I have a stack that includes some fairly complicated containers that can’t be simply restarted. They are running legacy software and we don’t want to re-engineer these apps to behave more like micro-services.
So I want to use Rancher to monitor the stack, and if one of the containers is sluggish of non-responsive, I don’t want Rancher to take any action, other than perhaps changing the state of the container, service, etc, and changing the color of the container in the UI.
Is there a way to tell Rancher not to restart a container on a healthcheck failure?
I wish I could provide a call-back or something like that for the healthcheck.