Open lbergesio opened 2 months ago
Not sure why the ci failed, any hint with that would be appreciated
Hi @LiilyZhang is anything preventing this to be merge I should address?
we are still investigating how to test this in openshift environment @lbergesio
we are still investigating how to test this in openshift environment @lbergesio
I think kubelet works with the pod the same way than docker with the container: https://kubernetes.io/docs/concepts/workloads/pods/pod-lifecycle/#pod-termination I would say just run the pod, get in, check the process is running. Destroy the pod, this will send sigterm to the pod process?
Hi! any update on this? Thanks
Description
When run in a container anax is run via a shell script that does not handle SIGTERM and hence when the container is stopped this signal does not arrive to the binary for graceful exit. A consequence of this is when run via docker compose up, docker compose down will have to wait until the timeout in order to kill the process.
The same happens for the ccs-api.
This change adds proper SIGTERM handling to agbot_start.sh and css_start.sh.
Type of change
Please delete options that are not relevant.
How Has This Been Tested?
Running the container image with docker compose up and checking the output of docker compose down: Running
docker compose down -t 20s
without this change :With this change:
Checklist: