When a set of clients is configured at once to a new backup-server, sometimes a race-condition is met during the multiple delegate_to for creating the backup-user on the target backup-machine.
During the useradd, the previous useradd is still running or hasn't finished completely and the new 'useradd' seems to run into trouble.
Fixing by using 'run_once' in this task.
Before:
TASK [borgbackup : server | create user] ***************************************
The end result is the same, the user is created. I only started noticing this problem when using Ubuntu Bionic as a backupserver, it seems to be fine when using Debian Jessie..
When a set of clients is configured at once to a new backup-server, sometimes a race-condition is met during the multiple delegate_to for creating the backup-user on the target backup-machine.
During the useradd, the previous useradd is still running or hasn't finished completely and the new 'useradd' seems to run into trouble.
Fixing by using 'run_once' in this task.
Before:
After (with run_once):
The end result is the same, the user is created. I only started noticing this problem when using Ubuntu Bionic as a backupserver, it seems to be fine when using Debian Jessie..