Resolved Plesk backups failing with error

[AmaZili Communication]

Hi,

In order to share information, I want to let the community knows that everything is working back again.
It seems the error disapared on the day I posted my message when Obsidian was updated to plesk 18.0.28 Update #2.
We do not know if it is a coincidence...

Hope this can help.

Philippe

 

[Pieter van Kampen]

I have this same issue still after upgrading to 18.0.28, using FTP as backup medium. What I have not read before is that the error is always about the same file:

pmm-ras failed (Error code = 1): Repository error: Failed to read backup backup_2006101819_2006141819.tar: Import error: Unable to find archive metadata. The archive is not valid Plesk backup or has been created in an unsupported Plesk version

I tried to delete the backups before and after this particular date from the Plesk UI and got this error:

Unable to get ftp dumps list: pmm-ras failed (Error code = 1): Repository error: Failed to read backup backup_2006101819_2006141819.tar: Import error: Unable to find archive metadata. The archive is not valid Plesk backup or has been created in an unsupported Plesk version

 

[Bitpalast]

Please log in to your FTP backup space and verify that the file actually exists, because the same error message is displayed if the backup file is missing.

 

[itoop]

It seems that the problem is solved now . I don't get error messages anymore and the backup is valid.

 

[Tag]

I think this might be an issue with too many concurrent connections to the FTP storage space. Could you please try this solution:

Add the option ftpForbidReuseConnection into [pmm] section in /usr/local/psa/admin/conf/panel.ini:

[pmm]
ftpForbidReuseConnection = 1

Then run
# service psa restart
on the console, then try the backup again. Is the same issue still there?

This solved the problem for me too, thank you!

I have an off-site backup (actually hosted at home on a Raspberry Pi using passive FTPES) which has been running for years.

I have three Plesk instances. My CentOS (latest Plesk 30 license) one is fine but had the above issue with both Ubuntu 20.04 (latest Plesk, 10 licence) servers.

Perhaps it's specific to the Plesk edition or, more likely, OS - specifically Ubuntu?

I also initially had a different error (similar to this bug which also led me to this solution which is the same fix as suggested here).

 

[Tag]

Nope, I'm still getting errors for the two Ubuntu boxes for each daily backup, this time something slightly different as well now:

The backup Feb 18, 2021 02:07 AM was created, although some issues occurred and the backup might not be entirely valid.

Error:
Export error: The size of the volume backup_2102121207.tar 72775168 does not match the expected one 73332957. The remote backup may not be restored.

Error:
Unable to validate the remote backup. It may not be restored. Error: Failed to exec pmm-ras: Exit code: 119: Import error: Unable to find archive metadata. The archive is not valid Plesk backup or has been created in an unsupported Plesk version

 

[epertinez]

Just a quick line...
In my case the error appeared when I was forced to hard reset a backup process that I started by mistake (I only wanted to backup plesk config database).
I understood the problem was with those backups being only half cooked so I deleted them from the server.
The problem remained even when those backup files were not in the server anymore.
Today I deleted those files from Dropbox. I hope that finally solves the problem.
If that is true, it seems that backup process not only preprocess old backup files (for incremental reasons?) but also preprocess old backup files saved in ftp/dropbox server.