Storage device reported an error on a request to query ...

[EGART]

I've got Backup Exec 8.0 installed on a MS Windows NT Server 4.0 + SP 6a (BDC) and the backup hardware is a COMPAQ Storageworks TL891 robot with 2 internal DLT drives.

The following error was registered by the Job's log file:

----------------------------------------------
Storage device "COMPAQ 4" reported an error on a request to query entity information.

Error reported:
The object invoked has disconnected from its clients.

^ ^ ^ ^ ^
An unknown error occurred on device "COMPAQ 4".^ ^ ^ ^ ^
----------------------------------------------

MS Windows NT Event Viewer reports the following error IDs:

Event ID 57665, Event ID 57600 and Event ID 57745

I've tried to find something on the Veritas KB without success. Can you please give me some help?

 

[Tourist]

I've had several different issues with Veritas backup Exec before, including the error you're getting. I managed to resolve the problem by defragmenting my data before the backup.

Here's what I usually do when faced with a problem with Veritas Backup Exec:

1) Stop and start the services
2) shut down the machine, and restart (no reboot, a real shut down, some machines need time to empty buffers)
3) defrag the data
4) Replace the DLT drive In a world without walls and fences, who needs Windows and Gates?

 

[bobafet]

Hi Egart,

Try upgrading the software to 8.5 build 3572 rc9 (which is a free upgrade from version 8.0).

http://seer.support.veritas.com//docs/235770.htm

BENT v8.5 build 3572 rc9 (Intel) (English) (SBS Support) Serial # Req.

(if you have any remote agents or remote open file option agents installed, you will have to re-push out the agents from the new install)

I am confident that this build should resolve your issue...

Cheers
Bobafet

Note: Backup Exec for NT/2000 Ver 8.6 3878 is the latest version but is a purchasable upgrade from 8.5 and below as it has new features like backup to disk, enhanced lotus notes option etc.

 

[Tourist]

If upgrading to version 8.5 build 3572 rc9, like Bobafet mentions, doesn't do the trick, you could try deleting your schedule and running the job that fails manually once.

I was faced with the same problem again last week and this time I got around it by deleting the schedule, running the job manually (Run Now) and rescheduling the job.

Let us know how it turns out

In a world without walls and fences, who needs Windows and Gates?