Tek-Tips is the largest IT community on the Internet today!

Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!

  • Congratulations Chris Miller on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

strange (and random) error

Status
Not open for further replies.

fabio9

IS-IT--Management
Sep 21, 2002
59
IT
Hi all...
I am trying to backup a 10g installation raw/rac/asm under 2003 srv and brightstor 11.5, of course using rman (because there is no other way right now...).
Random the backup fails with error "ec=backup agent error (279) cannot create ipc objects" . The dbora8.log says "03/14/2006 15:40 Cannot create ipc objects".

This error does not seems to exist in the ca support site..(and in google too...)
Sometimes, however, with the SAME rman script the backup runs with no problems.

Any idea ?!


Thanks in advance
 
Fabio,

well done for googling etc.

I don't know enough to say exactly what the error is, but I wonder if you could carry out an experiment.

Configure RMAN to back up to a local folder on the hard disk, and ignore any tape technologies (or anything else) that may be in use.

If the backup succeeds, then you know that the problem is not Oracle or RMAN, but something about your environment downstream of the backup. Basically, divide the whole thing down into smaller chunks and see where and when it fails.

Make sure that before you take a backup of your RMAN configuration, before you make any changes.

Regards

Tharg

Grinding away at things Oracular
 
Thanks for your answer...
Your is the rigth way to approach this kind of problem, but our dba are not capable of this...so I have not so many options. During the initial tests using the disk as a tape (brightstor filesystem tape) I encountered the same error. It seems to be a brightstor problem o, at least, of the layer between the db and the tape.
This layer is quite huge (form asm to the tape phisically) there are a lot of objects between.

Tnx.

Fabio

 
Fabio,

I suspected as much, because RMAN is usually very reliable.

If your DBA's can't help, then you should get them to study the RMAN documentation - it's their job!

However, I cannot over emphasise the importance of backups.

From a suitably authorised account, log in to RMAN and do the following:-

Code:
Microsoft Windows [Version 5.2.3790]
(C) Copyright 1985-2003 Microsoft Corp.

H:\>rman target /@dev_5

Recovery Manager: Release 9.2.0.1.0 - Production

Copyright (c) 1995, 2002, Oracle Corporation.  All rights reserved.

connected to target database: DEV_5 (DBID=2870968792)

RMAN> SHOW ALL;

using target database controlfile instead of recovery catalog
RMAN configuration parameters are:
CONFIGURE RETENTION POLICY TO RECOVERY WINDOW OF 3 DAYS;
CONFIGURE BACKUP OPTIMIZATION ON;
CONFIGURE DEFAULT DEVICE TYPE TO DISK;
CONFIGURE CONTROLFILE AUTOBACKUP ON;
CONFIGURE CONTROLFILE AUTOBACKUP FORMAT FOR DEVICE TYPE DISK TO 'e:\oracle\rman_
backups\dev_5\cab_%F';
CONFIGURE DEVICE TYPE DISK PARALLELISM 1;
CONFIGURE DATAFILE BACKUP COPIES FOR DEVICE TYPE DISK TO 1;
CONFIGURE ARCHIVELOG BACKUP COPIES FOR DEVICE TYPE DISK TO 1;
CONFIGURE CHANNEL 1 DEVICE TYPE DISK FORMAT   'e:\oracle\rman_backups\dev_5\back
up_%U';
CONFIGURE MAXSETSIZE TO UNLIMITED;
CONFIGURE SNAPSHOT CONTROLFILE NAME TO 'E:\ORACLE\RMAN_BACKUPS\DEV_5\SNCF_DEV_5.
ORA';

RMAN>
Whatever the setup is, take a copy of it, and save it safely somewhere, so that you can always restore to the current situation if something goes wrong.

Notice from my setup above that channel 1 is DEVICE TYPE DISK. The backup writes to the local disk, and is backed up to tape by the ordinary OS backup routines.

Alter your RMAN to backup to local hard disk, by configuring the device as above. Issue the command
Code:
RMAN> CONFIGURE DEFAULT DEVICE TYPE TO DISK;

old RMAN configuration parameters:
CONFIGURE DEFAULT DEVICE TYPE TO DISK;
new RMAN configuration parameters:
CONFIGURE DEFAULT DEVICE TYPE TO DISK;
new RMAN configuration parameters are successfully stored

RMAN>

Obviously, because my existing default device type was disk, this has made no difference to me, but it will to you.

Then issue the command 'BACKUP DATABASE PLUS ARCHIVELOG;' and watch for errors. If the backup appears to complete ok, issue the command
Code:
RMAN> RESTORE DATABASE VALIDATE;

Starting restore at 16-MAR-06

allocated channel: ORA_DISK_1
channel ORA_DISK_1: sid=11 devtype=DISK
channel ORA_DISK_1: starting validation of datafile backupset
channel ORA_DISK_1: restored backup piece 1
piece handle=E:\ORACLE\RMAN_BACKUPS\DEV_5\BACKUP_4SHDVVT6_1_1 tag=TAG20060315T01
1518 params=NULL
channel ORA_DISK_1: validation complete
Finished restore at 16-MAR-06

RMAN>

The line params=NULL shows that the backup is valid.
If you get this far, then Oracle and RMAN are behaving properly. The problem lies elsewhere. Your system administrators are then responsible for the matter, not you. It's their job to sort out backups from disk to tape, regardless of how complex the SAN (or whatever) is.

Regards

Tharg

Grinding away at things Oracular
 
Hey,
It's been a while since I worked with RMAN nad BrightStore, but that message looke familiar. I think the resolve was to increase a timeout value, either in the Oracle owner environment or in a parameter associated with the SBT channel. Looks like the media managment layer is failing when establishing inter-process communication to the BrightStore agent, increasing the timeout may help. You could also see if there's any correlation between the failures and other BrightStore activity, which might explain why the agent is busy or appears to be hung.

I would open a support issue with CA if you don't find how to set the timeout in the doc.

Tanloco (a.k.a. Tim from Evanston)
 
This strange error is still alive and I'm out of ideas.
Now I have bab 11.5sp1 on win2k3 ent.edition sp1, but the error is the same so I opened an issuse to c.a.



 
End of problem.

After a case opened (and after MORE THAN 30 DAYS) computer associates send me a new orasbt.dll .
It was a Brightstor bug.


Thanks all

 
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top