This post is more than 5 years old
2 Intern
•
139 Posts
0
5576
Authentication error (0:5:111) error during Online Oracle DB backup using NMO 5.0
Hi,
larger database backup have been failing with below error.
channel CH5: starting piece 1 at 06-JUN-12
RMAN-03009: failure of backup command on CH4 channel at 06/07/2012 05:24:16
ORA-27192: skgfcls: sbtclose2 returned error - failed to close file
ORA-19511: Error received from media manager layer, error text:
Authentication error (0:5:111)
continuing other job steps, job failed will not be re-run
channel CH4: starting full datafile backup set
channel CH4: specifying datafile(s) in backup set
Kindly suggest...........
ble1
2 Intern
2 Intern
•
14.3K Posts
0
June 14th, 2012 05:00
Hopefully you did the same on backup server too (and you must restart daemons after that for change to take effect).
I have seen this on two different occasions:
a) nsrauth
b) client has been ugraded and there was an issue with NMDA/NMO licenses (server was lower version and client was higher using NMDA)
However, in both cases if I remeber correctly (for b) for sure) this error would happen imediatelly. In you case, it looks more as if control session is killed by timeout of some kind (firewall, TCP_KEEPALIVE, similar).
coganb
736 Posts
0
June 8th, 2012 05:00
Hi,
If the problem only happens on larger backups, then you should implement a tcp keepalive. For details of this, you can look at the Configuring TCP Networks and Network Firewalls for EMC NetWorker technical note page 29 which explains how to implement this for each OS.
If this doesn't fix it, increase the debug level on the backup and check the debug log for more detail around the error message. You do this by adding NSR_DEBUG_LEVEL and NSR_DEBUG_FILE to the parms.
-Bobby
SabarnaDeb
2 Intern
2 Intern
•
139 Posts
0
June 8th, 2012 08:00
I disabled nsrauth in client and storage node. I will start backup...lets see
Bebo2k
544 Posts
0
June 9th, 2012 16:00
Hi Sabarna_Deb,
I think this error produced because of a miscommunication between NMO and NetWorker client, This error is not related to RMAN, it is either the nsrauth , or the username and password you are using for backups has been changed ( Permissions ). I believe by disabling the nsrauth ( As you done already ) this behavior will be changed and fixed, otherwise double check the username and password you are using for backing up.
Hope this helps you. Waiting your updates.
Thanks,
Ahmed Bahaa
SabarnaDeb
2 Intern
2 Intern
•
139 Posts
0
June 11th, 2012 07:00
No Backup failed again. I am using current username:sys and password. Kindly suggest...........:)
masonb
445 Posts
0
June 11th, 2012 07:00
Sabarna Deb,
Do you have any ORA-xxxx messages? Has this backup ever worked or is it new installation?
Does the backup work when running the RMAN script manually?
Regards,
Bill Mason
SabarnaDeb
2 Intern
2 Intern
•
139 Posts
0
June 11th, 2012 08:00
Previously backup was working fine from networker . No Authentication error. From oracle end, backup is completing. larger database backup is failing, not for small database like "user" tablespace with control file. At the end, backup is failing...............
fx8282
38 Posts
0
June 14th, 2012 06:00
Change the tcp keepalive to 10,000 in the networker startup script on the backup server and restart networker.
This should solve it.
SabarnaDeb
2 Intern
2 Intern
•
139 Posts
1
June 19th, 2012 06:00
TCP_keepalive was already set. Now issue has been resolved.
Steps: 1. Delete NSR_Peer information from another oracle physical server paired of effected server (Oracle database in RAC enviroment) because i believe somewhere else second server was interrupting the backup from first oracle client.
2. Again configure oracle client using legacy method.
makapil
2 Posts
0
January 8th, 2014 11:00
Could you please explain,, what do u mean by legacy method.
Did you re-configure it in NMC or Oracle DBA re-confiure it via RMAN script ?
ble1
2 Intern
2 Intern
•
14.3K Posts
0
January 8th, 2014 12:00
It means disable nsrauth, Check KB to details.
SabarnaDeb
2 Intern
2 Intern
•
139 Posts
0
January 9th, 2014 03:00
Longtime back i configured legacy method . Below i am pasting the procedure for taking online oracle database backup using Networker NMO. This may help you. In Legacy, backup RMAN script need to be kept in Oracle server (Vi file). Mention backup Path to RMAN script in Networker oracle client and few parameters we need to enter in Client's property. I forget. I need to recheck.
