File based backups create a lot of warnings

[caras]

I always get warnings...

Code: Select all

 [warn] Failed to backup <file name>

...for a lot of files when running backups.
At each backup other files are criticized, I never get an error free backup.
These messages typically come in "waves", not at accidentally point of time.
Later Veaam agent gives up with:

Code: Select all

"[warn] Too many items could not be synchronized. Stop logs flooding"

I'm very sure the hard disks are not damaged: Outside Veaam agent everything works properly at this computer.

Is such an problem already known?
I suspect Veeam agent tries to open too many files in parallel...

Seen at Ubuntu Focal, Veeam Agent 5.0.0.4318, but also with some years old versions.
Sector based backups are not affected by this problem (until new Linux kernels became incompatible to Veeamsnap).

Case #04884123

[Natalia Lupacheva]

Hello,

It does not seem like a common issue, so please keep working with Support.
Also, may I ask you to check if these are temporary/cache files? Just to make sure you don't lose some important data.

Thank you!

[caras]

It affects accidentally any kind of files: system files, data files, program settings, old and new files,...
The affected files are not open/in use, not locked and don't have special permissions.
The files exist before and after backup and are modified during backup.

[caras]

It occurs with magnetic disks and with SSD's (as sources).

Correction to my last post: The files exist before and after backup and are NOT modified during backup.

[SITSAGCH]

Hi, did you solve the issue?
i have the same on a Linux Agent Backup.
Pure File Backup. he is not able to backup some DB files.
Too many items xould not be Synchronized. Stop logs Flooging.

[JaySt]

same issue here. latest version of agent on ubuntu 20.04 doing a file backup. I'll get into contact with support as well.

[JaySt]

permission errors were visible in the agent log file. so big clue was to be found there.

[PTide]

Any chance you are using file-level without a snapshot?

Thanks!

[JaySt]

correct. it was without snapshot

[PTide]

The issue can be seen the most often if multiple files change/disappear during the backup session.

Thanks!

[Werzus]

Hello,
I have the same on a Oracle Linux Server 7.9 involving backup of files created on the same days.
Has anyone managed to solve the problem?

[PTide]

Hi

The only way to solve the problem is to use snapshot-based backiup for those files.
The root of the problem is that the message is supposed to be generated only if something prevented the Agent to access those files.
So far, the only two reasons for that are:

- the files had been deleted during the backup sessions before the agent managed to read them
- the underlying filesystem is busy and prevents Agent from reading the file

Thanks!

[Werzus]

Thanks for the feedback!

This happens with files that are created/transferred on the same day, causing the backup failure. When the job is run again... all the files that had the failure are able to be backed up normally, but if there are other new files that were not in the previous alert, these new files receive the failure alert and finish successfully again later.