DNS for xds-address

[stevesloka]

I'm testing through splitting envoy from contour and it doesn't seem to be working with a DNS name, but using IP address works fine. This was mentioned and resolved in #228. Following are the logs from my envoy pod and my deployment manifests are here.

Curious if anyone else has had luck with this or if something has changed since #238 landed.

[2018-05-02 14:42:58.083][1][warning][upstream] source/common/config/grpc_mux_impl.cc:36] Unable to establish new stream
[2018-05-02 14:42:58.083][1][warning][config] bazel-out/k8-opt/bin/source/common/config/_virtual_includes/grpc_mux_subscription_lib/common/config/grpc_mux_subscription_impl.h:66] gRPC update fortype.googleapis.com/envoy.api.v2.Cluster failed

[stevesloka]

Also I can confirm my DNS works fine:

$ kubectl exec -ti busybox -- nslookup contour.gimbal-contour.svc.cluster.local                                       [10:34:39]
Server:    10.96.0.10
Address 1: 10.96.0.10 kube-dns.kube-system.svc.cluster.local

Name:      contour.gimbal-contour.svc.cluster.local
Address 1: 10.108.227.232 contour.gimbal-contour.svc.cluster.local

My config also contains the change STRICT_DNS:

dynamic_resources:
  lds_config:
    api_config_source:
      api_type: GRPC
      cluster_names: [contour]
      grpc_services:
      - envoy_grpc:
          cluster_name: contour
  cds_config:
    api_config_source:
      api_type: GRPC
      cluster_names: [contour]
      grpc_services:
      - envoy_grpc:
          cluster_name: contour
static_resources:
  clusters:
  - name: contour
    connect_timeout: { seconds: 5 }
    type: STRICT_DNS
    hosts:
    - socket_address:
        address: contour.gimbal-contour.svc.cluster.local
        port_value: 8001
    lb_policy: ROUND_ROBIN
    http2_protocol_options: {}
    circuit_breakers:
      thresholds:
        - priority: high
          max_connections: 100000
          max_pending_requests: 100000
          max_requests: 60000000
          max_retries: 50
        - priority: default
          max_connections: 100000
          max_pending_requests: 100000
          max_requests: 60000000
          max_retries: 50
admin:
  access_log_path: /dev/null
  address:
    socket_address:
      address: 0.0.0.0
      port_value: 9001

[stevesloka]

Also, I tried using the service env variable by setting the xds-address to $(CONTOUR_PORT_8001_TCP_ADDR):

It seems to load, but the listeners never get added:

[2018-05-02 18:06:27.932][1][info][main] source/server/server.cc:178] initializing epoch 0 (hot restart version=9.200.16384.127.options=capacity=16384, num_slots=8209 hash=228984379728933363)
[2018-05-02 18:06:27.939][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:127] cm init: initializing cds
[2018-05-02 18:06:27.939][1][info][config] source/server/configuration_impl.cc:52] loading 0 listener(s)
[2018-05-02 18:06:27.940][1][info][config] source/server/configuration_impl.cc:92] loading tracing configuration
[2018-05-02 18:06:27.940][1][info][config] source/server/configuration_impl.cc:119] loading stats sink configuration
[2018-05-02 18:06:27.940][1][info][main] source/server/server.cc:353] starting main dispatch loop
[2018-05-02 18:06:27.942][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster default/kuard/80 during init
[2018-05-02 18:06:27.943][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster default/kubernetes/443 during init
[2018-05-02 18:06:27.943][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster default/kubernetes/https during init
[2018-05-02 18:06:27.943][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster default/nginx/80 during init
[2018-05-02 18:06:27.944][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster docker/compose-api/443 during init
[2018-05-02 18:06:27.945][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster docker/compose-api/api during init
[2018-05-02 18:06:27.945][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster gimbal-contour/contour/8001 during init
[2018-05-02 18:06:27.946][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster gimbal-contour/contour/xds during init
[2018-05-02 18:06:27.947][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster gimbal-contour/envoy/443 during init
[2018-05-02 18:06:27.948][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster gimbal-contour/envoy/80 during init
[2018-05-02 18:06:27.948][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster gimbal-contour/envoy/http during init
[2018-05-02 18:06:27.949][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster gimbal-contour/envoy/https during init
[2018-05-02 18:06:27.950][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster gimbal-monitoring/grafana/80 during init
[2018-05-02 18:06:27.951][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster gimbal-monitoring/grafana/http during init
[2018-05-02 18:06:27.951][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster gimbal-monitoring/prometheus-alertmanager/80 during init
[2018-05-02 18:06:27.952][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster gimbal-monitoring/prometheus-alertmanager/http during init
[2018-05-02 18:06:27.952][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster gimbal-monitoring/prometheus-node-exporter/9100 during init
[2018-05-02 18:06:27.953][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster gimbal-monitoring/prometheus-node-exporter/metrics during init
[2018-05-02 18:06:27.953][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster gimbal-monitoring/prometheus/9090 during init
[2018-05-02 18:06:27.953][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster gimbal-monitoring/prometheus/9093 during init
[2018-05-02 18:06:27.954][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster gimbal-monitoring/prometheus/alertmanager during init
[2018-05-02 18:06:27.954][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster gimbal-monitoring/prometheus/prometheus during init
[2018-05-02 18:06:27.955][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster kube-system/kube-dns/53 during init
[2018-05-02 18:06:27.955][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster kube-system/kube-dns/dns-tcp during init
[2018-05-02 18:06:27.955][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster kube-system/kube-state-metrics/8080 during init
[2018-05-02 18:06:27.956][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster kube-system/kube-state-metrics/8081 during init
[2018-05-02 18:06:27.956][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster kube-system/kube-state-metrics/http-metrics during init
[2018-05-02 18:06:27.956][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:382] add/update cluster kube-system/kube-state-metrics/telemetry during init
[2018-05-02 18:06:27.956][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:108] cm init: initializing secondary clusters
[2018-05-02 18:06:27.959][1][info][upstream] source/common/upstream/cluster_manager_impl.cc:131] cm init: all clusters initialized
[2018-05-02 18:06:27.959][1][info][main] source/server/server.cc:337] all clusters initialized. initializing init manager

[stevesloka]

I think the issue is with current build on master tag (#363), using v0.5.0 the ENV var works.