Procedure
I divided these procedures into three sections:
A. Installation procedure for taking Oracle Online tape backup and restoration.
B. Procedure for taking online oracle tape backup.
C. Restoration.
Section A. Installation procedure for taking Oracle Online tape backup and restoration.
1. Install legato client software 7.6.2 and NMO 5.0 on destination oracle client one by one as per below steps.
% cd $ORACLE_HOME/lib
% ln -s /usr/lib/libnwora.so libobk.so
Section B. Procedure for taking online oracle tape backups using EMC networker 7.6.3 and NMO 5.0.
CONFIGURE CONTROLFILE AUTOBACKUP ON;
CONFIGURE CONTROLFILE AUTOBACKUP FORMAT FOR DEVICE TYPE ‘SBT_TAPE’ TO 'NMO_%F';
Note: If we didn’t specify ‘AUTOBACKUP ON’ even though Networker will take control file backup but that would be unrecognizable by user.
RMAN_SCRIPT:
"RUN {
ALLOCATE CHANNEL CH1 TYPE 'SBT_TAPE';
ALLOCATE CHANNEL CH2 TYPE 'SBT_TAPE';
ALLOCATE CHANNEL CH3 TYPE 'SBT_TAPE';
ALLOCATE CHANNEL CH4 TYPE 'SBT_TAPE';
ALLOCATE CHANNEL CH5 TYPE 'SBT_TAPE';
SEND 'NSR_ENV=(NSR_SERVER=backup_server_name,NSR_CLIENT= )';
BACKUP
FULL
FORMAT '%d_%U'
DATABASE
INCLUDE CURRENT CONTROLFILE
PLUS ARCHIVELOG;
RELEASE CHANNEL CH1;
RELEASE CHANNEL CH2;
RELEASE CHANNEL CH3;
RELEASE CHANNEL CH4;
RELEASE CHANNEL CH5;
}";
Adjust the number of channels based on oracle server’s CPU utilization.
NOTE (Very Important): Whenever oracle tape backup failed whether manual or scheduled backup. Do contact DBA Team immediately to check whether any background RMAN session is running or not on effected database. If yes, we need to kill bad RMAN session on database and also always make sure that keep “client retries” value 0(zero) in group’s advanced properties. Below is one of the convenient ways we can check from database level.
Command:
SQL> l
1 SELECT s.sid,s.serial#,p.spid,s.client_info FROM v$process p, v$session s
2 WHERE p.addr = s.paddr
3* AND client_info LIKE '%rman%'
SQL> /
Output (For example)
SID SERIAL# SPID CLIENT_INFO
---------- ---------- ------------------------ ------------------
1038 15007 25507 rman channel=CH6
We need to kill these bad RMAN session with the below command at database level.
SQL> alter system kill session '1038,15007';
System altered.
Section C:Restoring oracle database to same host.
RUN {
SET DBID
ALLOCATE CHANNEL CH1 TYPE 'SBT_TAPE';
SEND 'NSR_ENV=(NSR_SERVER=backup_server_name,NSR_CLIENT= )';
RESTORECONTROLFILEFROM“Controlfile_name”;
RELEASE CHANNEL CH1;
}
It is also recommended for to insure successful database restoration do crosscheck of restored (above) control file with networker database not to point by control file to unnecessary datafiles. Database should be in mount mode.
RUN {
ALLOCATE CHANNEL CH1 TYPE 'SBT_TAPE';
SEND 'NSR_ENV=(NSR_SERVER=backup_server_name,NSR_CLIENT= )';
Crosscheck backup;
RELEASE CHANNEL CH1; }
Once cross check backup completed , execute below command in RMAN script.
RMAN>delete noprompt expired backup;
It will delete the expired images from restored control file.
Once control file restored successfully put database in ‘mount’ mode and execute below RMAN script.
RUN {
ALLOCATE CHANNEL CH1 TYPE 'SBT_TAPE';
ALLOCATE CHANNEL CH2 TYPE 'SBT_TAPE';
ALLOCATE CHANNEL CH3 TYPE 'SBT_TAPE';
ALLOCATE CHANNEL CH4 TYPE 'SBT_TAPE';
ALLOCATE CHANNEL CH5 TYPE 'SBT_TAPE';
SEND 'NSR_ENV=(NSR_SERVER=backup_server_name,NSR_CLIENT= )';
RESTORE DATABASE;
RECOVER DATABASE;
RELEASE CHANNEL CH1;
RELEASE CHANNEL CH2;
RELEASE CHANNEL CH3;
RELEASE CHANNEL CH4;
RELEASE CHANNEL CH5;
}
Moniter and check the log file for successful database restoration.