cancel
Showing results for 
Search instead for 
Did you mean: 

NB Catalog recovery keeps failing on NBDB recovery

zmlat
Level 4

Hello,

I have been trying to do a catalog recovery, I keep getting error code 2850 on the NBDB recovery.
Master server is a windows 2021 VM. Its a "restore only" backup master running NBU 7.7.3. I was having issues accessing the Java console, and I think windows team corrupted NB on the host by pushing out the 8.1.2 client. In any event, I figured just re-install, and restore the catalog. Hostname is the same. I see the following in the recover log:

18:00:27 (15.001) Status of restore from copy 3 of image created 08/30/19 09:36:42 = client/server handshaking failed

18:00:27 (15.001) The following files/folders were not restored:
18:00:27 (15.001) UTF - /D/Program Files/Veritas/NetBackupDB/staging/DARS_DATA.db
18:00:27 (15.001) UTF - /D/Program Files/Veritas/NetBackupDB/staging/DARS_INDEX.db
18:00:27 (15.001) UTF - /D/Program Files/Veritas/NetBackupDB/staging/DBM_DATA.db
18:00:27 (15.001) UTF - /D/Program Files/Veritas/NetBackupDB/staging/DBM_INDEX.db
18:00:27 (15.001) UTF - /D/Program Files/Veritas/NetBackupDB/staging/EMM_DATA.db
18:00:27 (15.001) UTF - /D/Program Files/Veritas/NetBackupDB/staging/EMM_INDEX.db
18:00:27 (15.001) UTF - /D/Program Files/Veritas/NetBackupDB/staging/JOBD_DATA.db
18:00:27 (15.001) UTF - /D/Program Files/Veritas/NetBackupDB/staging/NBAZDB.db
18:00:27 (15.001) UTF - /D/Program Files/Veritas/NetBackupDB/staging/NBAZDB.log.1
18:00:27 (15.001) UTF - /D/Program Files/Veritas/NetBackupDB/staging/NBDB.db
18:00:27 (15.001) Suspending the display of 8 additional messages for this image.

18:00:28 INF - Server status = 2850
18:00:28 (15.xxx) INF - Status = Restore error.

18:00:38 ERR - Failed to recover NBDB on MASTERSERVER.com (5)

Note that the media server doing the recovery did not exist at the time the catalog backup was taken, but I was able to get this far by editing the DR file, and doing the Media override.

Thanks

1 ACCEPTED SOLUTION

Accepted Solutions

zmlat
Level 4

Quick update...

I noticed errors in media server bptm:

10:28:18.458 [13561] <16> emmlib_initializeEx: (-) Exception! CORBA::NO_PERMISSION
10:28:18.459 [13561] <16> bptm: (-) Translating EMM_ERROR_Authorization_Denied(3000007) to 118 in the NetBackup context
10:28:18.459 [13561] <16> bptm: error (118) initializing EMM interface.
10:28:18.459 [13561] <16> main: (-) Translating ERROR UNDEFINED(118) to 196 in the Media context

When I googled that it seemed to indicate a dns issue, which there is: on the master, the reverse IP lookup for the media server fails. It did not occur to check that since I was able to add the media server, and recover db images. So I figured, let me duplicate the NBDB backup image to local disk on the media server, and recover from that copy. bpduplicate failed with error code 47. Seems some process keeps removing the media server. Added it back to the Masters additional server and media server config, and bpduplicate ran, and NBDB recovery from the disk copy worked. I got DB validation errors, but I can at least do restores now. 

View solution in original post

6 REPLIES 6

quebek
Level 6
   VIP    Certified

Hello

Windows server 2021 and NBU 7.7.3 - I think this pair is not supported...

Too less data - can you enable logging prior NBU catalog recovery? This from detailed status is not helpful. So please run mklogdir and put verbose to 3 or 5 for all logs. I would check tar logs and few others for clues.

the source and target NBU install dir is the same? here and there it is on D drive?

 

That was a typo...Windows 2012 master, NB version is 7.7.3. It only used for old restores. Looks like the images part of the catalog recovery worked, but I guess its failing on the last part, the NBDB recovery.

Recover log on master shows: 

17:30:23 (15.xxx) Restore job id 15 will require 1 image.
17:30:23 (15.xxx) Media id 123456 is needed for the restore.

17:30:24 (15.001) Restoring from copy 3 of image created 08/30/19 09:36:42 from policy CATALOG_BACKUP
18:00:27 (15.001) Status of restore from copy 3 of image created 08/30/19 09:36:42 = client/server handshaking failed

18:00:27 (15.001) The following files/folders were not restored:
18:00:27 (15.001) UTF - /D/Program Files/Veritas/NetBackupDB/staging/DARS_DATA.db
18:00:27 (15.001) UTF - /D/Program Files/Veritas/NetBackupDB/staging/DARS_INDEX.db
18:00:27 (15.001) UTF - /D/Program Files/Veritas/NetBackupDB/staging/DBM_DATA.db
18:00:27 (15.001) UTF - /D/Program Files/Veritas/NetBackupDB/staging/DBM_INDEX.db
18:00:27 (15.001) UTF - /D/Program Files/Veritas/NetBackupDB/staging/EMM_DATA.db
18:00:27 (15.001) UTF - /D/Program Files/Veritas/NetBackupDB/staging/EMM_INDEX.db
18:00:27 (15.001) UTF - /D/Program Files/Veritas/NetBackupDB/staging/JOBD_DATA.db
18:00:27 (15.001) UTF - /D/Program Files/Veritas/NetBackupDB/staging/NBAZDB.db
18:00:27 (15.001) UTF - /D/Program Files/Veritas/NetBackupDB/staging/NBAZDB.log.1
18:00:27 (15.001) UTF - /D/Program Files/Veritas/NetBackupDB/staging/NBDB.db
18:00:27 (15.001) Suspending the display of 8 additional messages for this image.

18:00:28 INF - Server status = 2850
18:00:28 (15.xxx) INF - Status = Restore error.

BPTM log on media server shows:

 

quebek
Level 6
   VIP    Certified

hi

the source and target server name are identical, right?

etc/hosts is pointing to correct server - or you do use dns - which I doubt as two same hostnames cant appear on AD...

Apologies for regurgitating the same info.

Same host name. I did a manual deinstall of the corrupted 7.7.3, and just re-installed on same host.

I had some challenges due to the media server not having been the original catalog backup media server, but I got around that.

zmlat
Level 4

Quick update...

I noticed errors in media server bptm:

10:28:18.458 [13561] <16> emmlib_initializeEx: (-) Exception! CORBA::NO_PERMISSION
10:28:18.459 [13561] <16> bptm: (-) Translating EMM_ERROR_Authorization_Denied(3000007) to 118 in the NetBackup context
10:28:18.459 [13561] <16> bptm: error (118) initializing EMM interface.
10:28:18.459 [13561] <16> main: (-) Translating ERROR UNDEFINED(118) to 196 in the Media context

When I googled that it seemed to indicate a dns issue, which there is: on the master, the reverse IP lookup for the media server fails. It did not occur to check that since I was able to add the media server, and recover db images. So I figured, let me duplicate the NBDB backup image to local disk on the media server, and recover from that copy. bpduplicate failed with error code 47. Seems some process keeps removing the media server. Added it back to the Masters additional server and media server config, and bpduplicate ran, and NBDB recovery from the disk copy worked. I got DB validation errors, but I can at least do restores now.