Unsolved

This post is more than 5 years old

4 Posts

18386

May 23rd, 2004 18:00

120T DLT7000-errors on Verify and Restore only

Quick background: Purchased a Dell PowerVault 120T DLT7000 a few months ago from eBay. It worked for a while but the "cleaning tape" indicator started appearing constantly. I replaced the Quantum TH6XF-ES drive with a refurbished drive. The indicator no longer appears, but now I have a new problem which started immediately after replacing the drive.

I’m using Backup Exec 9.1 on Windows 2003 with an Adaptec AHA-3944AUWD SCSI card. I can inventory tapes, catalog tapes, erase tapes, re-label tapes, even back up files-all successfully. However if I submit a job asking for verification or if I simply try to restore a file, I get a job failure notice stating that an I/O error with the device has occurred. The unit itself does not report an error. When a verification starts, the drive will beep three times and make a loud grinding sound as the heads recalibrate. The device displays “Reading� and then after a few minutes, the tape is ejected, but it appears that Backup Exec does not receive the data from the tape read.

I’ve replaced the terminator and both SCSI cables (external plus the one inside the tape unit). I’ve change the SCSI card settings try asynchronous, 10m/s and the default 40m/s settings, all with the same results. I even put the original drive back into the unit and I get the same error message. I’ve used a cleaning tape and I’ve tried several different tapes for the backup job. I've tried both channels on the SCSI card as well. I am able to restore using the original drive connected to an internal channel on the SCSI card, but I get lots of soft read errors (assuming from the bad tape heads).

I’m running the latest firmware on the PC, SCSI card and tape drive. I’m unable to upgrade the firmware for the loader from 114 to 116 because the scsiwm.exe application is unable to find the ADIC FastStore changer (although it appears in the device list and within Backup Exec). Bacup Exec has all updates applied.

I’ve already put a lot of money in getting this unit going again (since it is no longer under warranty) and would even be willing to pay $50 to the first person who places the “winning suggestion.� SERIOUSLY!!!!!

I’m thinking it is the circuit board located directly behind the drive, but I’m afraid to replace it without getting other professional advice. I’ve included the Event Viewer error logs below. Thanks in advance for any help!


Application Log Errors:
---------------------------------------------------------------
Event ID: 33152
Adamm Mover Error: Read Failure!
Error = ERROR_IO_DEVICE
Drive = "QUANTUM 1"
    {8E69C15E-8AD8-4D0D-820E-B6D98387F492}
Media = "Test.B"
    {900BC442-816E-4739-A317-0523C7D7282A}
Read Mode: SingleBlock(0), ScsiPass(0)
Write Mode: SingleBlock(1), ScsiPass(1)
---------------------------------------------------------------
Event ID: 33152
Adamm Mover Error: Read Not Ready Failure!
Error = ERROR_DEVICE_NOT_CONNECTED
Drive = "QUANTUM 1"
    {8E69C15E-8AD8-4D0D-820E-B6D98387F492}
Media = "Test.B"
    {900BC442-816E-4739-A317-0523C7D7282A}
Read Mode: SingleBlock(0), ScsiPass(0)
Write Mode: SingleBlock(1), ScsiPass(1)
---------------------------------------------------------------
Event ID: 57665
Storage device "QUANTUM 1" reported an error on a request to read data from media.

Error reported:
The request could not be performed because of an I/O device error.
.

 For more information, click the following link:
http://eventlookup.veritas.com/eventlookup/EventLookup.jhtml
---------------------------------------------------------------
Event ID: 34113
Backup Exec Alert: Job Failed
(Server: "SERVER") (Job: "2") 2 -- The job failed with the following error: The request could not be performed because of an I/O device error.
 
For more information, click the following link:
http://eventlookup.veritas.com/eventlookup/EventLookup.jhtml
 
 
System Log Errors:
---------------------------------------------------------------
Event ID: 11
The driver detected a controller error on \Device\Scsi\aic78xx1.

For more information, see Help and Support Center at http://go.microsoft.com/fwlink/events.asp.
---------------------------------------------------------------
Event ID: 15
The device, \Device\Tape0, is not ready for access yet.

For more information, see Help and Support Center at http://go.microsoft.com/fwlink/events.asp.

May 25th, 2004 19:00

It is hard to say what the state of the autoloader is in since it was picked up from ebay and the drive replaced.  However, since you can execute utility jobs and backups indicates that communications to the unit may be ok.  Look in your system eventn viewer and check if there are any aic78xx error messages.

I suspect that a bus reset is occuring thus interfering with a verify or restore operations.

The errors you posted indicate that the OS is losing communication with the tape device, thus reporting back that the drive is no longer connected. 

