Guten abend, alles!

In getting things set up, I have a long batch file to import a cleaned up configuration into cobbler. Surely, I am not the only person who has done bulk adds/edits. Because everytime we edit a system, we resync and restart named, this really quickly results in systemctl flagging named-chroot as restarting too rapidly, and then things quickly devolve into trying to restart named instead. This was not an issue when the resync/restart had to be explicitly triggered by a cobbler sync, but now, at least on RL 9, it results in changes to the /var/named tree which require manual fixing.

On workaround would be to add sleeps between each cobbler command in the shell script, but this is sub-optimal at best. Failing that, there are two possible fixes beyond the non-starter of going back to requiring a sync being manually requested. They are:

1. Recover the ancient functionally of disabling the sync/restart by specifying a flag on the commands. IIRC, this flag was --batch.
2. Rather than detecting named vs. named-chroot dynamically, make it a config option in say settings.yaml.

BTW... the threshold is really low to have systemctl think that the service is being restarted too quickly... like only 3-5 times in less than a second, which is easily reachable at least early on in a bulk script.

Thoughts?