C1537A Write / read? problems?

[CA1347421]

I am receiving a set of puzzling messages regarding my tape drive, specifically, when I attempt a backup, Backup Exec tells me part way into it that the drive couldn't recover a write error.
Using the latest version of StorageWorks L&TT, I perform the DDS Drive Assessment Test, and everything checks out okay. (I installed the most recent Firmware as well).
When I return to try a backup, Backup Exec fails to complete the process with the same error message.
Any thoughts or suggestions anyone?

[CA713937]

Does the backup fail on new media?

Can you post the LTT support ticket so we can see what the fault logs indicate?

What is the exact message in the backup app?

[CA1347421]

Thanks for the reply.

Yes the backup fails on new media (the LTT assessment failed on new media once as well - I discarded that tape).

I've posted the most recent event/result log, but it is from an LTT session that didn't report any problems...

Also, in all cases, upon failure, the front amber LED blinks 2x/sec.

I have just attempted to reproduce the backup error to capture the message but the failure didn't occr...sigh (1st time in a week, of course). As soon as it fails I'll send the exact error messages along.
Leo

[CA713937]

Leo,

Thanks for the additional details. It sounds like the backup failure, and the assessment test failure, are intermittent.

The log you attached is the event log from the tool logs folder, and not a support ticket. The support ticket is generated from the "Support" tool, and then "save ticket to file".

Is your product still under warranty? The event log states that the drive may benefit from recalibration, which typically the support center will walk you through.

[CA713937]

> Also, in all cases, upon failure, the front amber LED blinks 2x/sec.

That's typically the "Clean" light, which when flashing indicates the drive may require cleaning (typically as the result of a hard write or read error). The event log did not indicate that cleaning was suggested however, so the drive logs may have indicated that the drive was recently cleaned.

Have you cleaned the drive recently with a non-expired cleaning tape?

How often do you clean the drive (with a cleaning tape) in general?

[CA1347421]

The drive is not under warranty as best as I figure. It was replaced about 18 months ago.

Since then, I have cleaned it regularly - at least once/week before a major backup. When this latest event occurred I cleaned it several times with a new tape, and each time it blinks I clean it (about 9 times now in the past three days).

I don't think it needs any more cleaning.

After my morning message, I tried a full backup and got the dreaded message you were asking about. Backup Exec says:
"Tape Alert: Your data is at risk" blah blah blah
"Tape Alert: The operation has stopped because an error has occurred while reading or writing data which the drive cannot correct."
Closing that box, Exec posts a new box:
"An error occurred while writing the backup data. A defective or damaged area was detected on the media.... (it goes on to provide gratuitous information)."

I thought a recalibration might be in order, but the LTT utilities don't suggest it's required and I'm loathe to do it as I only get one shot at it.

I'll try and create a support ticket in the event I can reproduce the failure and post it later.

[CA713937]

Thanks Leo. The tape alert messages are a generic error reporting mechanism, so they aren't all that detailed, but they all suggest that the drive is having read/write problems. The drive fault logs will give us the best information.

BTW, you can run the device analysis script at any time (not just after a failure), and it will review previous faults. It had previously suggested recalibration (when run automatically at the beginning of the assessment test). Providing a support ticket will give us the error history and other use details to help understand the root cause.

[CA1347421]

Thanks David for your continuing interest.

Okay. I ran three tests: a complete tape erasure (took 11,000+ seconds!) which completed successfully, a "device self test" which caused the LTT app to hang (sent a report separately on that) and the "device analysis" which completed successfully ::with the one warning to suggest I run the recalibration::
A ticket is attached for your viewing pleasure.

Perhaps this will lead to a positive resolution. Thanks again.

[CA713937]

Leo, I've asked a few of our DDS experts to review the ticket and make their best recommendation.

Did you report the LTT hang to ltt_team@hp.com? That's the best place to report possible tool issues.

Is the problem reproducable?

Sometimes we see issues like that when using an unsupported/untested HBA. What specific HBA are you using? I'm guessing the sym_hi driver indicates some type of LSILogic or Symbios card.

Also, is windows RSM (Removable Storage service) disabled for this device? If not, it may try to manage it while LTT is accessing it.

[CA1347421]

David,

Thanks for sending along the ticket.

Re: HBA - I assume you mean the adapter - Symbios Logic

Re: RSM - I can't see how to disable it on a device specific basis - it's set as an auto-start service at present

Re: LTT hangs - yes I sent those reports directly to the LTT team.

Thanks again.
Leo

[CA713937]

Leo,

Which specific model of Host Bus Adapter (HBA) do you have?

Have you tried to see if the hang duplicates? The fault log of the drive did indicate one firmware hang.

The cause of the hang could also be from RSM.

To disable RSM in win2k, go to Adminstative tools, Services, right click on "Removable Storage", select Properties, click on "Stop" if it is not stopped, and then change the startup type to "disabled".

[CA1347421]

Hi David,

The HBA is a Symbios Logic 896, 22910 PCI SCSI Adapter

I remember removing RSM before, but I think windows updates keeps re-initializing it.

It's now disabled.

Regarding the hanging of LTT - it's occurred twice and both times I have sent the automated email to the LTT team. It has occurred from two different operations - one associated with a device analysis and another I can't recall.

[CA1347421]

David,

Just checking in - should I wait for further word on running the recal script or should I just go ahead and do it?

Thanks again for your ongoing help.

Leo

[CA713937]

Leo, sorry my engineering contacts in Bristol (UK) were out for the last two days. Hopefully I'll hear back by tomorrow morning US time.

[CA713937]

> The HBA is a Symbios Logic 896, 22910 PCI SCSI Adapter.

It's not on the list of supported 3rd party HBAs, so it would not be a specific HBA that we tested with LTT. There other LSI logic HBAs supported however, which usually increases the chances that it should work.

http://www.hp.com/products1/storage/compatibility/tapebackup/ThirdParty/index.html#HBA

> I remember removing RSM before, but I think windows updates keeps re-initializing it. It's now disabled.

> Regarding the hanging of LTT ...

Have the hangs occured with RSM disabled?

[CA756694]

Leo,

If you post your email then I'll contact you directly.

I work for HP in the division where DDS is developed. I'll see what I can do.

cheers

[CA1347421]

John / David,

Thanks for all of the replies...

David - that's funny that the HBA listed isn't one of the supported devices - the HP NetServer 800 came with the dds built in...

John,

My email is in the attachment (don't know how public this forum is to spiders...)

Also, any of your replies come to my email automatically...

Thanks for getting in touch, I was about to recalibrate if I hadn't heard from anyone.

[CA713937]

Leo,

> David - that's funny that the HBA listed isn't one of the supported devices - the HP NetServer 800 came with the dds built in...

OK, that helps explain it. The supporability matrix on the external web is for the current products being sold (the DDS3 drive is around the 4th generation - yours was a 1st generation drive), and it does not list the legacy HBAs that were supported (such as in the discontinnued netservers).

We have had some problems running LTT on netservers with the internal netRaid controllers when customers have upgraded to 2003 (to the best of my knowledge, that is not a supported OS on netserver hardware platform).