refactor(operators): unify api and design of operators (#187)

this is rather radical list, as one can see from amount of changes. - remove mocks for kubernetes and use fake clients instead - decople hostname operator from cluster client - decople ip operator from client - decople inventory operator from cluster client - decople metallb operator client from cluster client - unify operators clients group them within same location - unify labels and selectors across all operators Signed-off-by: Artur Troian <[email protected]>
akash-network · Feb 12, 2024 · 82c31b7 · 82c31b7
1 parent 6ce7931
commit 82c31b7
Show file tree

Hide file tree

Showing 559 changed files with 6,442 additions and 99,274 deletions.
diff --git a/.github/workflows/integration-tests.yaml b/.github/workflows/integration-tests.yaml
@@ -107,11 +107,13 @@ jobs:
       - name: K8S dump running pods
         working-directory: ${{ env.GOPATH }}/src/github.com/akash-network/provider
         run: |
-          kubectl get pods,ingress,svc -A
+          kubectl get deployments,pods,ingress,svc -A
       - name: K8S wait for operator inventory
         working-directory: ${{ env.GOPATH }}/src/github.com/akash-network/provider
         run: |
           make -s -C _run/kube kube-deployment-rollout-operator-inventory
+        env:
+          KUBE_ROLLOUT_TIMEOUT: 360
       - name: Run E2E Tests
         working-directory: ${{ env.GOPATH }}/src/github.com/akash-network/provider
         run: |

diff --git a/_docs/kustomize/akash-operator-hostname/cluster-roles.yaml b/_docs/kustomize/akash-operator-hostname/cluster-roles.yaml
@@ -0,0 +1,45 @@
+---
+apiVersion: rbac.authorization.k8s.io/v1
+kind: ClusterRole
+metadata:
+  name: akash-operator-hostname
+  labels:
+    app.kubernetes.io/name: hostname
+    app.kubernetes.io/instance: hostname-service
+    app.kubernetes.io/component: operator
+    app.kubernetes.io/part-of: provider
+rules:
+  - apiGroups:
+      - networking.k8s.io
+    resources:
+      - ingresses
+    verbs:
+      - get
+      - list
+      - create
+      - update
+      - delete
+      - deletecollection
+      - watch
+  - apiGroups:
+      - ""
+    resources:
+      - namespaces
+      - services
+    verbs:
+      - get
+  - apiGroups:
+      - akash.network
+    resources:
+      - providerhosts
+      - manifests
+    verbs:
+      - get
+      - list
+      - watch
+  - apiGroups:
+      - ""
+    resources:
+      - pods
+    verbs:
+      - get
diff --git a/_docs/kustomize/akash-operator-hostname/cluster_role.yaml b/_docs/kustomize/akash-operator-hostname/cluster_role.yaml
diff --git a/_docs/kustomize/akash-operator-hostname/deployment.yaml b/_docs/kustomize/akash-operator-hostname/deployment.yaml
@@ -2,63 +2,79 @@
 apiVersion: apps/v1
 kind: Deployment
 metadata:
-  name: akash-hostname-operator
+  name: operator-hostname
   labels:
-    akash.network/component: akash-hostname-operator
+    app.kubernetes.io/name: hostname
+    app.kubernetes.io/instance: hostname-service
+    app.kubernetes.io/component: operator
+    app.kubernetes.io/part-of: provider
 spec:
   selector:
     matchLabels:
-      app: akash-hostname-operator
+      app.kubernetes.io/name: hostname
+      app.kubernetes.io/instance: hostname-service
+      app.kubernetes.io/component: operator
+      app.kubernetes.io/part-of: provider
   replicas: 1
   revisionHistoryLimit: 1
   template:
     metadata:
       labels:
-        app: akash-hostname-operator
-        akash.network/component: akash-hostname-operator
+        app.kubernetes.io/name: hostname
+        app.kubernetes.io/instance: hostname-service
+        app.kubernetes.io/component: operator
+        app.kubernetes.io/part-of: provider
     spec:
-      serviceAccountName: akash-operator
+      serviceAccountName: operator-hostname
       containers:
-        - name: akash-hostname-operator
+        - name: operator-hostname
           image: ghcr.io/akash-network/provider:stable
           imagePullPolicy: IfNotPresent
-          args: ["provider-services", "hostname-operator"]
+          args: ["provider-services", "operator", "hostname"]
           ports:
-            - name: status
-              containerPort: 8085
+            - name: rest
+              containerPort: 8080
           env:
             - name: AP_K8S_MANIFEST_NS
               valueFrom:
                 configMapKeyRef:
-                  name: akash-provider-config
+                  name: operator-hostname
                   key: k8s-manifest-ns
             - name: AP_PRUNE_INTERVAL
               valueFrom:
                 configMapKeyRef:
-                  name: akash-hostname-operator-config
+                  name: operator-hostname
                   key: prune-interval
             - name: AP_IGNORE_LIST_ENTRY_LIMIT
               valueFrom:
                 configMapKeyRef:
-                  name: akash-hostname-operator-config
+                  name: operator-hostname
                   key: ignore-list-entry-limit
             - name: AP_WEB_REFRESH_INTERVAL
               valueFrom:
                 configMapKeyRef:
-                  name: akash-hostname-operator-config
+                  name: operator-hostname
                   key: web-refresh-interval
             - name: AP_RETRY_DELAY
               valueFrom:
                 configMapKeyRef:
-                  name: akash-hostname-operator-config
+                  name: operator-hostname
                   key: retry-delay
             - name: AP_IGNORE_LIST_AGE_LIMIT
               valueFrom:
                 configMapKeyRef:
-                  name: akash-hostname-operator-config
+                  name: operator-hostname
                   key: ignore-list-age-limit
             - name: AP_EVENT_FAILURE_LIMIT
               valueFrom:
                 configMapKeyRef:
-                  name: akash-hostname-operator-config
+                  name: operator-hostname
                   key: event-failure-limit
+            - name: AP_POD_NAME
+              valueFrom:
+                fieldRef:
+                  fieldPath: metadata.name
+            - name: AP_POD_NAMESPACE
+              valueFrom:
+                fieldRef:
+                  fieldPath: metadata.namespace
diff --git a/_docs/kustomize/akash-operator-hostname/kustomization.yaml b/_docs/kustomize/akash-operator-hostname/kustomization.yaml
@@ -6,15 +6,13 @@ resources:
   - deployment.yaml
   - service.yaml
   - ingress.yaml
-  - rbac.yaml
-  - service_account.yaml
-  - cluster_role.yaml
+  - role-bindings.yaml
+  - service-accounts.yaml
+  - cluster-roles.yaml
 configMapGenerator:
-  - name: akash-provider-config
+  - name: operator-hostname
     literals:
       - k8s-manifest-ns=lease
-  - name: akash-hostname-operator-config
-    literals:
       - prune-interval=600s
       - ignore-list-entry-limit=131072
       - web-refresh-interval=5s

