Merge pull request #529 from bogdando/netcheck

Add a k8s app for advanced e2e netcheck for DNS

docs/netcheck.md

@@ -0,0 +1,41 @@
+Network Checker Application
+===========================
+
+With the ``deploy_netchecker`` var enabled (defaults to false), Kargo deploys a
+Network Checker Application from the 3rd side `l23network/mcp-netchecker` docker
+images. It consists of the server and agents trying to reach the server by usual
+for Kubernetes applications network connectivity meanings. Therefore, this
+automagically verifies a pod to pod connectivity via the cluster IP and checks
+if DNS resolve is functioning as well.
+
+The checks are run by agents on a periodic basis and cover standard and host network
+pods as well. The history of performed checks may be found in the agents' application
+logs.
+
+To get the most recent and cluster-wide network connectivity report, run from
+any of the cluster nodes:
+```
+curl http://localhost:31081/api/v1/connectivity_check
+```
+Note that Kargo does not invoke the check but only deploys the application, if
+requested.
+
+There are related application specifc variables:
+```
+netchecker_port: 31081
+agent_report_interval: 15
+netcheck_namespace: default
+agent_img: "quay.io/l23network/mcp-netchecker-agent:v0.1"
+server_img: "quay.io/l23network/mcp-netchecker-server:v0.1"
+```
+
+Note that the application verifies DNS resolve for FQDNs comprising only the
+combination of the ``netcheck_namespace.dns_domain`` vars, for example the
+``netchecker-service.default.cluster.local``. If you want to deploy the application
+to the non default namespace, make sure as well to adjust the ``searchdomains`` var
+so the resulting search domain records to contain that namespace, like:
+
+```
+search: foospace.cluster.local default.cluster.local ...
+nameserver: ...
+```

inventory/group_vars/all.yml

@@ -35,6 +35,8 @@ kube_users:
 cluster_name: cluster.local
 # Subdomains of DNS domain to be resolved via /etc/resolv.conf
 ndots: 5
+# Deploy netchecker app to verify DNS resolve as an HTTP service
+deploy_netchecker: false
 
 # For some environments, each node has a pubilcally accessible
 # address and an address it should bind services to.  These are

roles/dnsmasq/tasks/resolvconf.yml

@@ -48,7 +48,20 @@
   when: resolvconf.rc == 0
   notify: Dnsmasq | update resolvconf
 
-- name: Add search domains to resolv.conf
+- name: Remove search and nameserver options from resolvconf cloud init temporary file
+  lineinfile:
+    dest: "{{resolvconffile}}"
+    state: absent
+    regexp: "^{{ item }}.*$"
+    backup: yes
+    follow: yes
+  with_items:
+    - search
+    - nameserver
+  when: ansible_os_family == "CoreOS"
+  notify: Dnsmasq | update resolvconf for CoreOS
+
+- name: Add search domains to resolvconf file
   lineinfile:
     line: "search {{searchentries}}"
     dest: "{{resolvconffile}}"
@@ -66,7 +79,7 @@
       nameserver {{ item }}
       {% endfor %}
     state: present
-    insertafter: "^search.*$"
+    insertafter: "^search default.svc.*$"
     create: yes
     backup: yes
     follow: yes

roles/download/defaults/main.yml

@@ -58,6 +58,12 @@ hyperkube_image_repo: "quay.io/coreos/hyperkube"
 hyperkube_image_tag: "{{ kube_version }}_coreos.0"
 pod_infra_image_repo: "gcr.io/google_containers/pause-amd64"
 pod_infra_image_tag: "{{ pod_infra_version }}"
+netcheck_tag: v0.1
+netcheck_kubectl_tag: v0.18.0-120-gaeb4ac55ad12b1-dirty
+netcheck_agent_img_repo: "quay.io/l23network/mcp-netchecker-agent"
+netcheck_server_img_repo: "quay.io/l23network/mcp-netchecker-server"
+netcheck_kubectl_img_repo: "gcr.io/google_containers/kubectl"
+
 nginx_image_repo: nginx
 nginx_image_tag: 1.11.4-alpine
 dnsmasq_version: 2.72
@@ -73,6 +79,21 @@ test_image_repo: busybox
 test_image_tag: latest
 
 downloads:
+  netcheck_server:
+    container: true
+    repo: "{{ netcheck_server_img_repo }}"
+    tag: "{{ netcheck_tag }}"
+    enabled: "{{ deploy_netchecker|bool }}"
+  netcheck_agent:
+    container: true
+    repo: "{{ netcheck_agent_img_repo }}"
+    tag: "{{ netcheck_tag }}"
+    enabled: "{{ deploy_netchecker|bool }}"
+  netcheck_kubectl:
+    container: true
+    repo: "{{ netcheck_kubectl_img_repo }}"
+    tag: "{{ netcheck_kubectl_tag }}"
+    enabled: "{{ deploy_netchecker|bool }}"
   weave:
     dest: weave/bin/weave
     version: "{{weave_version}}"

roles/kubernetes-apps/ansible/defaults/main.yml

@@ -1,3 +1,6 @@
+kube_config_dir: /etc/kubernetes
+kube_namespace: kube-system
+
 # Versions
 kubedns_version: 1.7
 kubednsmasq_version: 1.3
@@ -20,5 +23,14 @@ exechealthz_image_tag: "{{ exechealthz_version }}"
 calico_policy_image_repo: "calico/kube-policy-controller"
 calico_policy_image_tag: latest
 