When verify or restore operation starts, the tape media must be rewound back to the beginning of the session.  It is possible that the long rewind process may be timning out, thus causing a bus reset .

 

Here are some suggestions:

1.  Set the SCSI adapter 2944 adapter back to defaults (via adapter BIOS)

2.  make sure you installed the Veritas Backup Exec tape class drivers.

3.  Are you using an High-Voltage Differential (HVD) terminator.  (you must use this type, not an LVD terminator).

4.  Make sure you have the latest adaptec device drivers for the adapter.

5.  You didn't mention how long your backups were, but is a quick test.  Quick erase a tape, than submit a very SMALL backup (100 byte file) to the same tape cartridge.  See if the very operation completes.  If so, try to restore the single file to a temp directory.

 

4 Posts

May 26th, 2004 00:00

Thanks for the info Richard.

My original test file size was a little less than a gig; I've retried the test using a 652k file, but the results unfortunately were the same.

I'm currently using an Adaptec 3944, but I've seen that a majority of users are using the 2944. While the 3944 should be compatible (HVD), I've just purchased a 2944, new external SCSI cable and HVD terminator to rule out any problems outside of the autoloader.

Regarding your comment about the aic78xx errors, I am in fact getting an aic78xx error in the Event Log (see following). Can this simply be showing a bus reset or does it probably point out a problem with the SCSI card?

-------------------------------------------------------------------------------------------------------------------------------------------
Event ID: 11
The driver detected a controller error on \Device\Scsi\aic78xx1.
For more information, see Help and Support Center at http://go.microsoft.com/fwlink/events.asp.
-------------------------------------------------------------------------------------------------------------------------------------------

Your suggestions are wonderful. If you can think of anything else that might be causing the problem, I would be grateful beyond your imagination! Keep watching for a new post asking for your mailing address!!!

Larry

May 26th, 2004 01:00

Can you can match any of the aic78xx errors in the system event log with the same timeframe that the verify / restore fails.  If so, you can cut/paste the dump data into this forum message.  Double click the aic78xx entry, select WORDS, than copy the entire listing in the scroll box, than paste in a message.  The dump data has information that can narrow down what the error is.

Based on the Event ID 11, this does not indicate a bus reset, but rather a miniport driver timeout, nonetheless, the info in the Event ID 11 can help narrow done the problem.

Also, can you provide details about the server model and what type of software and hardware you have installed on the system

 

4 Posts

May 26th, 2004 02:00

Richard:

The machine is a standard desktop that I’m using as a server. It is only accessed by me so it works fine for my needs. It’s an HP Evo D530 desktop (2.66ghz) with 2gig of RAM, 40gig hard drive; the only “extra� installations are the Adaptec 3944 card and the external tape drive. It is running Windows Server 2003 and Backup Exec 9.1. I also have Microsoft Office installed along with Eudora, Nero and a few other small programs I’ve used for some time, and still do on other machines. I use the machine to backup other machines on the network, but I’ve only been troubleshooting with files on the local hard drive.

Below is the dump data for my latest aic78xx1 error:

0000: 0010000f 00680001 00000000 c004000b
0010: c1305022 00000000 00000002 00000000
0020: 00000000 00000000 00000000 00000003
0030: 00000000 00000005

I am also getting three additional errors following the aic78xx entry; all three appear to be the same (with a few seconds difference in the timestamp):

Event ID: 15
Type: Error
Source: QntmDLT
Category: None
The device, \Device\Tape0, is not ready for access yet.
0000: 0068000f 00c00001 00000000 c004000f
0010: 00000104 c000009d 00000000 00000000
0020: 00000000 00000000 0002d09f 00000000
0030: ffffffff 00000000 0a000040 00030000
0040: 12062004 10100108 00000000 000002d0
0050: 00000000 897cd200 00000000 897cccd8
0060: 89a2b000 00000000 00000000 00000000
0070: 00000000 00000000 00000000 00000000
0080: 00000000 00000000 00000000 00000000

I’m sure the latter message is a result of the aic78xx error…

One thing that I haven’t mentioned…I have removed both the changer and tape drive entries through Device Manager and had the OS redetect the equipment at next boot, but it didn’t appear to make any difference.

Will this be enough info to conjure up more ideas? Again thanks for the help!

May 26th, 2004 14:00

Looking at the aic78xx1 error dump is indicating a SCSI protocol error being returned. This is not a common error and I suspect that there is either a hardware issue or incompatibility.  This is telling you that proper SCSI communications (protocol) is not correct.

