Unable to mount tape in TLD

[SteveDown]

Hi all,

I've built a 2003 server with Netbackup 6 installed. When trying to do anything with the tape, such as erase or start a backup, I get an error and the drive goes down. When starting a job I can hear the tape loading up etc, then it just sits there for about 5mins saying mounting. I know there's nothing wrong with the TLD because I've used it on another server where I've been performing similar tests with the same software versions. I've compared both server configs however I can't spot anything out the ordinary.

The TLD is an HP MSL 2024 with HCART3 defined in the config. I'm using an LTO2 tape with a barcode, which has been read and correctly inventoried.

11/06/2015 10:22:01 - begin Erase
11/06/2015 10:22:01 - started process bplabel (3896)
11/06/2015 10:22:06 - mounting AG10L2
11/06/2015 10:29:03 - Error bptm(pid=2668) error requesting media, TpErrno = Robot operation failed    

Cheers,

Steve

[Nicolai · Accepted solution]

Delete tape devices and robot in Netbackup and re-configure using the device configuration wizard in the Netbackup GUI.

If you get the location wrong in the robot whereas Netbackup think it is would be is worth looking at.

What does the Windows event log say ?

[SteveDown]

I've rebuilt the server and configured the TLD via the wizard. All appeared to go well and the Event Log is clean.

When trying to run a backup I get the following.

11/06/2015 11:12:35 - requesting resource Any
11/06/2015 11:12:35 - requesting resource veritas-backup.NBU_CLIENT.MAXJOBS.veritas-backup
11/06/2015 11:12:35 - requesting resource veritas-backup.NBU_POLICY.MAXJOBS.Test-Backup
11/06/2015 11:12:35 - awaiting resource Any Reason: Drives are in use., Media Server: veritas-backup, Robot Number: 0, Robot Type: TLD, Media ID: N/A, Drive Name: N/A, Volume Pool: NetBackup, Storage Unit: veritas-backup-hcart3-robot-tld-0, Drive Scan Host: N/A
11/06/2015 11:20:56 - granted resource veritas-backup.NBU_CLIENT.MAXJOBS.veritas-backup
11/06/2015 11:20:56 - granted resource veritas-backup.NBU_POLICY.MAXJOBS.Test-Backup
11/06/2015 11:20:56 - granted resource AG10L2
11/06/2015 11:20:56 - granted resource HP.ULTRIUM3-SCSI.000
11/06/2015 11:20:56 - granted resource veritas-backup-hcart3-robot-tld-0
11/06/2015 11:20:57 - started process bpbrm (1044)
11/06/2015 11:20:57 - connecting
11/06/2015 11:20:57 - connected; connect time: 00:00:00
11/06/2015 11:20:59 - mounting AG10L2
11/06/2015 11:27:40 - Error bptm(pid=3748) error requesting media, TpErrno = Robot operation failed     
11/06/2015 11:27:40 - Warning bptm(pid=3748) media id AG10L2 load operation reported an error     
11/06/2015 11:27:48 - Error bptm(pid=3748) NBJM returned an extended error status: resource request failed (800)   
11/06/2015 11:27:48 - Error bpbrm(pid=1044) from client veritas-backup: ERR - bpbkar exiting because backup is aborting  
11/06/2015 11:27:40 - current media AG10L2 complete, requesting next media Any

11/06/2015 11:27:51 - end writing
extended error status has been encountered, check logs(252)

[mph999]

OK, try robtest from the robot control host

Run robtest, then try and move a tape too the drive, and back again, like this

[root@prod-master1 ~]# robtest
Configured robots with local control supporting test utilities:
  TLD(0)     robotic path = /dev/sg3
  TLD(1)     robotic path = /dev/sg4

Robot Selection
---------------
  1)  TLD 0
  2)  TLD 1
  3)  none/quit
Enter choice: 1

Robot selected: TLD(0)   robotic path = /dev/sg3

Invoking robotic test utility:
/usr/openv/volmgr/bin/tldtest -rn 0 -r /dev/sg3

Opening /dev/sg3
MODE_SENSE complete
Enter tld commands (? returns help information)
m s1 d1
Initiating MOVE_MEDIUM from address 1024 to 1
MOVE_MEDIUM complete
m d1 s1
Initiating MOVE_MEDIUM from address 1 to 1024
MOVE_MEDIUM complete
q

[watsons]

1) Have you checked and removed ghost devices (common in Windows)?  See http://www.symantec.com/docs/TECH93292

2) Have you applied the device mapping?  See http://www.symantec.com/docs/TECH228936

3) After Netbackup discovers devices, what do you see when running    \volmgr\bin\tpautoconf -report_disc 

Something is not right unless you get an empty output.

[SteveDown]

If i run the m s1 d1 command, this fails.

Opening {6,0,0,1}
MODE_SENSE complete
Enter tld commands (? returns help information)
m d1 s1
Initiating MOVE_MEDIUM from address 1 to 1001
move_medium failed
sense key = 0x5, asc = 0x3b, ascq = 0xe, MEDIUM SOURCE ELEMENT EMPTY

[SteveDown]

After checking Windows Device Manager, I can see the tape drive and autloader. However, on my other server these don't appear. I can only assume this is causing some sort of conflict.

I deleted them, however when I attempted an Inventory it failed. I've reinstalled the Windows drivers and I can run an Inventory again. Seems odd that one server displays these devices in Device Manager, yet the other server does not and it works perfectly well.

[watsons]

Which one fails?

m s1 d1    

or 

m d1 s1

You got to have a media tape in the source for this command to work (without that error you show). That means:

m s1 d1  (needs slot1 to have a media, check by doing "s s" : show slot)

m d1 s1  (needs drive1 to have a media, check by doing "s d" : show drive)

[Marianne]

You need to unload first: unload d1 Then m d1 s1 If I may ask - why W2003? Why NBU 6? Have you disabled Windows Removable Storage Manager service?

[SteveDown]

Annoyingly, I have fixed the problem by simply removing the config again and reinstalling (third time). I didn't do this any differently, though. :\

[Nicolai]

Did you follows Mariannes advice ?

Have you disabled Windows Removable Storage Manager service?

[SteveDown]

I haven't disabled it, however despite being set to Automatic, it hasn't started. Is it always best to disable Removable Storage Manager?

[Nicolai]

Please do

[Marianne]

It MUST be disabled. RSM will 'fight ' NBU for device control.