File not found

[Schmoo]

Technically a technical issue but it's just a simple question so I don't see the need to open a support case as warned above.

Backing up a simple directory to tape. Job ends with the warning "Volume D: have x problems. Details: File not found"

I understand from previous threads this is normally caused by Veeam doing file enumeration at the beginning of the job but the files weren't there when it got to the point of backing them up. I get that.

Simple question - where can I find a list of these files it didn't backup? That would increase my customer and my comfort level if I can just verify that the errors were generated because the files have been deleted/moved/whatever since the initial enumeration.

I've generated the log files and as we know, they are numerous and verbose. I just need to know which log file to look at and what error should I be searching for.

Thanks

[PTide]

Hi,

I just need to know which log file to look at and what error should I be searching for.

Try to search for "not found" string inside C:\ProgramData\Veeam\Backup\JobName\Job.JobName.TapeFilesBackup.log file on VBR server.

Thanks

[Schmoo]

Thanks for the info. I've been looking in that file and can't find any "not found" entries. I've tried searching for other terms as well "error" and stuff like that.

Seems like if the program knows I have "14 problems" it would at least be able to tell me what they are!

[B.F.]

I've experienced this as well. Any updates on how to identify the missing files.

Also, if Veeam is invoking Shadow Copy on the host which is being backed up, I don't understand how it can then miss any files. Is Veeam actually backing up the files based on the Shadow Copy?

Thanks!

[Dima P.]

Seems like if the program knows I have "14 problems" it would at least be able to tell me what they are!

Imagine you have 1000+ missing file names pushed to job session

Is Veeam actually backing up the files based on the Shadow Copy? Is Veeam actually backing up the files based on the Shadow Copy?

Yes, if you use VSS option in file to tape job. This error reflects that file added to the source for backup job was deleted (thus cannot be backed up)

[B.F.]

Seems like if the program knows I have "14 problems" it would at least be able to tell me what they are!

Imagine you have 1000+ missing file names pushed to job session

Imagine taking a vehicle to the shop to have work done. Upon picking up the vehicle, the repair receipt will then say "Found 14 problems. Cost is X"

Every backup I've done so far has this error / warning. If it doesn't tell us what files are omitted, why bother with the warning? What recourse do we have to try and correct this so that the reports aren't constantly yellow with warnings? Maybe instead of an error / warning, this could be informational stating "X files were omitted due to deletion" or something in that nature?

I'm probably sounding like a broken record from other posts but our production Backup Exec backing up the same data to tape does not throw any kind of error like this.

Yes, if you use VSS option in file to tape job. This error reflects that file added to the source for backup job was deleted (thus cannot be backed up)

Help me understand Veeam's process. In our job setup, I have it backup a server's E:\<folder> with a filemask of *.*. Does it then generate a list of specific files and folders from the filemask before starting the backup? If not, how else would it even know that a file is no longer there unless it has a reference point? If an unnamed file is part of the error / warning, will Veeam from that point on, always count that unnamed file in subsequent backups or just the current one?

Thanks much

[B.F.]

Update: Today's backup did not have any errors / warnings! So that answers the question if the errors are cumulative. It seemed like each day was adding more errors but this time it was green.

I just don't understand where it's getting the error / warning from. When VSS is being used and the backup is based off of the VSS snapshot, this warning should never come up.

So I did a test. I ran the tape backup and watched very closely the status of the job as it was happening while also monitoring VSS on the system being backed up. I think I see what's going on.

1. Veeam starts off building the file tree. This can take some time.

1a. I noticed that VSS was NOT invoked while the file tree was being built. This tells me it is generating the tree not based off of the VSS snapshot but from the actual file / folder structure of the system being backed up.

2. When the file tree is built in Veeam, then shortly after I see VSS invoked and the snapshot is taken.

3. Veeam then backs up files to tape.

If a file is deleted while the file tree is generated and after Veeam has already identified that file, it will not be in the VSS snapshot. Therefore, I can see this error being thrown.

If my assessment is true, would it be possible to rearrange the sequence a bit? First create the VSS snapshot, then build the file tree based off of the snapshot just taken?

Thanks!

[simcomit]

I 100% agree with this post, and have just run into the same issue trying to copy some files to tape.

1) As for the warning being virtually useless, I don't understand how you can claim this is a reliable backup software if it is hiding the details of the most important job it performs - making a copy of data from one media to another. How is this thing useful for anything if it keeps reporting files weren't transferred but doesn't tell you which ones? I want them ALL copied, and any that aren't I want to know exactly which ones so I can fix the process (no data should be able to be created and deleted without being backed up first). This should be a setting so those that don't want a list of 100's of files in the log can just get the generic "warning" as it does now, but for those of us that actually care if our data was backed up we should have the option of turning on the verbose info so the potential problem can be addressed (or a way to find said details by clicking on the warning).

2) If B.F.'s analysis of the VSS process is correct how is this a valid design? As I said in item #1 above, the whole purpose of this software is to reliably transfer specified data from one place to another. That means the second the job is started the source data needs to be "locked" in place until it can be copied. Why on earth would you take a file list first, then lock it, allowing this exact situation to occur? If you just fix this, the need for a full list of "deleted" files in item #1 would be irrelevant as it would never happen (because with a VSS lock even if the file is deleted after the backup started and built the file list it wouldn't actually be deleted until after the backup job released the VSS lock, hence the file would never miss the backup transfer and the warning wouldn't happen).

