cancel
Showing results for 
Search instead for 
Did you mean: 

Error 50 on FULL Catalog backups

Hello colleagues,

 

I've a serious problem since last week.

FULL catalog copies end with status 50 on DARS_DATA.db copy job.

(INCREMENTAL catalog copies run fine)

 

In detail status job, i see 1 error (nbjm):

...

11/11/2011 06:00:42 - Info bpbrm (pid=29462) validating image for client netbackup-sc
11/11/2011 06:11:39 - Error nbjm (pid=2699) nbrb status: RB deallocated orphaned resources
11/11/2011 06:27:50 - Info bphdb (pid=29341) done. status: 0: the requested operation was successfully completed
11/11/2011 06:27:50 - end writing; write time: 0:27:20
client process aborted  (50)
 

Checking the log of the daemon nbjm, I only find this error:

...

11/11/11 06:11:39.518 V-117-295 [JobManager_i::cancel] cancel job, jobid=3287703, parentid=0, flags=4
11/11/11 06:11:39.504 [Error] V-117-67 Resource(s) deallocated, canceling job ID = 3,287,703

11/11/11 06:27:50.924 V-117-239 [BackupJob::terminateThisJob] terminated job, jobid=3287703, status=50
 

This is Netbackup 7.1.0.1 on a Solaris10 clustered master server.

 

Any help will be apreciated.

Thanks in advance.

1 Solution

Accepted Solutions
Accepted Solution!

"the backup job tries to copy on the staging location...

...Do you think that all databases are copied the same way??"

 

Altho' the DARS_DATA.db is new to me (still @6.5.x & was intro'd @7.0?), that's the understanding I have.

http://www.symantec.com/business/support/index?page=content&id=HOWTO34800

Snippet:

"Normally, a hot, online catalog backup consists of one parent job and two or more child jobs. Events for these jobs appear in the dbm log.

An overview of the hot catalog backup process consists of the following process:

Make a temporary copy of database files to a staging directory by typing the following command:

Install_path\NetBackupDB\staging

Once the copy is made, NetBackup can back up the catalog files."

View solution in original post

7 Replies

I have not seen anything like

I have not seen anything like this up to now.

Please 'health check' your relational database:  http://www.symantec.com/docs/TECH126758

PLEASE remember to freeze your NBU service group to prevent failover while you are following the steps in the TN.

You may have to open a Support call with Symantec depending on the results.

Hello Marianne, I've

Hello Marianne,

I've performed a consistency check of the Netbackup DDBBs.

It was OK.

I've opened a support case with Symantec. No news till now.

Seen something similar before

If you have a firewall between the master and media servers, it's the firewall closing a inactive TCP session.

Doc: http://www.mass.dk/netbackup/quick-hints/87-status-code-50-nbrb-status-nbrb-deallocated-orphaned-res...

 

master & media on same machine

Hello Nicolai,

Master and media involved in copying catalog are the same machine.

Also, INCREMENTAL backups are succesfull.

Weird...

Few things to check: 1) Is it

Few things to check:

1) Is it always failing when backing up the catalog file: /usr/openv/db/data/DARS_DATA.db

2) If so, how big is that file?

I would think the file is big and therefore time taken to complete that backup (and the rest) is too long. The process might be terminated by a 3rd-party monitoring system, or crash therefore error 50 was reported. That also explain why an incremental works instead of full, because less data is needed for backup.

Not sure whether there is a way to reduce the catalog file size, all I know was to archive some of those.

I would suggest increasing the time-out value and check if there is a core dump for nb processes.

Hello, 1) Yes. The job's file

Hello,

1) Yes. The job's file list only contains that file : /opt/VRTSnbu/db/staging/DARS_DATA.db

I'm not sure if there're other DBs copied by this job.

2) This file is quite small. Only 25MB.

# ls -lah /opt/VRTSnbu/db/data/DARS_DATA.db
-rw-------   1 root     root         25M Jun 28 16:13 /opt/VRTSnbu/db/data/DARS_DATA.db
 

Thanks to your post I've realized that the backup job tries to copy on the staging location.

Do you think that all databases are copied the same way??

Thanks again.

Accepted Solution!

"the backup job tries to copy on the staging location...

...Do you think that all databases are copied the same way??"

 

Altho' the DARS_DATA.db is new to me (still @6.5.x & was intro'd @7.0?), that's the understanding I have.

http://www.symantec.com/business/support/index?page=content&id=HOWTO34800

Snippet:

"Normally, a hot, online catalog backup consists of one parent job and two or more child jobs. Events for these jobs appear in the dbm log.

An overview of the hot catalog backup process consists of the following process:

Make a temporary copy of database files to a staging directory by typing the following command:

Install_path\NetBackupDB\staging

Once the copy is made, NetBackup can back up the catalog files."

View solution in original post