Help I'm in autoloader hell ! ! !

[Mats Larsson]

Hi !

We are running a HP SureStore Tape AutoLoader Model 1/9 together with ArcServe 2000 (NT4 server SP6).

HP Library & Tape Tools is version 2.7 SR4.

The Library (C7145-8000) is using firmware 2400 and the drive is using firmware version v059.

The drive is a Quantum DLT 8000.

I'm having lots of difficulties with it recently.
It began with me not being able to restore an old backup. After that I couldn't write new backups.

When running the tests in the "hp library & tape tools" the "DLT SRAM check (short) fails. Looking under the "Test Result" log I get the following message (among others):

FIFO(SRAM)CRC errors detected during write phase.
Test concluded: this drive exhibits FIFO(SRAM) CRC error symptoms. Please replace the drive !

What does this mean ??? What is the test doing ? It completes the test after writing the 2GB of data and then this message.

Most of the other tests seem to run fine except for the "data compression test" wich seems to completely freeze the library & tape tools application.
The Device Analysis test passes with warnings saying that "this device has logged one or more 'log page 7' errors this drive may need exchanging."

The errormessages I got when trying to restore data from a backup in Arcserve2000 was this:

TAPE E6084 Drive hardware error - servo or sensor [ABSL:2010 CMD:8h]
TAPE E6921 The tape drivehas a hardware fault:
Turn the tape drive off and on again.
813 E3713 Unable to read from media (MEDIA=F, EC=CONTROLLER FAILURE or HARDWARE ERROR)


Anyone who has any ideas or suggestions PLEASE reply to this !

Many, many thanks in advance !
/Mats

[David Ruska]

> Test concluded: this drive exhibits FIFO(SRAM) CRC error symptoms. Please replace the drive !

> What does this mean ???

Please see the following support document:
http://www.hp.com/cposupport/information_storage/support_doc/lpg75068.html

You most likely updated to newer firmware (V59 for libraries), and this test (actually the drive) is now detecting a CRC error (aka checksum error) in the FIFO memory used to buffer data.

The drive needs to be REPLACED.

> I'm having lots of difficulties with it recently.
> It began with me not being able to restore an old backup. After that I couldn't write new backups.

Your new backup failures are due to the SRAM errors being detected on writes. In earlier firmware versions, this error was not detected and incorrect data could have been written. Your restore failures are either due to FIFO CRC errors on the reads, or a corrupted tape format from incorrect writes. Any backups made with firmware prior to V59 are suspect.

Note: The potential for this problem appears to begin with drives having more than 5000 power-on hours.

Please see the above referenced document for more information.

[Mats Larsson]

Thanks a lot for the fast reply ! I'm calling support now for a replacement drive !

Any ideas why the compression test freezes the aplication ?

Many thanks for the reply ! (I would have given it 10 points but I really don't know if it will solve my problems until I get the new drive)

/Mats

[David Ruska]

> Any ideas why the compression test freezes the aplication ?

If it is encountering the FIFO (SRAM) errors, it may not be handling the error condition correctly. We will attempt to reproduce that.

Thanks for pointing this out.