diff --git a/_docs/kustomize/akash-operator-hostname/rbac.yaml b/_docs/kustomize/akash-operator-hostname/rbac.yaml
diff --git a/_docs/kustomize/akash-operator-hostname/role-bindings.yaml b/_docs/kustomize/akash-operator-hostname/role-bindings.yaml
@@ -0,0 +1,18 @@
+---
+apiVersion: rbac.authorization.k8s.io/v1
+kind: ClusterRoleBinding
+metadata:
+  name: akash-operator-hostname
+  labels:
+    app.kubernetes.io/name: hostname
+    app.kubernetes.io/instance: hostname-service
+    app.kubernetes.io/component: operator
+    app.kubernetes.io/part-of: provider
+subjects:
+  - kind: ServiceAccount
+    name: operator-hostname
+    namespace: akash-services
+roleRef:
+  kind: ClusterRole
+  name: akash-operator-hostname
+  apiGroup: rbac.authorization.k8s.io
diff --git a/_docs/kustomize/akash-operator-hostname/service-accounts.yaml b/_docs/kustomize/akash-operator-hostname/service-accounts.yaml
@@ -0,0 +1,10 @@
+---
+apiVersion: v1
+kind: ServiceAccount
+metadata:
+  name: operator-hostname
+  labels:
+    app.kubernetes.io/name: hostname
+    app.kubernetes.io/instance: hostname-service
+    app.kubernetes.io/component: operator
+    app.kubernetes.io/part-of: provider
diff --git a/_docs/kustomize/akash-operator-hostname/service.yaml b/_docs/kustomize/akash-operator-hostname/service.yaml
@@ -2,10 +2,18 @@
 apiVersion: v1
 kind: Service
 metadata:
-  name: akash-hostname-operator
+  name: operator-hostname
+  labels:
+    app.kubernetes.io/name: hostname
+    app.kubernetes.io/instance: hostname-service
+    app.kubernetes.io/component: operator
+    app.kubernetes.io/part-of: provider
 spec:
   selector:
-    akash.network/component: akash-hostname-operator
+    app.kubernetes.io/name: hostname
+    app.kubernetes.io/instance: hostname-service
+    app.kubernetes.io/component: operator
+    app.kubernetes.io/part-of: provider
   ports:
-    - name: status
-      port: 8085
+    - name: rest
+      port: 8080
diff --git a/_docs/kustomize/akash-operator-hostname/service_account.yaml b/_docs/kustomize/akash-operator-hostname/service_account.yaml
diff --git a/_docs/kustomize/akash-operator-inventory/cluster-roles.yaml b/_docs/kustomize/akash-operator-inventory/cluster-roles.yaml
@@ -2,7 +2,7 @@
 apiVersion: rbac.authorization.k8s.io/v1
 kind: ClusterRole
 metadata:
-  name: operator-inventory
+  name: akash-operator-inventory
   labels:
     akash.network: "true"
     app.kubernetes.io/name: inventory
@@ -94,7 +94,7 @@ rules:
 apiVersion: rbac.authorization.k8s.io/v1
 kind: ClusterRole
 metadata:
-  name: operator-inventory-hardware-discovery
+  name: akash-operator-inventory-hardware-discovery
   labels:
     akash.network: "true"
     app.kubernetes.io/name: inventory

diff --git a/_docs/kustomize/akash-operator-inventory/deployment.yaml b/_docs/kustomize/akash-operator-inventory/deployment.yaml
@@ -54,20 +54,20 @@ spec:
             requests:
               cpu: 100m
               memory: 128Mi
-          livenessProbe:
-            httpGet:
-              path: /metrics/health
-              port: api
-              scheme: HTTP
-            initialDelaySeconds: 15
-            periodSeconds: 15
-          readinessProbe:
-            httpGet:
-              path: /metrics/ready
-              port: api
-              scheme: HTTP
-            initialDelaySeconds: 15
-            periodSeconds: 5
+#          livenessProbe:
+#            httpGet:
+#              path: /metrics/health
+#              port: api
+#              scheme: HTTP
+#            initialDelaySeconds: 15
+#            periodSeconds: 15
+#          readinessProbe:
+#            httpGet:
+#              path: /metrics/ready
+#              port: api
+#              scheme: HTTP
+#            initialDelaySeconds: 15
+#            periodSeconds: 5
           ports:
             - containerPort: 8080
               name: api