VMDK level backups

[Bmitche]

Can I do vmdk level backups through the vmware option and selecting "enable file recovery from VM backup" straight to tape? Or does it have to goto a disk storage unit first?

[Marianne]

Yes - you can backup straight to tape.

VM backups uses FlashBackup, not GRT.

[Bmitche]

I guess I need to call support...unless you can decipher this mess

5/12/2012 1:38:24 AM - Info nbjm(pid=6452) starting backup job (jobid=537040) for client app01atl1d.iam.intranet, policy VMDK_Dev-Cluster_DeDupe---test, schedule WF 
5/12/2012 1:38:24 AM - estimated 37481397 Kbytes needed
5/12/2012 1:38:24 AM - Info nbjm(pid=6452) started backup (backupid=app01atl1d.iam.intranet_1336801104) job for client app01atl1d.iam.intranet, policy VMDK_Dev-Cluster_DeDupe---test, schedule WF on storage unit Bak32atl1p-MSL4096-1Drive
5/12/2012 1:38:25 AM - started process bpbrm (5696)
5/12/2012 1:38:27 AM - Info bpbrm(pid=5696) starting bptm          
5/12/2012 1:38:27 AM - Info bpbrm(pid=5696) Started media manager using bpcd successfully      
5/12/2012 1:38:29 AM - Info bpbrm(pid=5696) app01atl1d.iam.intranet is the host to backup data from    
5/12/2012 1:38:33 AM - Info bpbrm(pid=5696) telling media manager to start backup on client    
5/12/2012 1:38:33 AM - Info bptm(pid=7636) using 262144 data buffer size       
5/12/2012 1:38:33 AM - Info bptm(pid=7636) using 256 data buffers        
5/12/2012 1:38:34 AM - Info bptm(pid=7636) start backup          
5/12/2012 1:38:34 AM - Info bptm(pid=7636) Waiting for mount of media id 0110L5 (copy 1) on server bak32atl1p.iam.intranet.
5/12/2012 1:38:41 AM - mounting 0110L5
5/12/2012 1:38:42 AM - Info bpbrm(pid=6288) sending bpsched msg: CONNECTING TO CLIENT FOR app01atl1d.iam.intranet_1336801104    
5/12/2012 1:38:42 AM - connecting
5/12/2012 1:38:43 AM - Info bptm(pid=5064) setting receive network buffer to 262144 bytes     
5/12/2012 1:38:43 AM - Info bpbrm(pid=6288) start bpbkar32 on client        
5/12/2012 1:38:43 AM - connected; connect time: 00:00:01
5/12/2012 1:38:48 AM - Info bpbkar32(pid=7820) Backup started          
5/12/2012 1:38:48 AM - Info bpbrm(pid=6288) Sending the file list to the client     
5/12/2012 1:38:51 AM - Error bpbrm(pid=6288) from client app01atl1d.iam.intranet: ERR - C:\Program Files\Veritas\NetBackup\online_util\fi_cntl\bpfis.fim.app01atl1d.iam.intranet_1336800996.1.0.NBU_DATA.xml: XML parser error: load() failed in file XPathParser.cpp at line 84, hr: 1 error code: -2146697211 reason: System error: -2146697211. line: 0 line position: 0
5/12/2012 1:38:51 AM - Error bpbrm(pid=6288) could not send server status message      
5/12/2012 1:38:53 AM - Info bpbrm(pid=5696) child done, status 20        
5/12/2012 1:38:53 AM - Info bpbrm(pid=5696) sending message to media manager: STOP BACKUP app01atl1d.iam.intranet_1336801104    
5/12/2012 1:38:53 AM - Info bpbrm(pid=5696) media manager for backup id app01atl1d.iam.intranet_1336801104 exited with status 150: termination requested by administrator
5/12/2012 1:38:53 AM - end writing
invalid command parameter(20)

[Marianne]

Sorry, never seen anything like this before. Please share all relevant details about your environment - maybe someone with a similar setup can advise. Detals of backup host - physical or VM? OS on backup host? Data stores zoned to backup host? ESX or vSphere? Version? Any errors in Event Viewer on backup host?

[Bmitche · Accepted solution]

Details:

master server=win2008R2 physical...NB7.5.01

media server(s)=win2008R2 physical...NB7.5.01

vSshpere=5.0, datastores=v5

using NBD as transport

No errors in event viewer

I have no problems backing up to my MSDP pool first then duplicating to tape...but when I change my policy to go straight to tape...is when I get the errors. I have the problem on ALL vm's not just a few. I can backup straight to tape via the client agent, ms windows policy type.

[Omar_Villa]

Is your Backup Host doing the staging? sounds to me that when you change the from DSSU to STU is missing the snapshot stagging unit in the Backup or Discovery Host, just a hunch, also will not hut if you share you policy confing and some snapshot (bpfis) and bptm logs.

Also I see you have NBU 7.5, is your policy FlashBackup or VMware? this changed on the 7.5 and the policy needs to be migrated and the Backup Host must be upgraded to 7.5 too, check this out maybe will explain a pice of the funcky behavior.

Regards.