Modifying Shutdown Sequence (Stop LXC's prior to Networking)

[_chris]

Does anyone have suggestions for modifying the shutdown process? It would seem that Prox (8.1.3 and others) drop elements of the networking and NFS mounts before stopping the running LXC's. Due to this current order the automated shutdown of the lxc's can take 300 plus seconds to timeout and force stop if an nfs mount point is used in the lxc.

Is there a way to latch onto say mnt-remote.mount to say don't stop until pve-manager (or similar) is fully stopped?

 

[dlasher]

Does anyone have suggestions for modifying the shutdown process? It would seem that Prox (8.1.3 and others) drop elements of the networking and NFS mounts before stopping the running LXC's. Due to this current order the automated shutdown of the lxc's can take 300 plus seconds to timeout and force stop if an nfs mount point is used in the lxc.

Is there a way to latch onto say mnt-remote.mount to say don't stop until pve-manager (or similar) is fully stopped?

Fair question, having the same issue myself.

Using containers that use NFS mounts inside them, and they refuse to stop or shutdown, have to kill the "lxc-start" to get them to die. Welcome any suggestions.

 

[Retro1982]

Anyone some idea? Having the same shutdown problematic.

Nice Regards

 

[dlasher]

Anyone some idea? Having the same shutdown problematic.

Nice Regards

I've had better luck using:

pct shutdown %x --forcestop 1 --timeout 60"

then doing whatever reboot/shutdown I need to do