Ingress-nginx: Enabling OpenTracing makes ingress controller unhealthy

[rafaeljesus]

Is this a request for help? Yes I've asked for help in slack channel

What keywords did you search in NGINX Ingress controller issues before filing this one? OpenTracing, Jaeger

---

BUG REPORT

NGINX Ingress controller version: 0.17.1
Kubernetes version (use kubectl version): v1.10.6
Environment:

• Cloud provider or hardware configuration: AWS
• OS (e.g. from /etc/os-release): Ubuntu 16.04.5 LTS
• Kernel (e.g. uname -a): 4.4.0-1065-aws

What happened:
I am trying to enable opentracing jaeger to my ingress controller, however the controller pod health check fails and it does not work. I am editing the controller configmap and then I delete one pod to test and tee latter health check fails.

configmap

enable-opentracing: "true"
jaeger-collector-host: jaeger-collector.tracing.svc.cluster.local
jaeger-collector-port: "14267"
jaeger-sampler-type: "probabilistic"
jaeger-sampler-param: "0.3"
jaeger-service-name: "ingress-nginx"

error on the logs

nginx_status.go:207] unexpected error obtaining nginx status info: unexpected error scraping nginx status page: unexpected error scraping nginx : Get http://0.0.0.0:18080/nginx_status: dial tcp 0.0.0.0:18080: connect: connection refused

The pod works fine after I disable opentracing, when I check the opentracing.json I see:

"reporter": {
    "localAgentHostPort": "jaeger-collector.tracing.svc.cluster.local:14267"
  }

What you expected to happen:
At least some error message describe any error about the opentracing config is properly set.

Thank you

[aledbf]

nginx_status.go:207] unexpected error obtaining nginx status info: unexpected error scraping nginx status page: unexpected error scraping nginx : Get http://0.0.0.0:18080/nginx_status: dial tcp 0.0.0.0:18080: connect: connection refused

This is not an error. The reason for this message is the scraping of prometheus metrics start before the ingress controller. This is fixed in 0.19

[rafaeljesus]

thx for the quick reply! Interesting, it only happen when I enable the trace and the pod is never ready. I will update to the 0.19 version anyway

[rafaeljesus]

@aledbf I was able to reproduce now and got Error: signal: segmentation fault (core dumped) the same as here https://github.com/kubernetes/ingress-nginx/issues/2856 so as you mentioned in that PR also it should be fixed in 0.19. Will reopen the issue in case the error persists, thank you

[rafaeljesus]

As a reference, using zipkin config fields worked