Also, are you using the Veritas tape class drivers?  The qntmDLT message is an indication that the OEM driver is installed.  To make sure you have the Veritas tape drivers installed, run the Backup Exec tapeinst.exe program and be sure to select VERITAS tape drivers.

It is hard to say where the problem resides because it is unknown if the adapter was validated in the server or if the autoloader is faulty.

As a last resort, you can try the following:

1.  If the PC has multiple PCI busses, try installed the SCSI adapter in a diferent slot.

2.  Try to install the adapter and autoloader in another model server.

3.  Check system POST to see if there are any errors reported by the adapter.

4.  Try a different SCSI ID on the autoloader and drive (change via LCD panel-  From the dump data, it appears you are set to ID 3.).

5.  Make sure that the PC, SCXI adapter have the latest BIOS installed and that you have the latested Adaptec driver installed. 

6.  Disable the OS Removable Storage manager (RSM) service.

 

4 Posts

May 28th, 2004 21:00

Hey Richard:

Thanks again for the suggestions.

I moved my existing card to another PCI slot and when I restarted, the OS found new hardware. Of course everything showed up as before since it was the same hardware, but I thought it was weird that it stated it found new hardware. I also changed the SCSI IDs of the drive and loader and upon restarting, the computer "found new hardware" again. Unfortunately the problem is exactly the same.

I replaced the SCSI card with a 2944 today, but the problem still remains.

You were correct with the driver; I was using the OEM and ran the Backup Exec tape drive utility and switched over the Veritas drivers; no resolution.

Do you think this problem could possibly be caused by a corrupted OS? Because the drive was working successfully on this machine before I replaced the tape drive, I assumed the problem was hardware. However the more and more I work on the problem, I'm thinking it might be OS related.

May 31st, 2004 23:00

Under W2K3, if you change the target ID or the slot, Windows will see this as  new device, thus popping up the box stating that new hardware is found.  This is normal for W2K3 and you would need to reload the tape drivers. 

Looking back at this entire forum thread, the only other item in question is the drive.  You mentioned that when you replaced the drive, you started seeiing the verify/restore problems.  Have you verified that you installed the new drive (DLT7000?)  correctly?  I have not taken one apart, but were there any jumpers that needed to be set, internal connections..etc?

As far as the OS being corrupted, I don't think this is the case because the error log you provided indicates a SCSI protocol error.  At most, you may want to uninstall the adaptec driver, reboot and reload the SCSI driver.  (you may also have to re-run the Veritas tape driver installer (tapeinst.exe).

Also, from the OS errors that were in the system event viewer, the "tape device not connected" is telling you that communications to the driver has been lost.  When the failure occurs, the the 120T LCD panel display any errors?  Have you run diagnostics on the autoloader.  (i think there is a small diag program via the LCD, but I will need to check on one to verify this)

 

June 25th, 2004 14:00

Richard,

 

Where can I find drivers for the same model tape drive autoloader for server 2003 for Netbackup 5.0 mp2?

6 Posts

February 17th, 2006 17:00

I am experiencing the same exact behavior with my PV120T.  I've repeated the problem on 2 different servers.  I've tried updating the firmware for the autoloader and drive and all the drivers.  Was any resolution ever found for this problem?

I am using Veritas Backup Exec 8.6 on Windows 2000.  The autoloader now has firmware version D116 and Microsoft Driver 5.0.2195.6717.  The tape drive has firmware version 296D, Veritas Driver version 4.0.1381.1.  The Adaptec Scsi has the Adaptec driver version d2.23.

I have spent an enormous amount of time trying to get this drive to work, any help would be greatly appreciated.

4 Posts

March 9th, 2006 15:00

I'm seeing the same issue and have tried updating the DLT firmware both using SDLT update and using the DFU utility from Quantum, but the drive stays resolutely on V97.

We're running Windows Server 2003, has anybody successfully updated to V109 using either of these on W2k3?

Steve

 

4 Posts

March 30th, 2006 08:00

Guess I'm alone in this!!

Anyone from Dell got any ideas, running the SDLT update identifies the drive as gaving v97 firmware (2561? I think) and then says it will update it to v109, we see teh 5 minutes remaining,4,3 etc then we get this error: "drive image checksum does not match the image file checksum" and the drive remains at v97.

I'm going to try moving the loader to an old windows NT4 box if I can find one just to try the upgrade there.

 

Are there any other tools that can write the .tz_image file to the the drive?  I've tried DFU and also xTalk from quantum but with no success.

 

Steve

 

 

 

1 Message

April 7th, 2006 18:00

FWIW, I also crash on verify. I gave up as I can backup and restore with no trouble. My standard monthly random restore assures  m that it is working.e

October 17th, 2006 21:00

See my post earlier today...
No Events found!

Top