Do not retry /run on action containers
It should be assumed that /run only happens after /init which should itself ensure that the container is running and ready. With that in mind, it does not make sense to attempt to retry posting to the /run endpoint on connection failures. Such a failure is an indication that something catastrophic has happened to the container, and no attempt at recovery should be made.
This is accomplished by adding a "retry" flag to the appropriate HTTP utilities to control whether or not a retry should be attempted.
1 file changed
