Veritas spits out tape before completion 6

[PalmStrike]

I have gone over this forum, and read so many threads with this same problem as i am having, i.e., one minute you are backing up fine, (in my case because somebody else configureded Back up exec) the next minute, it no longer wants to play ball.

Has anybody found out what it is yet that suddenly makes backup exec dicide it no longer wants to fit 12GB onto a 24GB DAT?

If it helps, all I did was be ill one day, so i failed to change a tape. In the past our old IT bloke would reset something and that would be fine.

 

[SQLBill]

Check to see what the tape is set for in regards to OVERWRITE and APPEND. If you are not allowing OVERWRITE, you can only APPEND so many times before you need a new tape. You may have gotten to that point and since you were out and didn't change the tape, it couldn't append so it failed.

-SQLBill

 

[PalmStrike]

I have been through pretty much every option on the whole peice of software, I have these settings at the moment:

When Job Begins - Overwrite media.

Backup Method for files - NORMAL -Back Up Files - Reset Archive Bit

Verify after Backup completes - Checked
Retension Media before Backup - Checked

Never back up open files
Backup files and directories following junction points
Backup Method - FULL - Database & Logs (Flush Committed Logs

In the Options:
Media Overwrite Protection level - None

Media overwrite options - Overwrite scratch media

And in the Drive properties, i have the default settings.

I've tried pretty much every other configuration over the last four months and even veritas can't help me, yet it can't be difficult as it used to take our old IT bloke less than 30 seconds to reset it, when this happened before.

Any ideas there?

I have just spotted something that I haven't tried:
Reset Cleaning Statistics.

A box pops up saying irreversable, though there are a few soft errors in there, and I did read a thread from someone about soft errors fast forwarding the DAT.

As theser DATs are only 7 months old is it likely that the drive just needs a clean and then a reset???

 

[2ffat]

I clean my DAT drives at least once a month. This seems to fix a plethora of problems. Another thing to try is to erase the tapes before using them. If a quick erase doesn't work try a full erase. James P. Cottingham

When a man sits with a pretty girl for an hour, it seems like a minute. But let him sit on a hot stove for a minute and it's longer than any hour. That's relativity.
[tab][tab]Albert Einstein explaining his Theory of Relativity to a group of journalists.

 

[PalmStrike]

Nope, didn't work, still the same messge, same error, same s**t, different day, different tape.

I am assuming that if veritas can't solve these problems is it correct to asume that Veritas backup simply does not work and is a waste of money

 

[jdeane]

check the job log and search for ^ and post the few lines above it.

Are you compressing the data, if the tape is a 12/24 then the tape will only back up 24 gig if the data is compressed.

are you backing up PST files? as these won't compress, before moving to DLT I could only backup 14Gb onto a 24Gb tape becuase of the PST files

 

[PalmStrike]

I can't find a ^ in the event log, either that or it is in the 202 pages of it, is that where I should be looking?

 

[jdeane]

load up the Veritas software and look in the Activity Monitor page, double click the job that failed and edit the Job log and do the search there.

you will see something like this

======================================================================
Job server: CLKFPS01
Job name: CLKFPS01
Job started: 07 February 2003 at 08:07:54
Job type: Backup
Job Log: BEX34.txt
======================================================================
Error - Mount failed.

^ ^ ^ ^ ^

 

[PalmStrike]

Here is the job log from last night's backup. I know it failed, because if I am to backup anything tonight, I had to cancel the job.

Does this tell you any of what you need to know?


======================================================================
Job server: EMAIL_SERVER
Job name: Backup 0008
Job started: 06 February 2003 at 23:30:04
Job type: Backup
Log file: BEX02.txt
======================================================================

Drive and media information from media mount:
Changer Name:
Drive Name: HP 1
Media Slot: 0
Media Cartridge Label: 4MM000005
Targeted Media Set Name: Media Set 1

======================================================================
Job Operation - Backup
Media operation - overwrite.
Hardware compression enabled.
======================================================================

Media Name: "Media created 06/02/03 23:30:04 "
Backup of "\\EMAIL_SERVER\Microsoft Exchange Directory "
Backup set #1 on storage media #1
Backup set description: "Backup 0008"
Backup Type: FULL - Database & Logs (flush committed logs)
Backup started on 06/02/03 at 23:30:53 .

