[bitnami/contour] Add ability to specify ingressClass and restart pods on config change

[mkilchhofer]

Description of the change
As mentioned in the documentation on projectcontour.io, Contour supports watching a specific ingressClass only.

Benefits

• Also as mentioned in the docs above:

  This can be useful while you are migrating from another controller, or if you need multiple instances of Contour.

• Contour pods are automatically restarted after we need to change configInline
• While I was already editing this chart I updated some missing settings in README.md

Possible drawbacks
-

Applicable issues
-

• fixes #

Additional information
-

Checklist

• Chart version bumped in Chart.yaml according to semver.
• Variables are documented in the README.md
• Title of the PR starts with chart name (e.g. [bitnami/chart])
• If the chart contains a values-production.yaml apart from values.yaml, ensure that you implement the changes in both files

⚠️ Keep in mind that if you want to make changes to the kubeapps chart, please implement them in the kubeapps repository. This is only a synchronized mirror.

[mkilchhofer]

As far as I can see the tests are failling because envoy DaemonSet has 300sec to terminate (Line 305 in values.yaml):

terminationGracePeriodSeconds: 300

Deleting release 'contour-gmywl0vz3q'...
>>> helm uninstall contour-gmywl0vz3q --namespace contour-gmywl0vz3q
release "contour-gmywl0vz3q" uninstalled
Deleting namespace 'contour-gmywl0vz3q'...
>>> kubectl delete namespace contour-gmywl0vz3q --timeout 180s
namespace "contour-gmywl0vz3q" deleted
error: timed out waiting for the condition on namespaces/contour-gmywl0vz3q
Namespace 'contour-gmywl0vz3q' did not terminate after 180s.
>>> kubectl get namespace contour-gmywl0vz3q
Namespace 'contour-gmywl0vz3q' did not terminate after 180s.
Force-deleting everything...
>>> kubectl delete all --namespace contour-gmywl0vz3q --all --force --grace-period=0
pod "contour-gmywl0vz3q-envoy-tpd2l" force deleted
warning: Immediate deletion does not wait for confirmation that the running resource has been terminated. The resource may continue to run on the cluster indefinitely.
>>> kubectl get namespace contour-gmywl0vz3q
>>> kubectl get namespace contour-gmywl0vz3q --output=json
Running 'kubectl proxy' on port 37631
>>> kubectl proxy --port=37631
Removing finalizers from namespace 'contour-gmywl0vz3q'...
>>> kubectl get namespace contour-gmywl0vz3q
Error: Error installing charts: Error processing charts
Namespace 'contour-gmywl0vz3q' terminated.

As far as I can see, we have these options to fix the failling tests:

• Skip the install tests for this ingress controller too (like for kong)?
  https://github.com/bitnami/charts/blob/master/.github/ct-install.yaml#L10
• Inject other values for the chart tests and set terminationGracePeriodSeconds below 180s?
• Increase timeout for all chart tests like the guys from jaegertracing (maybe we just double it from 180s to 360s?):
  jaegertracing/helm-charts@ed715ea#diff-b0299b9f73bfa783dc8d9e3d05712fb1

[marcosbc]

Hi @mkilchhofer, thanks for the contribution! The change looks fine, I'm proceeding to merge it.

As for the tests, we recently enabled PR chart testing so some failures like these are expected until we polish things up. Note that we run more reliable tests in our internal pipeline, and in the case of Contour they're currently succeeding.

[mkilchhofer]

Hi @marcosbc , you are welcome. Thanks for accepting this change 😎