site: add troubleshooting doc for unready Envoy

Also recommends setting resource requests
on containers.

Updates projectcontour#4851.

Signed-off-by: Steve Kriss <[email protected]>

site/content/docs/main/deploy-options.md

@@ -12,6 +12,19 @@ This secret can be auto-generated by the Contour `certgen` job or provided by an
 Traffic must be forwarded to Envoy, typically via a Service of `type: LoadBalancer`.
 All other requirements such as RBAC permissions, configuration details, are provided or have good defaults for most installations.
 
+### Setting resource requests and limits
+
+It is recommended that resource requests and limits be set on all Contour and Envoy containers.
+The example YAML manifests used in the [Getting Started][8] guide do not include these, because the appropriate values can vary widely from user to user.
+The table below summarizes the Contour and Envoy containers, and provides some reasonable resource requests to start with (note that these should be adjusted based on observed usage and expected load):
+
+| Workload            | Container        | Request (mem) | Request (cpu) |
+| ------------------- | ---------------- | ------------- | ------------- |
+| deployment/contour  | contour          | 128Mi         | 250m          |
+| daemonset/envoy     | envoy            | 256Mi         | 500m          |
+| daemonset/envoy     | shutdown-manager | 50Mi          | 25m           |
+
+
 ### Envoy as Daemonset
 
 The recommended installation is for Contour to run as a Deployment and Envoy to run as a Daemonset.

site/content/docs/main/troubleshooting.md

@@ -23,6 +23,9 @@ Learn how to profile Contour by using [net/http/pprof][11] handlers.
 ### [Contour Operator][8]
 Follow the linked guide to learn how to troubleshoot issues with [Contour Operator][12].
 
+### [Envoy container stuck in unready/draining state][13]
+Read the linked document if you have Envoy containers stuck in an unready/draining state.
+
 [0]: {{< param github_url >}}/issues
 [1]: {{< param slack_url >}}
 [2]: /docs/{{< param latest_version >}}/troubleshooting/envoy-admin-interface/
@@ -36,3 +39,4 @@ Follow the linked guide to learn how to troubleshoot issues with [Contour Operat
 [10]: https://www.envoyproxy.io/docs/envoy/latest/api-docs/xds_protocol
 [11]: https://golang.org/pkg/net/http/pprof/
 [12]: https://github.com/projectcontour/contour-operator
+[13]: /docs/{{< param latest_version >}}/troubleshooting/envoy-container-draining/

site/content/docs/main/troubleshooting/envoy-container-draining.md

@@ -0,0 +1,28 @@
+# Envoy container stuck in unready/draining state
+
+It's possible for the Envoy containers to become stuck in an unready/draining state.
+This is an unintended side effect of the shutdown-manager sidecar container being restarted by the kubelet.
+For more details on exactly how this happens, see [this issue][1].
+
+If you observe Envoy containers in this state, you should `kubectl delete` them to allow new Pods to be created to replace them.
+
+To make this issue less likely to occur, you should:
+- ensure you have [resource requests][2] on all your containers
+- ensure you do **not** have a liveness probe on the shutdown-manager sidecar container in the envoy daemonset (this was removed from the example YAML in Contour 1.24.0).
+
+If the above are not sufficient for preventing the issue, you may also add a liveness probe to the envoy container itself, like the following:
+
+```yaml
+livenessProbe:
+  httpGet:
+    path: /ready
+    port: 8002
+  initialDelaySeconds: 15
+  periodSeconds: 5
+  failureThreshold: 6
+```
+
+This will help the envoy container recover if it does become stuck in this state.
+
+[1]: https://github.com/projectcontour/contour/issues/4851
+[2]: /docs/{{< param latest_version >}}/deploy-options/#setting-resource-requests-and-limits