MSL6030

[William Grace_1]

Hello,
I have been trying to get HP to help with my problems, but since slugs are quicker and more responsive, I thought I would try here.
The tape library just stops working, and when using Tape Tools the exersize test will return Errors. After a while the Drive will work again, but the performance is getting slower and slower. I think the Tape Drive within the Unit itself is on its last legs. Any ideas?

[David Ruska · Accepted solution]

William,

I can take a look at it for you. The ticket you provided is missing the separate .dat files (the ticket is split into separate files for most multi-drive libraries).

Can you zip up the ticket and .dat files that were created when the ticket was saved and repost it?

[David Ruska]

If you can also provide the resultlog and eventlog files from the LTT logs folder (zipped) that will provide details on the exercise test failure.

[William Grace_1]

Thank you David. I have tryed sending the files Zipped,and I still get the returned email and when I check the call, they never get added. I have attached a zip of the tickets.

As things stand I have replaced the drive in the unit, but I am getting throughput of 53MB per minute. I should be getting 530MB per minute. With a 2 Tera byte backup to do, it will take forever at this rate.

My next guess is to replace the cables.

The tickets should show that there was an exersize error when testing, but this seems to have gone away, the arm is moving seems to swap tapes out, but at the rate of backup, it will not need to swap tapes for 4 days.

Anything assistance is greatly appreciated.

Yours,
William

[William Grace_1]

Here are the long from LTT

[William Grace_1]

With the DAT files this time

[David Ruska]

Thanks William, we will investigate this for you.

[David Ruska]

William,

First, regarding performance
----------------------------------------

From sticket2.ltt on Monday May 16 2005 - 10:17:21:
--
In the device analysis section it says:
The current SCSI configuration is likely to be limiting the performance of the drive.
Please check that your HBA is the correct type for the drive and that the cabling is good.
The SCSI configuration referenced is the one for which device analysis was run and/or the support ticket was pulled.
If this is not via your backup server then you may not have an issue.
Current SCSI configuration: narrow Recommended: wide
Current SCSI transfer rate limited to: 40 MB/sec. Recommended: 80 MB/sec. or better
The drives hardware compression algorithm is currently switched off.
This was probably set by the backup software.
The drives setup has not been altered.
----

Similar messages are shown in the eventlog.ltt file on June 02 2005 - 16:46:46:
----
Current SCSI configuration: narrow Recommended: wide
Current SCSI synchronization period: 200 nsec. Recommended: 25 nsec.
Current SCSI transfer rate limited to: 5 MB/sec. Recommended: 80 MB/sec. or better
----

How is the library & drive cabled? Does it use wide (68 pin) cables?

Is the HBA configured to run narrow, asynchronous, or at a limited transfer rate?

Some HBAs will reduce transfer rate if they fail a loopback test at power-up (usually indicates a severe cabling/termination issue).

Also in the eventlog.ltt file, there are a number of these messages when communicating with the LTO drive:
I/O Exception (Error: 2408 - I/O: An unexpected Win32 error has been reported)

This message is often seen when another program or service is accessing the device, or has it opened.

Before running LTT, was your backup app and services disabled (i.e. stopped)?

Is Windows RSM (Removeable Storage Manager) disabled? Unless your backup application needs this to manage the library, it should be disabled.

What backup app are you using?

Second, regarding the library exerciser test failure
----------------------------------------------------------------------------

The results log (from 6/2/2005) shows that the test failed as follows:

The test function encountered an unrecoverable error condition, the test failed.
- Application Exception 2 (invalid parameter passed to function)

This failure may be due to a conflict with another application/service also accessing the library, or it may be an LTT test bug. Let's determine if there's a conflict first.

The newest support ticket was from May 25, 2005. It did not show any library errors since updating to 5.07 firmware. There were one media load/eject failure a few hundred hours back with 4.30 firmware.

The drive logs show that it had been having no problems writing - one of the most recently loaded tapes had 196GB written with no retries and no errors, so there is no indication of any problem there that would cause slow backup performance.

*** First step would be to understand the SCSI transfer rate issues, and whether there are any application/services conflicts.

[William Grace_1]

The Cabling is good, unchanged from the time when I was getting 500MB throughput. I have replaced the Drive, its still slow, I have asked HP to send the Cables so I can replace the Cables (I will wait the Standard Weeks for a response).

I have checked the SCSI settings, all remain as before. Since I can not wait to get this resolved I have to wait the 10 days it will take to do a Backup, and then I will be able to work on it again.