+# Netchecker
+deploy_netchecker: false
+netchecker_port: 31081
+agent_report_interval: 15
+netcheck_namespace: default
+agent_img: "quay.io/l23network/mcp-netchecker-agent:v0.1"
+server_img: "quay.io/l23network/mcp-netchecker-server:v0.1"
+kubectl_image: "gcr.io/google_containers/kubectl:v0.18.0-120-gaeb4ac55ad12b1-dirty"
+
 # SSL
 etcd_cert_dir: "/etc/ssl/etcd/ssl"

roles/kubernetes-apps/ansible/tasks/calico-policy-controller.yml

@@ -1,5 +1,5 @@
 - name: Write calico-policy-controller yaml
-  template: src=calico-policy-controller.yml.j2 dest=/etc/kubernetes/calico-policy-controller.yml
+  template: src=calico-policy-controller.yml.j2 dest={{kube_config_dir}}/calico-policy-controller.yml
   when: inventory_hostname == groups['kube-master'][0]
 
 
@@ -7,7 +7,7 @@
   kube:
     name: "calico-policy-controller"
     kubectl: "{{bin_dir}}/kubectl"
-    filename: "/etc/kubernetes/calico-policy-controller.yml"
+    filename: "{{kube_config_dir}}/calico-policy-controller.yml"
-    namespace: "kube-system"
+    namespace: "{{kube_namespace}}"
     resource: "rs"
   when: inventory_hostname == groups['kube-master'][0]

roles/kubernetes-apps/ansible/tasks/main.yaml

@@ -1,6 +1,6 @@
 ---
 - name: Kubernetes Apps | Lay Down KubeDNS Template
-  template: src={{item.file}} dest=/etc/kubernetes/{{item.file}}
+  template: src={{item.file}} dest={{kube_config_dir}}/{{item.file}}
   with_items:
     - {file: kubedns-rc.yml, type: rc}
     - {file: kubedns-svc.yml, type: svc}
@@ -10,10 +10,10 @@
 - name: Kubernetes Apps | Start Resources
   kube:
     name: kubedns
-    namespace: kube-system
+    namespace: "{{ kube_namespace }}"
     kubectl: "{{bin_dir}}/kubectl"
     resource: "{{item.item.type}}"
-    filename: /etc/kubernetes/{{item.item.file}}
+    filename: "{{kube_config_dir}}/{{item.item.file}}"
     state: "{{item.changed | ternary('latest','present') }}"
   with_items: "{{ manifests.results }}"
   when: inventory_hostname == groups['kube-master'][0]
@@ -21,3 +21,7 @@
 - include: tasks/calico-policy-controller.yml
   when: ( enable_network_policy is defined and enable_network_policy == True ) or
     ( kube_network_plugin == 'canal' )
+
+- name: Kubernetes Apps | Netchecker
+  include: tasks/netchecker.yml
+  when: deploy_netchecker

roles/kubernetes-apps/ansible/tasks/netchecker.yml

@@ -0,0 +1,20 @@
+- name: Kubernetes Apps | Lay Down Netchecker Template
+  template: src={{item.file}} dest={{kube_config_dir}}/{{item.file}}
+  with_items:
+    - {file: netchecker-agent-ds.yml, type: ds, name: netchecker-agent}
+    - {file: netchecker-agent-hostnet-ds.yml, type: ds, name: netchecker-agent-hostnet}
+    - {file: netchecker-server-pod.yml, type: po, name: netchecker-server}
+    - {file: netchecker-server-svc.yml, type: svc, name: netchecker-service}
+  register: manifests
+  when: inventory_hostname == groups['kube-master'][0]
+
+- name: Kubernetes Apps | Start Netchecker Resources
+  kube:
+    name: "{{item.item.name}}"
+    namespace: "{{netcheck_namespace}}"
+    kubectl: "{{bin_dir}}/kubectl"
+    resource: "{{item.item.type}}"
+    filename: "{{kube_config_dir}}/{{item.item.file}}"
+    state: "{{item.changed | ternary('latest','present') }}"
+  with_items: "{{ manifests.results }}"
+  when: inventory_hostname == groups['kube-master'][0]

