I would like to use a full elastic stack instance on a production grade Kubernetes cluster managed by GKE. I'm pretty new to Kubernetes in general but especially Helm.

I have tried running the following for both versions 1.6.0 and 1.5.1 of the following chart:

helm install stable/elastic-stack --version 1.5.1 --name test-elastic

Most of the pods seem to spin up fine, except for the StatefulSet test-elastic-logstash, where the pod within seems to fail with a reproducible CrashLoopBackOff error. The readiness probe seems to continually fail with connection refused, but that's not something I find especially constructive.

Could anyone please offer any thoughts on what's required to get this working?