I’m wondering how containers will be killed when an update is being installed or when the shutdown / reboot endpoint is called via the supervisor API. Is there a graceful shutdown period with a SIGTERM signal, or is the container just being killed instantly?
I searched in the docs and forum, but couldn’t find an answer.
Thanks for the great question! When the reboot/shutdown endpoint is called, the supervisor will make a call to stop the engine which as outlined by the docker documentation, will first send a SIGTERM and then a SIGKILL to the main process inside the container.
When using Application Update Locks, the reboot endpoint request will fail as the application takes priority.