HttpPort parameter doesn't works

[leoblanc]

Is this a request for help?: No

Is this a BUG REPORT or FEATURE REQUEST? (choose one): BUG REPORT

Version of Helm and Kubernetes: Using Kubernetes 1.6.6 and Helm 2.5.0

Which chart: Consul (https://github.com/kubernetes/charts/tree/master/stable/consul)

What happened: After assign the HttpPort parameter to 80, the Consul UI can't be reached (I'm using LoadBalancer as Service Type)

What you expected to happen: See the Consul UI (as it's shown when you use the default port 8500)

How to reproduce it (as minimally and precisely as possible): Just set the HttpPort to a value different from 8500. For example:

   helm install \
     --name consul \
     --namespace default \
     --set "uiService.enabled=true,uiService.type=LoadBalancer,HttpPort=80" \
     stable/consul

Anything else we need to know: By the way, the "ui.enabled" parameter described on the documentation it's not used at all on the Helm Chart, but that's a different issue.

I have received the help of @netflash to investigate this issue. He discovered that, when "uiService.enabled" is set to true, only the "-ui" argument is passed to the Consul server (https://github.com/kubernetes/charts/blob/master/stable/consul/templates/consul.yaml#L179). This is ok, but is not enough if "HttpPort" was also used to indicate a different port. According to https://www.consul.io/docs/agent/options.html, the "ports/http" setting should be changed also:

Thank you very much in advance!

[fejta-bot]

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.

Prevent issues from auto-closing with an /lifecycle frozen comment.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or @fejta.
/lifecycle stale

[netflash]

/remove-lifecycle stale

[jonstacks]

@leoblanc,

Good find. Thanks for the great details. I was able to re-produce from the example you gave. I'm hoping the above PR will fix your issue. Also, just curious, is there any reason you are running the UI Service as a Load Balancer instead of an Ingress? Trying to find out how best this chart can be tailored to the community's needs.

[leoblanc]

@jonstacks Hi Jonathan, thank you very much for your patch! Yes, I think that it solves this issue, hope it can be merged soon. Regarding your question: we found this problem trying to create a minimal prototype of Consul on k8s on AWS, and that's why we used a LoadBalancer. We thought that fix this issue could be useful for other people too, that's why we reported the issue. By the way, we were contacted by a real client recently to deploy Consul on k8s, so I think that we will be using this chart in production soon. Thanks again!