V13.1.1 Missing directories and failing HA-Sync

[jbender81]

Hi,
After Reinstalling HA-Nodes with 13.1, which have originally being installed with 13.0, due to having to migrate them from VMware to Proxmox, I noticed, that the Job "HA Cluster Synchronization" failed on every run.
On downloading a Log-Bundle I had a look into the Logfile "Job.HAMaintenance.log. There I noticed this errors:

Code: Select all

[31.08.2026 13:25:07.437]    <07>    Info (3)    [HA binary executor] Executing command '/usr/bin/chown -R :veeam-grp-log /var/log/VeeamBackup/HighAvailability/consul'
[31.08.2026 13:25:07.459]    <07>    Info (3)    Command execution failed. Command '/usr/bin/chown', Exit Code: 1, Error: '/usr/bin/chown: cannot access '/var/log/VeeamBackup/HighAvailability/consul': No such file or directory

Fortunately we still had a staging system, where this directory existed, hence I SSH'd into the Nodes and created it with "consul" as owner and "veeam-grp-log" as group.
Afterwards I noticed, that also the directory '/var/log/VeeamBackup/HighAvailability/patroni' was nonexistant, which also got created, but with owner "postgres" and group "veeam-grp-log".

Afterwards the Sync-Jobs works.

This has the unfortunate side effect, that the relevant services can not write to their respective log files, as the respective directory was nonexistant during startup.

A good solution for this might be to add some ExecStartPre for the respective services, to create those directories and chown/chmod them as required. like so:

Code: Select all

ExecStartPre=-/usr/bin/mkdir /var/log/VeeamBackup/HighAvailability/consul
ExecStartPre=-/usr/bin/chown consul:veeam-grp-log /var/log/VeeamBackup/HighAvailability/consul
ExecStartpre=-/usr/bin/chmod 2770 /var/log/VeeamBackup/HighAvailability/consul

Cheers
Jörg

[david.domask]

Hi Jörg,

The behavior definitely sounds incorrect, so please open a Support Case and let Support review the situation. While I'm glad you were able to get the sync going again, it's best to check this with Support as there is never a reason you're required to do such edits manually, and it'd be best to check the situation and ensure that there aren't any other underlying issues that may impact HA operations down the line.

Please share your case number once created, thanks!