Backup completed on 06/02/03 at 23:31:00 .
Backed up 1 Exchange Server store(s)
Processed 8,405,280 bytes in 7 seconds.
Throughput rate: 68.7 MB/min
----------------------------------------------------------------------

Media Name: "Media created 06/02/03 23:30:04 "
Backup of "\\EMAIL_SERVER\Microsoft Exchange Information Store "
Backup set #2 on storage media #1
Backup set description: "Backup 0008"
Backup Type: FULL - Database & Logs (flush committed logs)
Backup started on 06/02/03 at 23:31:01 .

Backup completed on 07/02/03 at 01:46:29 .
Backed up 1 Exchange Server store(s)
Processed 7,972,356,572 bytes in 2 hours, 15 minutes, and 28 seconds.
Throughput rate: 56.1 MB/min
----------------------------------------------------------------------

Media Name: "Media created 06/02/03 23:30:04 "
Backup of "\\EMAIL_SERVER\Microsoft Exchange Mailboxes "
Backup set #3 on storage media #1
Backup set description: "Backup 0008"
Backup Type: NORMAL - Back Up Files - Reset Archive Bit
Backup started on 07/02/03 at 01:46:30 .

Media mount failed.
User aborted a Physical Volume Library operation.


Backup completed on 07/02/03 at 09:09:43 .
Backed up 25154 mail message(s) in 744 folder(s) in 30 mailbox(es)
Processed 3,985,177,299 bytes in 1 hour, 10 minutes, and 24 seconds.
Throughput rate: 54.0 MB/min
----------------------------------------------------------------------

======================================================================
Job ended: 07 February 2003 at 09:09:43
Job completion status: Aborted by EMAIL_SERVER\Administrator
======================================================================

 

[2ffat]

Backup started on 07/02/03 at 01:46:30 .

Media mount failed.
User aborted a Physical Volume Library operation.

Backup completed on 07/02/03 at 09:09:43 .

Did you abort this job manually or is this coming from the system?
James P. Cottingham

When a man sits with a pretty girl for an hour, it seems like a minute. But let him sit on a hot stove for a minute and it's longer than any hour. That's relativity.
[tab][tab]Albert Einstein explaining his Theory of Relativity to a group of journalists.

 

[Qdog]

PalmStrike:

Sorry for your problems.
It appears that whoever 'Administrator' is cancelled the job. (I would check your App/System logs in Event Viewer to see what happened here. Did anyone logon to server? Maybe there is other info here that may help, in the logs.)

Also looks like this bombed out during your Exchange mailbox backup. Look into conflicts between Antivirus on Exchange server and BE backing up mail. This is always a problem.

-Check for device errors:
Windows Event Viewer/System log...look for event id 7, 9, 11, 15 with a source of the SCSI controller or the tape drive. 9 and 11 w/source of SCSI controller indicates a timeout problem. Event id 7 is generally bad block when source is tape drive. 15 can also indicate a timeout problem.
The above event ids are errors reported to the Windows OS by the SCSI controller drivers or tape device drivers and are good indicators of hardware problems.

-Tape drive should be on the lowest SCSI target ID while avoiding 0 and 1. Next, make sure the tape drive is NOT plugged into a RAID Scsi controller. (Not supported config)

-***Check Devices tab in BE, right-click tape drive icon, choose Properties|Statistics tab. Do you have a lot of errors here? THIS IS IMPORTANT BECAUSE YOU MENTION YOU CANNOT GET 12GB ON A 12/24 TAPE. **Every soft-write error that occurs on a tape drive during a backup decreases the amount of tape capacity. Explanation: when a soft-write error occurs, the tape drive is actually fast-forwarding past what it thinks is a bad spot on the tape. If enough of these happen during a backup, I have see situations where the user gets LESS data on the tape then the native capacity! This may very well be your problem.