[busitech]

Sometimes, it doesn't even matter what files are missing, because they were deleted during the backup. This can happen when backing up files from a live Linux machine that has no concept of VSS - which can be especially problematic for Veeam when the backup is long. These jobs always end in Warning when in fact they were successful.

This specific case is one of a Linux machine sharing out a volume over AFP (netatalk) to Apple Macintiosh machines. There are changes to the Time Machine archives 24x7x365, as long as the machines are powered on. (The user receives an email, shortly thereafter, it is in the Time Machine backup on the file server. Awesome!)

I do not want to create a third copy of this data in the Veeam deduplicated store, because the backup volume is huge, and full of data that does not deduplicate at all.

The server is a VM, so I'd actually prefer to start a VM snapshot and copy the (full or incremental) backup of the volume straight to tape, but Veeam doesn't offer this option (it requires a disk copy first).

I would like to see an option to ignore missing files.

[B.F.]

Now that v9.5 has been released into the wild, does anyone know if this was addressed?

Thanks

[Dima P.]

Hi guys,

Same behavior remains in v9.5 We have this request on tack and noted this feature request for next versions. Meanwhile, detailed errors are displayed in the job log file – if required you can check the detailed output there.

[hannesbreit]

Hi guys,
are there any news on this case?
from out of nowhere i do experience the same Error now.
But B.F.s szenario does not fit 100% to me.
The file which is not backed up is definitly there before and after update. idk why veeam does not backup this file

[Dima P.]

Hannes,

If you are backing up file from a CIFS share (and backup is performed without VSS) file can be locked by another process. When it happens, we show notification in the job log to notify the user that some files where not processed. The best workaround would be to rerun the file to tape job when such files are not locked by users or applications.

[tomnewman]

I've used both B&R and Backup Exec for a while now and though B&R significantly better in most respects, I have always felt Veeam's tape support has sometimes been a bit half-hearted.

It is ridiculous for backup software to tell you it hasn't backed something up but not tell you what! It is also not satisfactory to say that the files are 'probably' files that have been deleted since the backup was started.

Furthermore, identifying the missing files requires access the programdata directory, which in a secure environment may well not be possible. And why should it be necessary?

This thread is two years old - come on Veeam, sort this out!

[tomnewman]

Can someone from Veeam comment as whether this is ever going to get resolved?

We have a 10TB volume that we backup direct to tape because we don't have NAS space to do an intermediate backup. It takes about 20 hours to back up and there is a good chance that the contents may change during the backup. It isn't critical data but we would rather not lose it if we don't have to so we backup to tape to have an off-site, air gap copy.

It is not unusual to have missed files, we would like to be able to check which files were missed without wading through long log files.

[Dima P.]

Hello Tom and sorry for the delay.

We have a 10TB volume that we backup direct to tape because we don't have NAS space to do an intermediate backup. It takes about 20 hours to back up and there is a good chance that the contents may change during the backup. It isn't critical data but we would rather not lose it if we don't have to so we backup to tape to have an off-site, air gap copy.

Is that a windows file server volume or NAS volume?

It is not unusual to have missed files, we would like to be able to check which files were missed without wading through long log files.

Whenever you get such warring in job stats you can open job session log to review the list of files that were skipped during backup process (you can find extended job logs here C:\ProgramData\Veeam\Backup\).

[tomnewman]

Hi Dima

I must also apologise for the late reply. I have been checking the thread but hadn't noticed that it had gone on to a second page so I missed your reply until now.

This is an agent backup of a Windows volume.

I appreciate the information is in the logs but to put that into context, the logs run to over two million lines spread over about 50 files (presumably because of a log file size limit). Would you fancy finding missed files in that?

Veeam is a great backup program but it is the only backup software I've used (and I've used a few over 30 years) that doesn't list skipped files in the standard log file. I can't believe that it is that programmatically difficult to record the exceptions and add them to the standard log.

[Dima P.]

Hello Tom.

Thanks for getting back to us.

This is an agent backup of a Windows volume.

If that's an agent backup then your backup is taken with VSS snapshot. It means you don't have to deal with the locked files at all and this data must be processed by Veeam agent.

[tomnewman]

Hi Dima

The problem is files (or folders) deleted after the start of the backup (and bearing in mind the backup takes 20+ hours that is pretty common).

I can't just assume that the failed files are because of this, I need a list to check that the results are as expected (because of deletions) rather than another reason.

[Tonst3r]

Necro'ing this thread from 2016 because this is still an issue. We've been using B&R for VM backups and love it...so figured let's check-out the tape option and hopefully ditch Backup Exec. Now I'm staring at a page full of yellow warnings, with no indication on what files were missed, and my first search goes to a thread from 2016 that was never resolved.

That's not super comforting. Is anyone familiar with this, and/or has anyone found a way to resolve it? Pretty much identical concern to the OP.

[Gostev]

If I remember correctly, there should be a link to a log file with all missing files in the backup job's Action Log. The reason this goes into a separate file is that it putting millions of files into a configuration database does not scale (and Veeam does not target SMB customers, like Backup Exec).