Backups Fail Status 800 - New Robot

[DanTheMan_2]

Hi,

NB 6.5 on Windows Server 2003 Standard SP2.  Attached to a Fujistu robotic library and it performs as it should.  But we connected another library to the same server and those new backups configured to use it are failing with status 800.  Is there something we've missed.  What other info do you need?


Thanks
:)

[Bill_Johnson1]

A status 800 is very generic. You should be able to look in the bptm log on the media server to get the exact error or issue. If you would like to post the bptm log from the time it occurs, I can look at it.

[Will_Restore]

I bet it did something. ;^)

[DanTheMan_2]

I don't have a bptm log.  

The whole message in the status of the failed backup is:

"Errro nbjm(pid=2108) NBU status : 800, EMM status: All compatible drives are down, but media is available resource request failed (800)"

And the drive (the new one) is showing as down.  I can 'up' it again no problem but backups still fail.

[Will_Restore]

Details:

Overview:
This status code is new in NetBackup 6.5.4 and covers some cases that may have formally been reported as Status 800.


Error Messages/Symptoms:
EMM_ERROR_DriveUnavailable:          = 2005001
EMM_ERROR_DrivesDown:                = 2005008
EMM_ERROR_DrivesDownMediaAvailable:  = 2005009

Troubleshoot these issues in the same manner as when the errors occured in versions prior to 6.5.4.  See the related documents section for a technote on "DrivesDownMediaAvailable" conditions.

 

Related Documents:

278533: STATUS CODE: 800, NetBackup 6.0 jobs are failing with Status 800 (resource request failed).
 


Verify the storage unit is configured with the correct number of available drives. Also, verify the density of the storage unit has been set correctly and matches the available media. If this density is set incorrectly the storage unit will be unavailable and the backup can fail with a Status 800.

[DanTheMan_2]

 Hmm good shout.

Windows Event id: 168

RSM cannot manage library Tape0.  Failed to communicate with device or obtain device setup information.



From what I've managed to find out this could be something to do with the RSM service and NB trying to manage the device.

[Andy_Welburn]

"...
6.  Removable Storage Manager (RSM) in Windows 2000 and 2003
In most configurations, the Operating System's Removable Storage Manager service is not required. In some configurations, stopping this service can resolve robot communication issues in NetBackup. Before disabling RSM, ensure that the service is not required for other unrelated purposes, such as removable hard disks.
..."

Taken from: 
GENERAL ERROR: How to troubleshoot robot communication issues in Windows ...

also:
3rd PARTY: Removable Storage Service set to Manual or Automatic and running. ...
 

[DanTheMan_2]

 I've disabled the service and am watching a job I just kicked off now...

It says that it's mounting a tape and has been for the last 8 minutes!

Other Events from the earlier failure:

ID: 1402

TLD(1)(1468) Times out after waiting 855 seconds for ready, drive2

And before that...

ID: 2636

Operator/EMM server has DOWN'ed Fujitsu-blah (device1)

[Marianne]

Any device errors in System log? 
Does the robot have a display panel? If so, are there any errors on the display panel?
If a tape gets loaded in a drive, does the drive come back with 'Ready' status (on display panel or status light on tape drive)?
How are the robots attached to the server? Are they SCSI or SAN attached? If SCSI - have you checked total cable length on each SCSI bus?
How many HBA's (fibre or SCSI) in the server?  How many devices per HBA?

[DanTheMan_2]

I have a Fujitsu engineer going to look at the robot today (hopefully).  It's in a remote data centre so I can't do much with it.  It is SCSI, but seeing as it was installed by a Fuji engineer I'm hopeful that they know what they were doing re the connections.  They added a new SCSI card to the server specifically for the new robot.

The robot can complete an inventory OK so I'm kinda leaning towards it not being  that kind of problem.  But then again, I'm quite new to this so maybe not!

Thanks for your efforts.
 

[Will_Restore]

Was kidding 2 days ago but may be more right than I knew! 

[DanTheMan_2]

Well nothing wrong with the library and I'm still getting 800's.

I disabled RSM and now some of my working backups have failed (event viewer says that Netbackup was expecting a certain barcode).

Thanks for help on this but I think it's time to get Symantec support involved.

[J_H_Is_gone · Accepted solution]

I sounds to me like you have a bad config going on.

check the server the tape drive is showing up on.
find out what rmt goes to what drive serial number

so now you have tape drive serial number 0001 going to rmt3
in netbackup see if rmt3 is serial number 0001

for example
if you have serial number 0005 showing up as rmt1 in netbackup
then NB loads a tape in 0001 drive, (which really is rmt1 on the server) and the server keeps looking at rmt3 for a tape and never sees it.

[DanTheMan_2]

No too sure what an rmt is.. as in /dev/rmt/ ?

Anyway, drive 0000 is attached to robot TLD0 (and works).  Drive 0001 is attached to robot TLD1.  Netbackup agrees with that as well.  All the serials match.  Are there any windows cmd lines I can confirm with?

[Marianne]

/dev/rmt are Unix device names.
You can run these commands from ...Veritas\volmgr\bin to confirm:
scan
tpconfig -l

I know you said there's 'nothing wrong with library', but please get the hardware engineer to tell you what the SCSI cable cable lengths are. Too long cables will not show any errors on the robot itself and comms up to the robot (including device config and inventory) will work fine. Problems are seen when the drive has to report 'READY' status back to the O/S.

[Karthikeyan_Sun]

 If you are well confident no problem with Library,

1. Check Tape Drivers installed, Try re installing it.

2. Try to verify the Drives from Veritas- run tpconfig -l,robtest

3. If all Tape Drives are visible and well, Just re run your Configure Storage Device Wizard.

4. If Storage Config is successful, try checking status in Deice Monitor and Test the backup !

Let us where ever you face problems in the above steps.

[DanTheMan_2]

 You guys were onto something with the drives.

Netbackup was scanning for devices (Storage device wizard) and telling me I had 2 drives available to the robot but could only use 1 of them, with the other blacked out.  I changed the drive number from 2 to 1 and managed to get a successful backup and restore!

I've scheduled a full backup for tonight so will know for sure tomorrow morning.

Thanks again.  Just one more question:  is it sane to have a policy dedicated to diffs with a policy dedicated to weekend fulls?  I.e. using seperate tapes for diffs and fulls?  Without a full to begin with diffs can't possibly work, but how is that handled by splitting them over policies?

[J_H_Is_gone]

you should have one policy with different schedules
if you want them to use different volume pools have the schedule override the selected volume pool and use a different one.