Restore issue in netbackup

[Anjan_Das]

Hi friends need your help.

i have a client 'A' taking backup through media server 'B'. another client it self media server 'c'.both 'A' and 'c' backup taking from same master server 'M'.when i am going to restore backup of client 'A' to restore on server 'c'.it showing error massage socket write failed(24).

i add line on bp.conf file as

FORCE_RESTORE_MEDIA_SERVER = B c

after that no success. i am able to telnet from master to client. client to master. media to master, master to media and media to client.
in job details status it showing like

bptm error TPerrno= check drive on server are available
bprd error requested image expire on DATE

please help. how can i restore that backup

[Mouse]

Hmm, maybe stupid question - can your media servers get an access the same tape?

[Marianne]

Did you restart NBU on master after adding FORCE_RESTORE entry?
Verify in bprd log on master that restore is indeed sent to media server listed in FORCE_RESTORE entry.
If that is all fine, look at bptm log on media server for error.

[Anonymous]

See this technote, as we had a recent issue with alternate redirected restore and when this was in place it all worked.

DOCUMENTATION: Clarification on usage of the No.Restrictions touch file
http://seer.entsupport.symantec.com/docs/280742.htm

Tip:
To reread bp.conf without restarting Netbackup on the master:

# bprdreq -rereadconfig -verbose

[Anjan_Das]

No i am not restarted NBU on master server. i run command bprdreq -rereadconfig after adding entery FORCE_RESTORE on master server. 

[Anjan_Das]

Media server access tape libary and all backup running fine with media server.but still having same problem.

[Android]

Socket errors are more often than not network related.  I would start by looking at the bpcd logs on the to determine which daemon/service is having the socket/port issue. 

When you say you can telnet between all of the machines, did you do that by hostname?  Make sure they can resolve each other's names, telnet by IP wont prove/disprove anything. 

When you see which daemon/service is giving the socket error, and at which step of the process(server a is trying to talk to server b), find out which port that daemon/service uses and be sure you can connect to it on server b with telnet.

#telnet serverb port#

[Anjan_Das]

08/21/2009 03:34:45 - begin Restore
08/21/2009 03:34:51 - number of images required: 1
08/21/2009 03:34:51- media needed: KC1284
08/21/2009 03:34:56 - restoration from image A_1249221889
08/21/2009 03:34:57- connecting
08/21/2009 03:34:59- started process bptm (pid=1089)
08/21/2009 03:34:59- mounting KC1284
08/21/2009 03:35:00- connected; connect time:0:00:00
08/21/2009 03:35:02- error bptm (pid=1089) error requesting media, TpErrno = Requested number of drives are not configurd.
08/21/2009 03:35:03- warning bprd(pid=11682) restore must be resumed prior to first image expiration on tue 03 Nov 2009 08:04:49 AM
08/21/2009 03:35:02- end restore; elaspsed time 0:00:18

Socket write failed(24)

now i am able to telnet bpcd port  from master server  to media and client, network communication is fine. after that i got that error. please help how to resolve issue.

Thanks
 

[J_H_Is_gone]

where are you starting the BAR from?

Server a b or c?

I would start it from the master if you are not already.

[Amit_Karia]

Are you able to do bpcoverage from media server C to client A ? also check all the drives are available in media server C..