Scheduled backups incomplete on errors, or at least error mail is incomplete

Virtualmin
1
SYSTEM INFORMATION
OS type and version Ubuntu 20.04 LTS
Webmin version 2.101
Virtualmin version 7.8.2
Related setting setting [Action on error] is ** Continue, but don’t purge old backups**

In a complete virtualmin backup of all sites, when there is not enough disk space to generate the gzipped-file, the backup stops, and the remaining sites are not backuped, despite setting [Action on error] is * Continue, but don’t purge old backups*

Moreover, the error mail of backup states:

Failed backup of Virtualmin on …com to /file on SSH server …com
Backup failed! See the progress output above for the reason why. Total backup time was 30 minutes, 35 seconds.
Virtual servers that failed :
** thebigsite.com**
Sent by Virtualmin at: https://…

…

Creating TAR file of home directory ..
.. TAR failed! cat: write error: No space left on device

gzip: stdout: Broken pipe
/bin/tar: -: Wrote only 2048 of 10240 bytes
/bin/tar: Error is not recoverable: exiting now

    Saving Virtualmin configuration ..
    .. done

    Saving templates and plans ..
    .. done

    Saving email templates ..
    .. done

    Saving custom fields, links, categories and shells ..
    .. done

    Saving custom script installers ..
    .. done

    Saving scheduled backups ..
    .. done

    Saving DKIM settings ..
    .. done

    Saving greylisting settings ..
    .. done

    Save mail rate limiting configuration ..
    .. done

    Saving mail server configuration ..
    .. done

.. done

Thus, the following virtual servers that don’t get backuped are NOT listed in the “failed servers”, giving the impression that they got backuped, but in fact they aren’t.

So, if thebigsite.com is a not important site, and you let it not get backuped (not great, but in stressful times can happen) for days, you end up with no backup at all of all the following sites.

IMHO, at very least, the list of erroneous servers in the mail should be completed by the following servers that didn’t get backuped because of that error, but actually, an error on one site should continue the backup, so that if others fail, they get listed too in the mail, and that others that don’t fail get actually backuped. Specially with the setting [Action on error] is ** Continue, but don’t purge old backups**

2

and

seem to me to be conflicting. If an error occurs during a backup then eventually a problem will exist the backup will grow in size to the point of no memory to hold it.

I think what you are looking for is something like “[Action on error] only do listed domains/listed mail from domains” (which I don’t think is available) however I would start by performing backups by site. at least that way you can identify which site is causing the problem

3

No, this is not conflicting.

“Continue, but don’t purge old backups” is a setting for the remote ssh backup destination.

The “disk full” is on the virtualmin server itself.

And the site with the size too big to backup is known, it is bigsite.com.

This thread/bug is not solved.

4

So if you exclude bigsite.com from backups there is no problem with “disk full”.

Just a guess - this has nothing to do with quotas set on the VS by any chance?

5

Yes (well previously) if I excluded a too big site, the backups worked.

In the last version of Virtualmin (or its default template) with 2 lists instead of a multiselect excluding a site does not work anymore (see my other bug report).

I don’t think it has something to do with quotas. The fact that there is not enough disk space is indeed correct. The error handling and the error message if that happens during a backup are imho incorrect and a bug.

6

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.