Constant DLT-S4 problems

Hi there,

What OS are you running? If Server 2003 make sure the Removable Storage service is stopped and disabled.

You should also be checking the tape errors as well. Open BE --> Devices --> check the tapes for any Soft Write Errors, and more specifically, Hard Write Errors. If you're getting those, try backing up with a new tape to see if this occurs again.

Some other things to try:

1. Make sure you've got SP1 and the latest drivers and patches loaded on your media server.

2. Download the Quantum drive utility, stop the BE services, and run this against the drive to rule out definite hardware issues.

3. Make sure the SCSI cable has no breaks/faults with it, and that it is terminated correctly.

Thanks!

Running Svr2003 R2 SP2

Removable Storage is disabled and always has been.

I checked the jobs and it reported there had been critical read write errors with the device.

We're not running the latest (which would upgrade the DLT drivers upto 5.1.37.0) because this issue has now happened three times after a succuessful period of backups. Even in the past when I have updated it's had no impact whatsoever and I see no reason why this would be the case now because backups were working (the last three days the byre count has failed at 7gb, 38gb and 138gb so no pattern with the job requesting I load media). If it was a new install it's probably the first thing I'd do.

I'm running xtalk now (even before this time xtalk reported no issues the last time it happened, though a replacement device resolved it).

Cables are fine.

As mentioned, the backups have been running fine since 22/11/11 and then out of the blue backups start failing.

So far the failure point has been the S4 on all three occasions. It's bizzare that these units are failing after 8 weeks of use.

Thanks though

Thought I'd provide an update on this.

xtalk health check ran and completed (awaiting feedback from Quantum on this).

Performed a full read/write and it stopped at 1h 30m

Having done nothing apart from re-starting services after quitting xtalk the backup of last night was successful. It's almost as if the xtalk diags have reset something on the unit.

Mmm...the diagnostics MIGHT reset error statistics if that is ever an option, otherwise they'd just pull logs and report on them.

Are there any xtalk services running on the device? What is the make of server as a matter of interest?

Dunno, but we're going Mmm...... a lot.

No xtalk services, it's simply an app that recommends you stop any services utilising the device. I always stop all the BE services prior to running.

Server is a HP Proliant DL380 G4 externally coupled up to a HP Tape storage unit.

...stop the HP storage services. This has been known to cause repeat SCSI timeouts as on certain occasions, it hammers away on the SCSI bus...there is an HP alert on this.

I guess you're refering to the HP Insight Storage Agents?

Would you expect it to take 6 weeks before the backups begin failing and it's only this type of unit we've experirenced problems with, SDL-600 was previously installed.

Btw, we received this error from a Catalog yesterday

(Server: "TIMOR") (Job: "Catalog 00039") Catalog 00039 -- The job failed with the following error: A backup storage read/write error has occurred.

If the storage is tape based, this is usually caused by dirty read/write heads in the tape drive.  Clean the tape drive, and then try the job again.  If the problem persists, try a different tape.  You may also need to check for problems with cables, termination, or other hardware issues.

If the storage is disk based, check that the storage subsystem is functioning properly.  Review any system logs or vendor specific logs associated with the storage to help determine the source of the problem.  You may also want to check any vendor specific documentation for troubleshooting recommendations.

So I stopped all the BE Services, ran another xtalk health check and it completed ok without any errors reported and last nights backup completed without error.

Also, having looked around we're not getting E00084F4 which relates to the management agent issue by all accounts.

...that's right.

The drive itself is on the BE 2010 R3 HCL so it is supported.

Have you tried to completely uninstall the library as follows:

1. Delete the library out of BE.

2. Delete the library out of Windows Device Manager and turn it off, and disconnect the SCSI cable.

3. Restart the server, let Windows boot and verify the library doesn't show up in DM.

4. Shut down the server, reconnect the library and start it up. Once initialised, start up Windows.

5. Load the Symantec drivers again and run a backup job.

I wouldn't know how long the backups will take to fail again, and hopefully they don't, but if I did, I'd make a LOT of money selling my services for preventative troubleshooting!

Pretty sure the device was on R1 also from what I recall from initial investigation.

I haven't tried deleting the Library because you have to go through all the existing jobs and re-assign them, if you tell the system to use/assign to another device, it doesn't work.

Also, when you add a new device, it creates new said library so I sought of removed that possibility a while back. Albeit the old device names are still vivible but inactive

Because it has worked immediately for a period of time after a new device is added I'm, though I recall deleting it once before and this made no difference.

I have a little black book of all the tasks I've performed with this issue, maybe I too could make a lot of money by turning it into a published article. LOL

...if not, why not. Go for it...it might actually be worth it to try removing the library completely, and this would rule out any sort of software corruption in drivers etc.