BleepingComputer.com: Backup Failing - Veritas

Hi everyone. Not sure if I am posting in the right place, sorry if not.

I am having some problems with a backup job on Veritas Backup Exec. One failed backup a week for past 4 weeks.

Failed backups =

20/08/2007 - Error : a00084af - The directory or file was not found, or could not be accessed.
14/08/2007 - Error : a00084af - The directory or file was not found, or could not be accessed.
10/08/2007 - Error : a00084af - The directory or file was not found, or could not be accessed.
07/08/2007 - Error : a00084af - The directory or file was not found, or could not be accessed.
01/08/2007 - Error : 17 - Data error (cyclic redundancy check).
26/07/2007 - Error : a00084af - The directory or file was not found, or could not be accessed.
19/07/2007 - Error : a00084af - The directory or file was not found, or could not be accessed.
13/07/2007 - Error : a00084af - The directory or file was not found, or could not be accessed.
09/07/2007 - Error : 17 - Data error (cyclic redundancy check).

I suspect a bad tape - seems to be almost same day each failed backup. Called the server tech, asked how many tapes they have - they have TWO weeks worth, and have been changing them every morning, they confirmed they are using the right tapes.

The error: "Data error (cyclic redundancy check)" is reported in the Backup Exec job log during a failed backup/verify operation

Veritas documentation on this subject suggest dirty heads or faulty medium or faulty hardware (tape drive). http://seer.entsupport.symantec.com/docs/192216.htm

The fact it is one day out of the week kind of cancels the theories of dirty heads or faulty hardware from my mind, but will try to organise it. Any theories / suggestions on how to figure this out?

Thanks!

This post has been edited by KoanYorel: 20 August 2007 - 09:41 PM
Reason for edit: moved to more appropriate forum

The failures don't appear to be happening on a regular basis - so it's anyone's guess as to what's causing it. Could be almost anything in the chain from the hard drive to the tape. If it was happening every Tuesday, or every 4th day (for example) that'd indicate (to me) a bad tape - but the days between each error aren't regular, so that doesn't seem to be the issue.

I'd suggest checking the hard drive with a diagnostic tool to see if there's indications of any problems with it.
I'd also suggest structuring the tape usage so that you can verify if it's one tape that's causing the problem (and being present when the backup is done so you can immediately take steps to start another backup if it fails).
Cleaning the tape recorder heads is also a good idea.
Dunno any tools to check the tape drive mechanism, but the manufacturer may be able to suggest something.

This post has been edited by usasma: 21 August 2007 - 06:29 AM

Excellent tip, have organised with the client to try the tests in a few hours...

Thanks again for all the advice, will let you know how I go ASAP

Well, as much as I don't want to believe it is a dirty head problem, it would be silly to try any other troubleshooting or diagnostic tests without out first going a pass of cleaning tape. BTW, the past two days have been backing up and verifying without incident. Last failed backup was Monday.

Let's sit back and wait for the next fail.

Cleaning tape passed through unit - still failing backups. Now trying to recreate all the backup sets.

Just a thought - but is it the same person doing the backup each day - or do several people do it?

I believe the same person changes the tapes - I do all the log checks and create the backup sets.

You thinking sabotage (listen up y'all)?

I'm now thinking it is a problem with the backup set - I am going to recreate it and see if I get improvement.

Nope, was thinking it might be a problem with technique (maybe slamming it in too hard, or not pushing hard enough).

When you eliminate all of the possibilities, you're left only with impossibilities - so it's gotta be one of the impossibilities!

With computers there's no such thing as "random" (IMO) - rather there's just things that occur for a reason that we don't know about. Troubleshooting is all about figuring out the reason.

Since most of the documention refers to impaired connections - have you checked the wires that hook the tape unit up to the system? Could they be damaged (in such a way that heat or cold could cause contraction/expansion to mess up the circuit)?

Do you have the time of day for each of the backups? Does this help to correlate the errors to an event (reason)? Are there any other things running at the time of the backup that might interfere with the process (network/internet stuff in particular)?

Have done a pass of the cleaning tape 2 days ago and has been peforming good, verified backups since. When we get through 2 weeks without a fail, I will claim success. If not, I will put other suggestions into play.

Spoke too soon. Backups failing again.

Escalating to the next level of support, I am at a loss here...