every FullBackup same error: "0xe00084ed - A hardware error occurred"

[Markus_Moench]

Hi there,
I am a little confused about the behaviour of my system here. Every weekday I run a Incremental Job but each Saturday a full backup. The amount of Data for full backup fills 3 Tapes (130GB). Incremental runs fine each time. Full Backup only about 100GB (10hours) an then getting "0xe00084ed - A hardware error occurred". That happens when the 2nd Tape is full and it should take the 3rd Tape in and should continue backup. I´m confused and at loss with that.

I had already read "http://support.veritas.com/docs/255501" and tested what recommended there. But with no success. Anyone similar experience?
We have Win2003 Server, Backup exe 10.0 rev 5520, Overland Robotic Libray LXL1B10 firmware 5.17, SCSI ID2 with DLT1, SCSI ID3 and 10x 40GB Tapes in Magazine on internal SCSI-Controller of HP Proliant ML350 (which worked well under Netware). All Software at latest Patchlevel.
Any ideas will be highly appreciated.What I don´t understand is: when the system is able to backup 2 Tapes (arount 100GB) why that error when accessing the 3rd Tape? Anyone a clue?

[Sharvari_Deshmu]

Hello,

Are there any event id 9 or 11 related to the device?
Try inserting a different tape when the job asks for third tape and then verify the result.



Thanks,

NOTE : If we do not receive your reply within two business days, this post would be marked assumed answered and would be moved to answered questions pool.

[Markus_Moench]

Thanks for your reply.

No, I dont have any event ids 9 or 11 from backupexec in the list. Instead I have 33152 and 57662 which listed below.
ID 33152:
Adamm Mover Error: Write Failure!
Error = ERROR_INVALID_FUNCTION
Drive = "QUANTUM 1"
{0E474334-1E16-43DB-B677-3DF3D48EAFEA}
Media = "1-3-2"
{94B2D898-CDB4-436E-9A2F-D695C3546C68}
Read Mode: SingleBlock(0), ScsiPass(1)
Write Mode: SingleBlock(0), ScsiPass(1)
ID 57662:
Storage device "QUANTUM 1" reported an error on a request to write data to media.
Error reported:
Die Anforderung konnte wegen eines E/A-Ger?tefehlers nicht ausgef?hrt werden. (means: no request possible cause of a I/O-Device error)

As figured out in my 1st post we have a Autoloader (Overland Robotic Libray LXL1B10) and so this happened with different Tapes, no matter which Tape. To be sure I got 20 new Tapes 1 Month ago - same error.

Last night I tested what recommended here: http://seer.support.veritas.com/docs/191087.htm -> figure3. But with same error.

I also tested to change the Device properties, like "Read / Write single block mode", "Read / Write SCSI pass-trough mode" switched on and off - but wit no luck.

The Library and built-in-Drive are connected to the built-in SCSI Controller of the Proliant ML350. This server has 2 built-in Controllers, and I tested both - with the same error. The Drivearray is running on a different SATA-Controller.

Any ideas?

[Markus_Moench]

I am not sure why the program is backing up several hours with no problem (about 100GB) and then suddenly getting "0xe00084ed - A hardware error occurred"? Is the program "surprised" about the ending of a Tape?
Also last night the incremental job was running, backed up about 2,5GB with success - no error occurred.
Any ideas?

[Markus_Moench]

I have found this post: http://forums.veritas.com/discussions/thread.jspa?messageID=4407991??

Sandy Williams was using similar SCSI Controller (LSI Logic Ultra320 (LSIU320)) as built-into my Proliant ML350:
LSI-Adapter, Ultra320 SCSI 2000 Series, (w/1020/1030)(StorPort)
I have Driver-Version 1.20.18.0 of 13.06.2005 installed

And I got same errors in eventlist...

Is there any common to that 2 cases?

[Deepali_Badave]

Hello,


Please perform the following steps:

1.From Backup Exec gui disable the device and delete it.

2.Recycle the Backup Exec services.

You can refer the following technote:

http://seer.support.veritas.com/docs/268626.htm

NOTE : If we do not receive your reply within two business days, this post would be marked assumed answered and would be moved to answered questions pool.

[Markus_Moench]

Thanks for your recommendation.

I had done what described in http://seer.support.veritas.com/docs/268626.htm.
Doing backup again at the moment, will see in approx 5 hours if it fix this issue.

[Markus_Moench]

I had done 2 Backups in the meantime

==> after 10hours19mins and 104 GByte Backup (2Tapes used) <==
Job ended: Freitag, 3. Februar 2006 at 23:25:40
Completed status: Failed
Final error: 0xe00084ed - A hardware error occurred.
Final error category: Backup Device Errors
For additional information regarding this error refer to link V-79-57344-34029
Backup- G: dataStorage device "QUANTUM 1" reported an error on a request to write data to media.
Error reported:
Die Anforderung konnte wegen eines E/A-Ger?tefehlers nicht ausgefuhrt werden.
V-79-57344-34029 - A hardware error occurred.

