Netbackup unable to restore MS Exchange objects

[Alain_Razafiman]

I have a problem when restoring MS Exchange mailbox.

Here is my conf :

Netbackup server 6.0MP4 on AIX 5.3,

Exchange agent 6.0MP4 on Windows 2003 R2 Enterprise

The job status window shows these errors :

........
begin reading
Error bpbrm (pid=405554) from client svrmtst: ERR - unable to create object for restore : Microsoft information Store:\.........
end reading; read time: 0:32:11
Error bpbrm (pid=405554) client restore EXIT STATUS 185: tar did not find all the files to be restored
...
the restore failed to recover the requested files
(5)

What is the problem ?

[Marianne]

Where are you initiating the restore from?

Does the mailbox exist on the Exchange server that you are restoring to?

[Alain_Razafiman]

The restore operation is initiated from the supervisor console . The backup on tape that I need to restore is from the production server and I want to restore it on the test server.

The mailbox does not exist on the test server.

Only the mailbox group has been created on the test server.

Maybe I should create the mailbox before restoring ?

[RiaanBadenhorst · Accepted solution]

Correct, you need to create it

Cut from Exchange Admin Guide (6.5)

In the Destination box, indicate where you want to restore the object.
■    You must indicate an explicit path (or full path) in the Destinationbox.
■    When you restore mailboxes, change the mailbox name to another existing mailbox.
For example, if you want to restore the contents of Mailbox 1 to Mailbox 2, specify one of the following in the Destination box: Microsoft Exchange Mailboxes:\Mailbox 2\ Microsoft Information Store:\Storage Group\Mailbox Store [Mailbox]\Mailbox 2\

[Alain_Razafiman]

OK,

I'll try today and inform you

Thanks

[Alain_Razafiman]

I created the mailbox and now the restore job seems to hang. I started the job on oct. 7th 11.42 am and still running on oct. 8th 3.00 pm. I had to kill the job.

Hereafter are the messages :

.......

7 oct. 2010 11:47:15 - begin reading

7 oct. 2010 11:47:47 - Warning bpbrm (pid=5777604) from client srvmtst: WRN - error writing object: Microsoft Information Store:\Groupe de stockage de récupération\Boite DG - DARF - RPCI

7 oct. 2010 11:47:47 - Warning bpbrm (pid=5777604) from client srvmtst: WRN - error writing byte: 0

7 oct. 2010 11:47:47 - Warning bpbrm (pid=5777604) from client srvmtst: WRN - desired byte count: 64000

7 oct. 2010 11:47:47 - Warning bpbrm (pid=5777604) from client srvmtst: WRN - actual byte count: 24

8 oct. 2010 15:03:12 - Error bptm (pid=581834) media manager terminated by parent process

..............

termination requested by administrator (150)

See attached in zip file some log files

[Marianne]

Have a look at the errors in the tar log:

11:54:23.857: [5004.4140] <2> ov_log::V_GlobalLogEx: ERR - beds_ese_access::V_Write:FS_WriteObj() Failed! 0xE000030B:The database must be created prior to restore.

11:54:23.857: [5004.4140] <2> tar_base::V_vTarMsgW: WRN - error writing object: Microsoft Information Store:\Groupe de stockage de récupération\Boite DG – DARF – RPCI
 

[RiaanBadenhorst]

Did you create this information store "Groupe de stockage de récupération"
 or a different one. It would seem this is where you're trying to restore to. So either create it, or update the location you're restoring to.

[Alain_Razafiman]

After checking and modifying some restore paths, the restore operation seems to start successfully but does not end. there is a network timeout error. (see attached files)

Other file level restore works well even for large data and within an acceptable amount of time.

the logs and screen capture attached concerns a 24GB restore of mailbox. 94 MB were restored before the network timeout error. A second attemp were issued and 400MB out of 24 GB were restore before a network timeout error.

I wonder if there is a problem in network, data spaca available, agent perf, ... ?

[Marianne]

Interesting - only 94 mb restored and timeout after more than 6 hours?

All that tar is telling us is that the restore started at 14:27 and that the network connection was terminated at 20:31 (because of the timeout).

You will need to check server logs as well.

Master: bprd

Media server: bptm and bpbrm

At this point it is not clear where exactly the problem is - destination client or server side.

If the logs on master/media server does not pinpoint the problem, you will probably have to increase logging level on client to see if we can get more info in the tar log.

Any Exchange logs in Event Viewer Application log that's related to this restore?

[Alain_Razafiman]

There are no logs for bprd, bptm and bpbrm (folders are empty)

No informations corresponding to the error is found in event viewer application

Another attempt were made and 400MB out of 24GB were restored before "network timeout error" . See attached the corresponding logs if they can be useful for diagnostics (the restore started on october 13th and aborted on october 14th). At this time I did not increase the logging level yet.

I will make another attempt today or tomorrow with logging level on client set to a higher value.

[Marianne]

NBU on master needs to be restarted for bprd log to take effect.

bptm and bpbrm needs to exist on media server at the start of restore process.