I just found something which I find quite annoying. Currently - due to a not-yet-perfect testing pipeline - we create images which might not run all the time.
Now if I deploy a catalog which contains such an image, the service won’t come up (it keeps restarting, and crashing, basically). If we create a new image from our build pipeline, I can’t force the unhealthy service to re-pull the image. I have to delete the full stack and re-create it. This is a bit annoying I think.
Is there a way around it? This might only occur on stacks which have never been healthy before, but still.
Thanks for ideas!