==> after 5hours17mins and 59 GByte Backup (1Tape used) <==
Job ended: Sonntag, 5. Februar 2006 at 04:17:19
Completed status: Failed
Final error: 0xe00084ed - A hardware error occurred.
Final error category: Backup Device Errors
For additional information regarding this error refer to link V-79-57344-34029
Backup- G: dataStorage device "QUANTUM 1" reported an error on a request to write data to media.
Error reported:
Die Anforderung konnte wegen eines E/A-Ger?tefehlers nicht ausgef?hrt werden.
V-79-57344-34029 - A hardware error occurred.

Before that I had cleaned the Drive and made sure that enough overwritable Tapes were in the Magazine.

Can you get in contact with Sandy Williams ( http://forums.veritas.com/discussions/thread.jspa?messageID=4407991??) and ask him what helped in his case?

I dont have a clue why this happens. Should I remove BackupExec and reinstall again? Can I get a newer Version of BackupExec from you?
Any other ideas?

[Markus_Moench]

Hi there.

Yesterday I had updated the Bios of the ML350 which includes a update of the integrated LSI-SCSI Contoller. I also had downgraded the LSI Drivers in Win2003 to the last version provided by HP (before I had used a version from LSI).

Running fullbackup last night which stops after 5hours15mins, backed up 59GByte to one Tape. Message: "Final error: 0xe00084ed - A hardware error occurred"

I�m stuck. Can someone give me a idea? Do the guys of Veritas read this? Do they have any recommendations? Is this error known to them? I feel not very helped by the staff of Veritas. :(

[Peter_van_Oord]

Hi,

I have the same issues with a MSL6030 Ultrium 3 drive on a HP DL380g4 and a LSI U320 scsi adapter(with the latest drivers).
All backups that go over the network are no problem, when i perform a local backup of that server it faills after a 70GB with the same error as you describe.

I also have done these "http://support.veritas.com/docs/255501" but no luck.

Anyone that have a possible solution ?

[Markus_Moench]

Hi Peter,

what I have here is:
HP ML350G4p with integrated LSI-SCSI 320, running Win2003 Server, driver ver. 1.10.5.2 from HP. Before I had ver. 1.20.18.0 from LSI.
Bios of machine is now 5.05.19, ROM D19.

Is your Streamer built in or external? How do you manage the termination? What throughput rate did your system reach? Mine have about 180MB/min, all data backed up from internal hdd-raid which runs on SATA controller.

regards
Markus

[Peter_van_Oord]

Hi Markus,

My drive is a external, it is a HP MSL6030 with a Ultrium 3 drive.
The throughput i have is around 4300MB/Min
termination is done on the library card.

But i think it is a hardware related issue at my site now. Because the HP LTT tools saying the the cabling is not right(and i am sure that all components are compatible) so i am going to call HP now.

Peter

[Markus_Moench]

Peter,

did you find out some news?

Did the staff of Vertias have any suggestions?

regards
Markus

[Bert_ter_Beest]

I too have the same "hardware error occurred" error.

I've done the following to try and resolve this with no effect whatsoever:

1 changed tapes
2 cleaned tapedrive
3 followed every document that was in the knowledgebase about this problem
4 changed the tapedrive to a new one (HP DAT72)
5 reinstalled veritas and updated it to the highest level.
6 altered the backup job so it only backups 1gb of data

Nothing seems to help, anyone got another idea I could try?

[Markus_Moench]

Bert, Peter, Staff of Veritas,

I am on reinstalling BackupExec right now. Will let you know if something changed then...

I am wondering about no one of Veritas has an idea for us 3 guys having the same issue (Bert, Peter and I)

regards
Markus

[Markus_Moench]

Had reinstalled Backupexec (this time in german language) applied all patches and latest drivers. Done fullbackup last night. It failed after 58GB, 1st Tape with same error.
I�m fiddling around now for several weeks with no success, anyone a hint or idea what we can try next?

Markus

[Bert_ter_Beest]

Since several people have this problem and I am not going to reinstall my server OS it would be nice if the Veritas support staff would come up with an awnser to this problem....

[Markus_Moench]

Bert,

Which type of Server and Controller do you have in use? Same as mine?
What can we do, if they will not get in touch with us?

Markus

[Bert_ter_Beest]

I use a self build server with a Promise Ultra 320-0 Raid adapter and a HP Dat72 tapedrive...

The thing worked like a charm before, but it suddenly stopped working giving the before mentioned error.

I tested the whole thing with a different tapedrive and several clean installations of veritas including updates and driver patches...

The only thing I can try next is changing the SCSI controller, but I fear it won't help...

[Hans_D]

Everybody,

my server is encoutering the same problem. Since a few weeks my job stops at the end of the data backup and gives the known error. I already reinstalled my tape drive, changed drivers, changed tapes, ... Sometimes a reboot fixes the problem for the first backup, but afterwards the error pops up again after 10Gb. I will send the specs of my own-build server with Server 2003 later, but for now I can tell you that it is a LSI SCSI card and a Sony AIT drive.

After several researches and tests, could it not be due to an update of the Veritas-software, or even a Windows update ?
Just as you guys, hoping for a reaction of Veritas...