roles/kubernetes-apps/ansible/templates/calico-policy-controller.yml.j2

@@ -2,7 +2,7 @@ apiVersion: extensions/v1beta1
 kind: ReplicaSet
 metadata:
   name: calico-policy-controller
-  namespace: kube-system
+  namespace: {{ kube_namespace }}
   labels:
     k8s-app: calico-policy
     kubernetes.io/cluster-service: "true"

roles/kubernetes-apps/ansible/templates/kubedns-rc.yml

@@ -2,7 +2,7 @@ apiVersion: v1
 kind: ReplicationController
 metadata:
   name: kubedns
-  namespace: kube-system
+  namespace: {{ kube_namespace }}
   labels:
     k8s-app: kubedns
     version: v19

roles/kubernetes-apps/ansible/templates/kubedns-svc.yml

@@ -2,7 +2,7 @@ apiVersion: v1
 kind: Service
 metadata:
   name: kubedns
-  namespace: kube-system
+  namespace: {{ kube_namespace }}
   labels:
     k8s-app: kubedns
     kubernetes.io/cluster-service: "true"

roles/kubernetes-apps/ansible/templates/netchecker-agent-ds.yml

@@ -0,0 +1,25 @@
+apiVersion: extensions/v1beta1
+kind: DaemonSet
+metadata:
+  labels:
+    app: netchecker-agent
+  name: netchecker-agent
+  namespace: {{ netcheck_namespace }}
+spec:
+  template:
+    metadata:
+      name: netchecker-agent
+      labels:
+        app: netchecker-agent
+    spec:
+      containers:
+        - name: netchecker-agent
+          image: "{{ agent_img }}"
+          env:
+            - name: MY_POD_NAME
+              valueFrom:
+                fieldRef:
+                  fieldPath: metadata.name
+            - name: REPORT_INTERVAL
+              value: '{{ agent_report_interval }}'
+          imagePullPolicy: {{ k8s_image_pull_policy }}

roles/kubernetes-apps/ansible/templates/netchecker-agent-hostnet-ds.yml

@@ -0,0 +1,26 @@
+apiVersion: extensions/v1beta1
+kind: DaemonSet
+metadata:
+  labels:
+    app: netchecker-agent-hostnet
+  name: netchecker-agent-hostnet
+  namespace: {{ netcheck_namespace }}
+spec:
+  template:
+    metadata:
+      name: netchecker-agent-hostnet
+      labels:
+        app: netchecker-agent-hostnet
+    spec:
+      hostNetwork: True
+      containers:
+        - name: netchecker-agent
+          image: "{{ agent_img }}"
+          env:
+            - name: MY_POD_NAME
+              valueFrom:
+                fieldRef:
+                  fieldPath: metadata.name
+            - name: REPORT_INTERVAL
+              value: '{{ agent_report_interval }}'
+          imagePullPolicy: {{ k8s_image_pull_policy }}

roles/kubernetes-apps/ansible/templates/netchecker-server-pod.yml

@@ -0,0 +1,21 @@
+apiVersion: v1
+kind: Pod
+metadata:
+  name: netchecker-server
+  labels:
+    app: netchecker-server
+  namespace: {{ netcheck_namespace }}
+spec:
+  containers:
+    - name: netchecker-server
+      image: "{{ server_img }}"
+      env:
+      imagePullPolicy: {{ k8s_image_pull_policy }}
+      ports:
+        - containerPort: 8081
+          hostPort: 8081
+    - name: kubectl-proxy
+      image: "{{ kubectl_image }}"
+      imagePullPolicy: {{ k8s_image_pull_policy }}
+      args:
+        - proxy

roles/kubernetes-apps/ansible/templates/netchecker-server-svc.yml

@@ -0,0 +1,15 @@
+apiVersion: v1
+kind: Service
+metadata:
+  name: netchecker-service
+  namespace: {{ netcheck_namespace }}
+spec:
+  selector:
+    app: netchecker-server
+  ports:
+    -
+      protocol: TCP
+      port: 8081
+      targetPort: 8081
+      nodePort: {{ netchecker_port }}
+  type: NodePort

roles/kubernetes/node/meta/main.yml

@@ -9,6 +9,15 @@ dependencies:
     file: "{{ downloads.nginx }}"
   - role: download
     file: "{{ downloads.testbox }}"
+  - role: download
+    file: "{{ downloads.netcheck_server }}"
+    when: deploy_netchecker
+  - role: download
+    file: "{{ downloads.netcheck_agent }}"
+    when: deploy_netchecker
+  - role: download
+    file: "{{ downloads.netcheck_kubectl }}"
+    when: deploy_netchecker
   - role: download
     file: "{{ downloads.kubednsmasq }}"
     when: not skip_dnsmasq_k8s|default(false)