What is going on during the backup of Exchange? Remember, Exchange 5.5 does maintenance between hours of 1:00AM and 5:00AM every nite. Try to avoid those hours for your Exchange backups.

Look into network performance, you have some poor throughput. (Although the mailbox backup throughput is actually pretty good considering BE uses MAPI.)

If I were you, I'd run a test backup-to-disk folder. (Assuming you have Backup Exec 8.6 or higher.) This test would bypass your tape device entirely. See how the throughput rates and whether the job completes.

Take care, hope you work it out. My suggestion, call Veritas support if the above doesn't help. Stick with them and you may get your case escalated and finally get the problem resolved.

 

[PalmStrike]

It was me that cancelled the job, as the tape was spat out, but I need to know why it no longer backs up on one tape, it always used to, and this is the problem.

if anything, it is backing up less now than when it used to backup fully.

I can back up either the information store, or the individual mailboxes, but I understand one is no use without the other.

the jobs only started to fail when I was ill one day, and didn't put in a new tape.

In the past the old it man used to clear it and it would be fine, so I know it is not complicated, unfortunately this fellow is no longer around.

any ideas???

 

[SQLBill]

Try this...go to the tape, right click on it and select properties. Go to the statistics tab. Take a look at that information (you can also do the same for the device). Are there a lot of errors? How many bytes are written?

Something else to consider, I noticed you ERASED the tape, but have you tried FORMATting it?

I'm thinking that there might be bad blocks on the tape and that's what's cutting down the amount of data you can store.

-SQLBill

 

[PalmStrike]

It won't let me format the tape, it is darked out?

 

[alwayslearning25]

I though I'd put my 2 cents in here. I also run Backup Exec, and when it works it works well, but those times seem to be far and few between.

I have this same exact problem PalmStrike is having. I have the same problems outlined above ( soft write errors ) and having troubles fitting 15gb of data on a 24gb tape.

We clean the tape drive every day.

I'm going to try erasing the tape before each backup and see if that helps. I also backup PST files which I didnt know that the can't be compressed.

I get errors in the alerts requesting multiple tapes and the backup job fails backing up approx 15GB.

I'll let you know the outcome of formating the tapes. Is there away to not backup a certian file extention?

 

[PalmStrike]

Hey Always,

I have tried the erase tape thing, but unless you have the facility in backup exec, it won't let you format it, it might just be something you download from veritas, but if it involves money, my boss it seems would rather his company burned.

Good luck, I have not yet found the problem, and have had to resort to veritas help desk, but last time I tried that, they just sent me something telling me to up grade. I can't see why this is necessary, Why sell backup software that does not fit your needs in the first place?

Rob

 

[PerryMc]

Just in case you don't know, running a cleaning dat through a dat unit once does not give it a thorough clean. I believe that 3 times is best for a good clean. Once cleaned, do the reset under cleaning statistics and use a new dat for your backup. After the backup, check the error counts under cleaning, if they are high repeat the above and try the backup again with another new dat. Still high, your dat unit is naff and needs replacing.

Smile anyway,
Perry.

 

[PalmStrike]

The DAT's we use here are only 8 months old, are are used only once a week, is that enough to make them redundant?

 

[PerryMc]

Dats vary in quality, 120m dats are particularly poor. Once a week for 8 months is about 33 to 36 times used, so yes that is ample usage to have dats go bad on you. Before switching to dlt, I had 120m dats that only lasted being written over 3 times, before they needed to be replaced. If you go to media, you can get the stats on particular dats, replace the ones with high error counts.

Smile anyway,
Perry.

 

[PalmStrike]

The highest error counts on them so far is 10, is that high, on the DLT (Dave Lee Travis???) tapes we use for our other back up, they back up fine with over 1000 soft write errors. is DAT a pretty useless media then?