B2Tape submit as queued and never start

[exxcomm]

As the title says, any job (newly created or old scheduled) with the tape drive as the target resource will read as "Active - queued" and never start. I'm using Backup Exec version 12.5 Rev. 2213 installed on M$ Server 2003 Standard SP2 using a HP Ultrium 3 drive with HP 1x8 G2 autoloader.
The backup jobs to tape would work correctly until we started using a B2D solution using an iSCSI SAN attached with M$ iSCSI iniitiator. New D2T backup jobs will work fine until the first B2D job was tested without putting the D2T job on hold, so they both ran at the same time and locked up the server.
Since rebooting the server from that freeze any D2T job will start for 15 to 30 sec.  and then report a status of "queued" until it is manually cancelled. D2D runs as expected, D2T just hang in "queued" limbo forever.
The BE Support tool doesn't find anything wrong except the agent using localsystem credentials instead of the login accct. that all other services are running under. When credentials are changed BE support complains it's not a localsystem acct. (go figure).

This is not a new issue and goolging "Backup Exec starts jobs queued" finds many complaints of this behavior on Experts Exchange, TekTips, and this forum. Suggestions to fix this issue (submitting a job to fail due to hardware paused, pausing and unpausing the media server, etc.) have not worked and I would REALLY prefer not to un/re-install the BE software as this has created HUGE sql issues for me when it was tried at a previous time using BE 11.

Any suggestions folks?

[pkh]

Have you tested a backup to the library using NTBackup?

You have obviously tried a lot of the suggestions in the forum.  Maybe you can list them so that we can see if there is any other suggestion.

[Dev_T]

Hello,

The job is in queued state means backup exec is unable to communicate with the library

The best way would be to powercycle the network.
1  : Turn OFF the server and the Library
2  : Disconnect the Library and turn ON the server (The Library entries will no longer be present in the OS)
3  : Turn OFF the server and conect the library
4  : Turn ON the library and let it come up
5  : Turn ON the server and check the device manager for the changer and the drive
6  : Install Symantec Drivers and Change the changer to Unknown Medium Changer (Microsoft Drivers)

Hope this helps....

[RahulG]

You can try the following step
Make sure you have the latest drivers install
Delete the robotic library from the console
restart the servcies which woul reinitalize the Rl
Check in the event log if there is any error
Try running the job and see if you have any errors in the event log
 

[exxcomm]

pkh - I had tried to view/write to the tapes using NTbackup but this is a server that has been around a while and lists 4 different tape drives a 4mm DAT with media from 2005 (no longer used), File (no media), LTO Utrium (no media), Unrecognized (listing 5 of the 8 tape slots in the media library).
All of these devices have greyed check boxes so I cannot look at any of the drives/media. I don't think NTbackup has been used for quite a while.
MediaTools from the tools entry in the menu bar is greyed out too, but device manager does see the devices and rediscovers them after removal.

Dev T - I had done what you suggest except for changing the drivers for the the tape drive/library and physically detach the SCSI cable. I had uninstalled the devices, turned off the tape/library device, rebooted the server so it didn't "see" the drive/library, shutdown server again, turned on the tape/library, and booted up the server so it could rediscover the tape which it has done every time.
Using your exact procedure has the same results as before. Drive/library is rediscovered and found in device manager, but a catalog of the tape library results in "Ready, no servers found"  This is the same as I was experiencing after a stop/start of the services without rebooting the server. After a reboot it would report "Queued".

RahulG - Your procedure was the last listed and seemed the most invasive (and has proven so, so far) so I saved it for last. I get VERY nervous about deleting devices within BE as then you start really messing with the SQL backend which is NOT a good time.
But, i did as you suggested and deleted the library and then restarted the services. It did re-find the library, but did a renumbering of the tape drive (from HP1 to HP2 which used to be the name of the library) and then renaming the library to HP3. It sees all the slots in the tape library but has them as "empty" (which I know they are not. Magazines have been verified). An initialization ran for 1hr 10 min. so far with no activity showing up on the tapes in the devices tab. The drive will flash a "play" icon for about 10 sec. and then it goes away. Killing the initialization is taking quite a long time (~40 minutes so far) so this fix attempt may have just added to the issue I'm having.

Thank you all for your assistance. This is an OLD server whose history I have not been a part of (I'm a contractor). It seems to have a few other issues besides these, but except for the server freeze when it tried to run both FULL jobs, it did see the tape library and write to it just fine.
I've used BE since it was a Veritas product (BE 8 or so), but have become less and less impressed with it since it started using the SQL backend.

[pkh]

My suggestion of using NTBackup is to test whether there is any hardware problem with the tape drive.  If NTBackup can backup to the tape drive, this means that there is something wrong between BE and the tape drive.

[exxcomm]

Using HP LTT utility it reports "Changer0 is opened exclusively by another application/service. It cannot be accessed until it was <sic> released."
This is why neither BE or NTbackup can access the drive. I'm opening a support ticket with HP to find out why/how this could occur (RSM is disabled on the server so it's not that.)

[Khue]

If this is an old server, when was the last time the Ultrium drive got a lil' tlc? Maybe it needs a firmware upgrade? Perhaps I missed it so forgive me for asking, but what kind of interface does the auto loader have to the media server?

[Ken_Putnam]

Using HP LTT utility it reports "Changer0 is opened exclusively by another application/service. It cannot be accessed until it was <sic> released."

You need to shut down all BackupExec services before running either NTBackup or  LTT

[exxcomm]

But HP sent a new one after a 3hr. and 44min. support "trial". Installed the new drive today. Un/re-installed BE and M$SQL and it was looking good.
Then I copied over the Data and Catalog folders (to try to rescue my previously built jobs as suggested by Symantec) and it renumbered drive/library and an inventory says "Ready, no media server found".
Next step is un/re-install BE and M$SQL AGAIN and rebuild all backup jobs. Fun.

I did learn something from the HP support call though. It seems that adding D2D jobs to a BE install already doing D2T will kill the D2T ability as found here:

http://kbalertz.com/842411/Windows-Server-cannot-perform-backup-devices-Storage-Network.aspx

I'll chime back in after the un/re-install, rebuild of jobs, and testing.

[exxcomm]

I would have to suggest NOT deleting the tape library from BE as that is when I started having the drive/library renumbering issue.

[gray_CG]

I have exactly the same issue, however this is only happening on two of my 17 backup jobs. I'm certainly not going to mess about with removing/adding the tape drive or recycle the server when this problem is only happening to two of the jobs, all the rest backup without any issues... I've been trying to find a solution for this for about 2 weeks now to no avial!