mem# 0: R:\ORACLE\ORADATA\ABEL\REDO\REDO02A.RDO
Sat Feb 28 21:02:38 2004
ARC0: Beginning to archive log# 3 seq# 84
ARC0: Failed to archive log# 3 seq# 84
Sat Feb 28 21:02:38 2004
ARCH: Completed archiving log# 3 seq# 84
Sun Feb 29 21:02:03 2004
Thread 1 advanced to log sequence 86
Sun Feb 29 21:02:03 2004
ARCH: Beginning to archive log# 2 seq# 85
Sun Feb 29 21:02:03 2004
Current log# 1 seq# 86 mem# 0: R:\ORACLE\ORADATA\ABEL\REDO\REDO01A.RDO
Sun Feb 29 21:02:03 2004
ARC0: Beginning to archive log# 2 seq# 85
ARC0: Failed to archive log# 2 seq# 85
Sun Feb 29 21:02:03 2004
ARCH: Completed archiving log# 2 seq# 85
Tue Mar 02 21:02:31 2004
Thread 1 advanced to log sequence 87
Tue Mar 02 21:02:31 2004
ARCH: Beginning to archive log# 1 seq# 86
Tue Mar 02 21:02:31 2004
Current log# 3 seq# 87 mem# 0: R:\ORACLE\ORADATA\ABEL\REDO\REDO03A.RDO
Tue Mar 02 21:02:31 2004
ARC0: Beginning to archive log# 1 seq# 86
ARC0: Failed to archive log# 1 seq# 86
Tue Mar 02 21:02:31 2004
ARCH: Completed archiving log# 1 seq# 86

What version are you running?

There was a problem on 8.1.7.2 - but I thought that was a speed problem. The device couldn't keep up.

Are the files actually there? Are they usable? (use logminer to see if they are readable)

I see the same behaviour in 8.1.7.4 The archlogs apply to the standby without any problems, so I have worked on the assumption that this behaviour is acceptable.

Quote:

---

Originally posted by DaPi
I see the same behaviour in 8.1.7.4 The archlogs apply to the standby without any problems, so I have worked on the assumption that this behaviour is acceptable.

---

Excatly what I thought too even if I have been a bit worried during a while as I noticed it first after a PITR with RMAN on one database.

Thanks

Latyr

This is from Metalink: I had similar situations but needed the
additional arch process:

CAUSE
------

'Failed to archive log' messages appear in the alert log as multiple arch procresses are started
They are trying to start archiving of filled redo logs but as another arc process
already had started archiving them, other process faile and simply write a message in the alert log.

If you look atthe alert log , you will observe that, the arc process that starts archiving the log
is the process which eventually completes it


FIX
-------

This messages do not cause any problem to the working of database
Except that they fill up the alert log quite fast!!

to stop these messages from appearing in the alert log

decrease the archiver processes equal to 1


..... Please note, its still a draft.

Regards,
Rowena Serna
Oracle Corporation




HTH
Gregg

I'm not sure that's it . . . . :confused:

The archiver process are supposed to be ARC0 to 9, so I assummed ARCH (which reports success) was a "generic" message ? ? ?

LOG_ARCHIVE_MAX_PROCESSES is 1 by default (my case) - not sure that 0 is a good value!

Have I missed something ? ? ?

Quote:

---

Originally posted by DaPi
I'm not sure that's it . . . . :confused:

The archiver process are supposed to be ARC0 to 9, so I assummed ARCH (which reports success) was a "generic" message ? ? ?

LOG_ARCHIVE_MAX_PROCESSES is 1 by default (my case) - not sure that 0 is a good value!

Have I missed something ? ? ?

---

I ran this query and here is what I have

select name, value from v$parameter where name='log_archive_max_processes';
NAME VALUE
----------------------------------------------- -------
log_archive_max_processes 1

Found on Metalink
Note 119547.1 : Message 'Failed to archive' in Alert.log
fact: Oracle Server - Enterprise Edition 8.1
symptom: RMAN errors when backing up archive logs
symptom: ARCHIVE LOG CURRENT command is issued
symptom: Errors appear in alert file
symptom: ARCH: Beginning to archive log# %s seq# %s
symptom: ARC0: Beginning to archive log# %s seq# %s
symptom: ARC0: Failed to Archive log# %s seq# %s
symptom: ARCH: Completed archiving log# %s seq# %s
symptom: Database in archive log mode
symptom: The redolog is archived correctly
cause: The command 'Alter system archive log current' he is forces manual
archiving of
the current log, hence the shadow process on which the command was issued will
perform
the archiving. It will archive the logfile and write messages to the
trace/alert using
the prefix ARCH.

If automatic archiving is turned on then the archiver process will also try
to archive this logfile, however will not be able to aquire the lock 'kcrrlt'
ie the lock to protect multiple arch processes from archiving the same logfile.

The failing process process will write to the trace/alert message that it was
unable to archive the logfile.

In Oracle 9i the message in the alert.log is more informative:
ARC1: Beginning to archive log# 2 thrd# 1 seq# 24
ARC1: Unable to archive log# 2 thrd# 1 seq# 24
Log actively being archived by another process

By setting the parameter log_archive_trace = 127 in the init.ora file
the archiving steps for both the shadow as well as the archiver process can be
traced.


Reference :
'ARCHIVE LOG CURRENT' causes 'FAILED TO ARCHIVE'



fix:

The command 'Alter system archive log current' is commonly used in hot-backup
scenarios. The message 'Failed to archive' can be ignored, as long as it is
followed by a 'Completed archiving' message.

Workaround:
Do not issue the command 'Alter system archive log current'.

Reference PARAMETER:
LOG_ARCHIVE_TRACE

That explains it for my case - I issue an Alter System Archive Log Current every 10 minutes to feed my standby.