External DNS Failing to Discover Services Workaround

Many external DNS issues can be worked around by directly connecting to the desired backend service. This can circumvent authentication and authorization protections, but it may be necessary to access specific services when mitigating critical issues.

Istio’s ingress gateway uses Gateway and VirtualService objects to configure how traffic is routed to backend services. Currently, there is only one gateway supporting the Customer Access Network (CAN), which is services/services-gateway. It is configured to support traffic for any host. It is the VirtualService objects that ultimately control routing based on hostname.

Use this procedure to resolve any external DNS routing issues with backend services.

Procedure

  1. Search for the VirtualService object that corresponds to the desired service.

    The command below will list all external hostnames.

    ncn-mw# kubectl get vs -A | grep -v '[*]'

    Example output:

    NAMESPACE        NAME                              GATEWAYS                       HOSTS                                                      AGE
istio-system     kiali                             [services/services-gateway]    [kiali-istio.SYSTEM_DOMAIN_NAME]                           2d16h
istio-system     prometheus                        [services/services-gateway]    [prometheus-istio.SYSTEM_DOMAIN_NAME]                      2d16h
istio-system     tracing                           [services/services-gateway]    [jaeger-istio.SYSTEM_DOMAIN_NAME]                          2d16h
nexus            nexus                             [services/services-gateway]    [packages.local registry.local nexus.SYSTEM_DOMAIN_NAME]   2d16h
services         gitea-vcs-external                [services/services-gateway]    [vcs.SYSTEM_DOMAIN_NAME]                                   2d16h
services         sma-grafana                       [services-gateway]             [sma-grafana.SYSTEM_DOMAIN_NAME]                           2d16h
services         sma-kibana                        [services-gateway]             [sma-kibana.SYSTEM_DOMAIN_NAME]                            2d16h
sysmgmt-health   cray-sysmgmt-health-alertmanager  [services/services-gateway]    [alertmanager.SYSTEM_DOMAIN_NAME]                          2d16h
sysmgmt-health   cray-sysmgmt-health-grafana       [services/services-gateway]    [grafana.SYSTEM_DOMAIN_NAME]                               2d16h
sysmgmt-health   cray-sysmgmt-health-prometheus    [services/services-gateway]    [prometheus.SYSTEM_DOMAIN_NAME]                            2d16h

  2. Inspect the VirtualService objects to learn the destination service and port.

    Use the NAME value returned in the previous step. The following example is for the cray-sysmgmt-health-prometheus service.

    ncn-mw# kubectl get vs -n sysmgmt-health cray-sysmgmt-health-prometheus -o yaml

    Example output:

    apiVersion: networking.istio.io/v1beta1
kind: VirtualService
metadata:
  creationTimestamp: "2020-07-09T17:49:07Z"
  generation: 1
  labels:
    app: cray-sysmgmt-health-prometheus
    app.kubernetes.io/instance: cray-sysmgmt-health
    app.kubernetes.io/managed-by: Tiller
    app.kubernetes.io/name: cray-sysmgmt-health
    app.kubernetes.io/version: 8.15.4
    helm.sh/chart: cray-sysmgmt-health-0.3.1
  name: cray-sysmgmt-health-prometheus
  namespace: sysmgmt-health
  resourceVersion: "41620"
  selfLink: /apis/networking.istio.io/v1beta1/namespaces/sysmgmt-health/virtualservices/cray-sysmgmt-health-prometheus
  uid: d239dfcc-a827-4a51-9b73-6eccfb937088
spec:
  gateways:
  - services/services-gateway
  hosts:
  - prometheus.SYSTEM_DOMAIN_NAME
  http:
  - match:
    - authority:
        exact: prometheus.SYSTEM_DOMAIN_NAME
    route:
    - destination:
        host: cray-sysmgmt-health-promet-prometheus
        port:
          number: 9090

    From the VirtualService data, it is straightforward to see how traffic will be routed. In this example, connections to prometheus.SYSTEM_DOMAIN_NAME will be routed to the cray-sysmgmt-health-prometheus service in the sysmgmt-health namespace on port 9090.

External DNS will now be connected to the backend service.