Trickiest ever problem with BackupExec 10d, Dell Tl2000, Windows 2003

[AntonyPaul]

Hello,

Have a major problem with BackupExec 10d backing up to a Dell TL2000 (IBM TotalStorage 3573 Medium Changer) which uses the IBM ULT 3580HH-4 tape mechanism.

Communication is working between BE and the backup device as I can Initialise the library, run cleans etc that do work, it simply will not write any bytes to the tapes.

BE DIag file here - http://pastebin.com/JuWvCp10

And some errors in event viewer

Backup Exec Alert: Catalog Error
(Server: "CHEFFINS-FILE") ODBC access error. Possible lost connection to database or unsuccessful access to catalog index in the database.

Update to catalog index (Catalog index database) failed.
Reason:  [Microsoft][ODBC SQL Server Driver]Invalid attribute/option identifier cat_RecordSet::Delete()
e:\caicosmp2\5629r\becat\beplugins\beimage\catmedia.cpp(780) 

This all started after a PAT tester turned off the Tape Library...

Grateful for any advice.

Thanks

[CraigV]

Hi Antony,

Chances are that you have no connection to the BE database.
First thing to do is check to make sure you have that connection. Check the link below:

http://seer.entsupport.symantec.com/docs/256661.htm

Post the results...

[AntonyPaul]

Thank you for your fast response.  I followed the instructions and it reported Test Connection Succeeded.

I realise now that those error messages were probably generated when I was stopping and restarting the Backupexec services as that seems to be the only way to clear the jobs which, after cancelling, are just stuck on Cancel Pending.

BE reports that all services are currently running.

[CraigV]

OK...next step:

1. Stop your services for BE.
2. Go to your installation directory, and rename your Catalogs folder to Catalogs.old
3. Start up BE again.
4. Try another backup job again...

[AntonyPaul]

OK, did that.  It has created a new Catalogs directory.  Now it is showing the "No Idle Devices are available" message.  This is what it started saying right at the start of this ordeal.

They say a picture paints a thousand words so maybe this might help.

http://img704.imageshack.us/img704/3115/symss.jpg

I should add that the Library is online and enabled under the devices tab and nothing seems out of the ordinary there...

cheers

[CraigV]

Temporarily fix 1 problem, go back to the original...got to love it :)

OK, in this case, just try create a new backup job. It doesn't have to be very big. Run it, and post the results. We can take it from there...

[AntonyPaul]

OK, good plan, started  a new job, fiddled around with the target device and 4GB in so far so good.  I'm not entirely certain what caused all this, although I read on another thread here that reinstalling drivers for Medium changers doesn't always properly update pre-existing job device settings within Backup Exec.  Bit unclear on all this, I'm not sure if I'll need to tweak each specific daily job after this, will investigate this more once this backup has completed.  Thanks a lot for your help Craig

[AntonyPaul]

Alas, it was too good to be true.

The job that was running had stalled on 449GB

I started another which has now stalled on 467GB

Usual amount of data is c. 920GB

When I have cancelled these jobs it stays on Cancel Pending and the only way for me to clear it is to restart the services.  Most peculiar.

When I look at the log for the first job. I get this error message: One or more errors were encountered converting the job log into a web page for display.  The job log will be displayed in text format, I presume because I had to force the job closed by restarting the services.

Job log then opens http://pastebin.com/heJWdSxy

2nd job log (after forcing cancel by restarting the 4 BE services() http://pastebin.com/Rr8CVfqX

There's some concerning alerts in there, it's just trying to figure which one is responsible for the stalling.

Used different media both times.

Do  you think this is caused by Volume shadow copies. I had disabled these on the file server however the 2nd log kind of implies I didn't - or that traces of them remain - which I thought wasn't the case when disabled VSC.

[CraigV]

Mmm...OK, if you're running AOFO, deselect the option in the job to snap each volume 1 at a time and see if it goes further.
Is it prompting you to replace the tape perhaps?