Sidecar Injection Problems

    1. Ensure your pod is not in the kube-system or kube-public namespace. Automatic sidecar injection will be ignored for pods in these namespaces.

    2. Ensure your pod does not have hostNetwork: true in its pod spec. Automatic sidecar injection will be ignored for pods that are on the host network.

      The sidecar model assumes that the iptables changes required for Envoy to intercept traffic are within the pod. For pods on the host network this assumption is violated, and this can lead to routing failures at the host level.

    3. Check the webhook’s namespaceSelector to determine whether the webhook is scoped to opt-in or opt-out for the target namespace.

      The namespaceSelector for opt-in will look like the following:

      The injection webhook will be invoked for pods created in namespaces with the istio-injection=enabled label.

      1. $ kubectl get namespace -L istio-injection
      2. NAME           STATUS    AGE       ISTIO-INJECTION
      3. default        Active    18d       enabled
      4. istio-system   Active    3d
      5. kube-public    Active    18d
      6. kube-system    Active    18d

      The namespaceSelector for opt-out will look like the following:

      1. $ kubectl get mutatingwebhookconfiguration istio-sidecar-injector -o yaml | grep "namespaceSelector:" -A5
      2.   namespaceSelector:
      3.     matchExpressions:
      4.     - key: istio-injection
      5.       operator: NotIn
      6.       values:
      7.   rules:
      8.   - apiGroups:
      9.     - ""

      The injection webhook will be invoked for pods created in namespaces without the istio-injection=disabled label.

      1. $ kubectl get namespace -L istio-injection
      2. NAME           STATUS    AGE       ISTIO-INJECTION
      3. default        Active    18d
      4. istio-system   Active    3d        disabled
      5. kube-public    Active    18d       disabled

      Verify the application pod’s namespace is labeled properly and (re) label accordingly, e.g.

      1. $ kubectl label namespace istio-system istio-injection=disabled --overwrite

    4. Check default policy

      Check the default injection policy in the istio-sidecar-injector configmap.

      1. $ kubectl -n istio-system get configmap istio-sidecar-injector -o jsonpath='{.data.config}' | grep policy:
      2. policy: enabled

      Allowed policy values are disabled and enabled. The default policy only applies if the webhook’s namespaceSelector matches the target namespace. Unrecognized policy causes injection to be disabled completely.

    5. Check the per-pod override annotation

      The default policy can be overridden with the sidecar.istio.io/inject annotation in the pod template spec’s metadata. The deployment’s metadata is ignored. Annotation value of true forces the sidecar to be injected while a value of false forces the sidecar to not be injected.

      The following annotation overrides whatever the default policy was to force the sidecar to be injected:

      1. $ kubectl get deployment sleep -o yaml | grep "sidecar.istio.io/inject:" -C3
      2. template:
      3.   metadata:
      4.     annotations:
      5.       sidecar.istio.io/inject: "true"
      6.     labels:
      7.       app: sleep

    Run kubectl describe -n namespace deployment name on the failing pod’s deployment. Failure to invoke the injection webhook will typically be captured in the event log.

    1. Warning  FailedCreate  3m (x17 over 8m)  replicaset-controller  Error creating: Internal error occurred: \
    2.     x509: certificate signed by unknown authority (possibly because of "crypto/rsa: verification error" while trying \
    3.     to verify candidate authority certificate "Kubernetes.cluster.local")

    x509: certificate signed by unknown authority errors are typically caused by an empty caBundle in the webhook configuration.

    Verify the caBundle in the mutatingwebhookconfiguration matches the root certificate mounted in the istiod pod.

    1. $ kubectl get mutatingwebhookconfiguration istio-sidecar-injector -o yaml -o jsonpath='{.webhooks[0].clientConfig.caBundle}' | md5sum
    2. 4b95d2ba22ce8971c7c92084da31faf0  -
    4. 4b95d2ba22ce8971c7c92084da31faf0  -

    no such hosts or no endpoints available errors in deployment status

    Injection is fail-close. If the istio-sidecar-injector pod is not ready, pods cannot be created. In such cases you’ll see an error about no endpoints available.

    1. Internal error occurred: failed calling admission webhook "istio-sidecar-injector.istio.io": \
    2.     Post https://istio-sidecar-injector.istio-system.svc:443/admitPilot?timeout=30s: \
    3.     no endpoints available for service "istio-sidecar-injector"
    1. $  kubectl -n istio-system get pod -listio=sidecar-injector
    2. NAME                            READY     STATUS    RESTARTS   AGE
    3. istio-sidecar-injector-5dbbbdb746-d676g   1/1       Running   0          2d
    1. $ kubectl -n istio-system get endpoints istio-sidecar-injector
    2. NAME           ENDPOINTS                          AGE
    3. istio-sidecar-injector   10.48.6.108:15014,10.48.6.108:443   3d

    If the pods or endpoints aren’t ready, check the pod logs and status for any indication about why the webhook pod is failing to start and serve traffic.

    1. $ for pod in $(kubectl -n istio-system get pod -listio=sidecar-injector -o jsonpath='{.items[*].metadata.name}'); do \
    2.     kubectl -n istio-system logs ${pod} \
    3. done
    5. $ for pod in $(kubectl -n istio-system get pod -listio=sidecar-injector -o name); do \
    6.     kubectl -n istio-system describe ${pod} \
    7. done

    When the Kubernetes API server includes proxy settings such as:

    With these settings, Sidecar injection fails. The only related failure log can be found in kube-apiserver log:

      Make sure both pod and service CIDRs are not proxied according to *_proxy variables. Check the kube-apiserver files and logs to verify the configuration and whether any requests are being proxied.

      One workaround is to remove the proxy settings from the kube-apiserver manifest, another workaround is to include istio-sidecar-injector.istio-system.svc or .svc in the no_proxy value. Make sure that kube-apiserver is restarted after each workaround.

      An was filed with Kubernetes related to this and has since been closed. https://github.com/kubernetes/kubernetes/pull/58698#discussion_r163879443

      Tcpdump doesn’t work in the sidecar pod - the container doesn’t run as root. However any other container in the same pod will see all the packets, since the network namespace is shared. will also see the pod-wide configuration.

      Communication between Envoy and the app happens on 127.0.0.1, and is not encrypted.