Video Screencast Help

NBEMM DB Connection Issues

Created: 17 Sep 2012 | 10 comments

A few days ago the logging on I am having problems with the NBU Admin Console. When I try to access the robot inventory it recieve an error. I can query the EMM DB using nbemmcmd.exe and it returns the correct results. However when accessing through the console i get this message:

 

"Error connecting to the Media Manager on [MASTER]: network protocol error(39)."

 

I have cycled the services(including PBX) and rebooted the Master server. This is a Windows 2003 server with NBU 7.1.0.3 installed.

I have done some research on error 39, and I have deleted the entire contents of the <install_path>\VERITAS\NetBackup\Remote_Versions at the suggestion of http://www.symantec.com/business/support/index?pag....

 

I have a ticket open with support however was trying to get it fixed sooner. Has anyone else seen this issue before?

 

-Thanks

 

 

Comments 10 CommentsJump to latest comment

Marianne's picture

Has anyone perhaps changed or deleted MM_SERVER_NAME in vm.conf?

Is only Inventory failing? Backups (including tape mount/dismount) working fine?

What is robot type?
TLD?
ACS?

Please post device config:

tpconfig -l
(in ...volmgr\bin)

Supporting Storage Foundation and VCS on Unix and Windows as well as NetBackup on Unix and Windows
Handy NBU Links

MilesVScott's picture

vm.conf has not been touched.

 

I can not even open the "Media" tab so I am unable to use the GUI for anything that is realated to media.

 

It is TLD.

Device Robot Drive       Robot                    Drive                 Device                            
Type     Num Index  Type DrNum Status  Comment    Name                  Path                              
robot      0    -    TLD    -       -  -          -                     {2,0,0,0}
  drive    -    0 hcart2    5    DOWN  -          HP.ULTRIUM2-SCSI.002  MISSING_PATH:{2,0,16,0}:HUM5M001  
  drive    -    1 hcart2    6      UP  -          HP.ULTRIUM2-SCSI.003  {2,0,15,0}                        
  drive    -    2 hcart2    8    DOWN  -          HP.ULTRIUM2-SCSI.005  MISSING_PATH:{2,0,13,0}:MXP9E03K  
  drive    -    3 hcart2    7      UP  -          HP.ULTRIUM2-SCSI.004  {2,0,14,0}                        
  drive    -    4 hcart2    4      UP  -          HP.ULTRIUM2-SCSI.001  {2,0,9,0}                         
  drive    -    5 hcart3   10      UP  -          HP.ULTRIUM3-SCSI.000  {2,0,4,0}                         
  drive    -    6 hcart3   11      UP  -          HP.ULTRIUM3-SCSI.001  {2,0,3,0}                         
  drive    -    7 hcart3   12      UP  -          HP.ULTRIUM3-SCSI.002  {2,0,7,0}                         
  drive    -    8 hcart3   13      UP  -          HP.ULTRIUM3-SCSI.003  {2,0,6,0}                         
  drive    -    9 hcart3    9      UP  -          HP.ULTRIUM3-SCSI.004  {2,0,8,0}                         
  drive    -   10 hcart2    1    DOWN  -          HP.ULTRIUM2-SCSI.006  MISSING_PATH:{2,0,12,0}:HUM4M012  
  drive    -   11 hcart2    2      UP  -          HP.ULTRIUM2-SCSI.007  {2,0,10,0}                        
  drive    -   12 hcart2    3    DOWN  -          HP.ULTRIUM2-SCSI.000  MISSING_PATH:{2,0,9,0}:HUM6K0012  
robot      1    -    TLD    -       -  -          -                     qflexmail3
  drive    -   13  hcart    1      UP  -          HP.ULTRIUM5-SCSI.000  {2,0,1,0}                         
  drive    -   14  hcart    2      UP  -          HP.ULTRIUM5-SCSI.001  {2,0,5,0}                         

 

Thanks,

If my answer was helpful, please mark as a solution so others can benefit as well.

NBU 7.1.0.3,Win 2k3 Enterprise x64 SP2

Miles Scott | Holland & Knight
IT Infrastructure Specialist

Karthikeyan Sundaram's picture

Who is the media server? is media and master on same server? if media server is different, did you get to bounce the NBU daemons and see the status?

post us vm.conf file content

Do you see any wrong on the <Install_path>\volmgr\debug or ltid log files

Thanks, Karthikeyan Sundaram.

Andy Welburn's picture

Something else to try?

It's a very simple change but it was in relation to "39" issues I was having following an upgrade to 6.5.6 a couple of years ago, so it may well have absolutely no relevance to your issue, but .....

https://www-secure.symantec.com/connect/forums/loo...

MilesVScott's picture

Andy:

 

I had actually looked into that. Funny thing is, when i first experienced the issue, it took forever for the console to return the error code. When I saw this I was like hum maybe thats my issue. Well I went to look for it, and I guess the previous admin had set the timeout to 1000. So i bumped that back down to a more reasonable time.

Karth:

 

This is the contents of the vm.conf file:

 

MM_SERVER_NAME = [MASTER SERVER]
VERBOSE = 5
 

As far as the ltid logs, the only thing I saw in there was this:

 

16:53:35.909 [5040.3676] <4> tpreq: ENOFILE3: cannot find pending mount_req
16:53:35.909 [5040.3676] <4> tpreq: unlink failed (F:\Program Files\Veritas\NetBackup\db\media\tpreq\drive_HP.ULTRIUM2-SCSI.007), errno = 2 (No such file or directory)

 

Not sure but that doesn't look like it would be my actual problem. The oddest thing is, as I walked in today, its working fine. Not sure what happend. I know that sometimes the postgresql db for PureDisk can take a while, but has anyone ever heard of the EMMdb taking 8+ hours to work?

 

Thanks,

If my answer was helpful, please mark as a solution so others can benefit as well.

NBU 7.1.0.3,Win 2k3 Enterprise x64 SP2

Miles Scott | Holland & Knight
IT Infrastructure Specialist

Mouse's picture

Check the SERVER LIST and other relevant SERVER entries in Registry under config. Maybe some parasite entry left there.

I've seen weird things like this, when you kill, for example, bpresolver it will leave DAG name in the server list and after restart your master starts feeling itself as a media server and so on, looks scary at first glance.

MilesVScott's picture

I took a look at the reg keys this is all I saw under config. Just a list of the media servers. Now I did notice that some old decommisioned media servers were still in this list. Would that cause a problem such as this?

 

If my answer was helpful, please mark as a solution so others can benefit as well.

NBU 7.1.0.3,Win 2k3 Enterprise x64 SP2

Miles Scott | Holland & Knight
IT Infrastructure Specialist

Mouse's picture

If there are any entries that are not accessible from the master at the moment it's better to remove them. The first entry in SERVER list should be the master hostname IIRC

In one of these weird cases the stale entry forced the master to think it's a media server

MilesVScott's picture

The Master server is lsited first. But I dont see anything out of the norm.

If my answer was helpful, please mark as a solution so others can benefit as well.

NBU 7.1.0.3,Win 2k3 Enterprise x64 SP2

Miles Scott | Holland & Knight
IT Infrastructure Specialist

Marianne's picture

Check 'nbemmcmd -listhosts -verbose' output.

Any decommissioned media server that was not properly decommissioned WILL cause problems.

 

Supporting Storage Foundation and VCS on Unix and Windows as well as NetBackup on Unix and Windows
Handy NBU Links