Search code examples
sidekiqmesosmarathon

monitoring dockerized sidekiq workers in marathon/mesos

How to specify a healthcheck for a sidekiq worker inside a docker spawned by mesos?

it has no listening tcp endpoint to query

from marathon api docs i have a possibility of using

 {
            "protocol": "COMMAND",
            "command": { "value": "curl -f -X GET http://$HOST:$PORT0/health" },
            "maxConsecutiveFailures": 3
        }

but I don't see it working in this case...

Solution

  • Clarification: When a command dies for whatever reason, the docker container is stopped and re-spawned.