Merge 5936b36ab5 into 80d62e30a4

observability/README.md

@@ -0,0 +1,28 @@
+
+# Observability Stack
+
+This directory contains an observability implementation based on Grafana tooling
+
+## Caveats
+1) reliance on ref-implementation for SSO
+   - This is possible to work around by removing the `auth.generic_oauth` section from `prometheus.yaml` and removing the `grafana-config.yaml` and `grafana-external-secret.yaml` files
+2) using `tls_skip_verify_insecure` for oauth
+    - This is due to using the ingress certificate. Once this is addressed, we can remove this
+3) Bigger memory requirement required for kind cluster
+    - Due to using a more robust loki deployment, the memory limits have been increased. 16 GB seems to work while leaving ample room in the cluster. 
+
+## Components
+The observability stack is built upon:
+- Prometheus - metrics
+- Loki - logging
+  - Promtail - log delivery
+- Opencost - cost accounting
+- Grafana - visualization
+- Alertmanager - alerting
+
+## Installation
+Note: The stack is configured to use Keycloak for SSO; therefore, the ref-implementation is required for this to work. 
+
+`idpbuilder create --use-path-routing --package-dir ./ref-implementation --package-dir ./observability`
+
+A `grafana-config` job will be deployed into the keycloak namespace to create/patch some of the keycloak components. If deployed at the same time as the `ref-implementation`, this job will fail until the `config` job succeeds. This is normal

observability/loki.yaml

@@ -0,0 +1,50 @@
+apiVersion: argoproj.io/v1alpha1
+kind: Application
+metadata:
+  name: loki
+  namespace: argocd
+  labels:
+    env: dev
+  finalizers:
+    - resources-finalizer.argocd.argoproj.io
+spec:
+  project: default
+  sources:
+    - repoURL: 'https://grafana.github.io/helm-charts'
+      targetRevision: 6.6.3
+      helm:
+        releaseName: loki
+        values: |
+          deploymentMode: SingleBinary
+          loki:
+            commonConfig:
+              replication_factor: 1
+            storage:
+              type: 'filesystem'
+            schemaConfig:
+              configs:
+                - from: "2024-01-01"
+                  store: tsdb
+                  index:
+                    prefix: loki_index_
+                    period: 24h
+                  object_store: filesystem # we're storing on filesystem so there's no real persistence here.
+                  schema: v13
+          singleBinary:
+            replicas: 1
+          read:
+            replicas: 0
+          backend:
+            replicas: 0
+          write:
+            replicas: 0
+      chart: loki
+  destination:
+    server: "https://kubernetes.default.svc"
+    namespace: monitoring
+  syncPolicy:
+    syncOptions:
+      - CreateNamespace=true
+      - ServerSideApply=true
+    automated:
+      selfHeal: true

observability/opencost.yaml

@@ -0,0 +1,33 @@
+apiVersion: argoproj.io/v1alpha1
+kind: Application
+metadata:
+  name: opencost
+  namespace: argocd
+  labels:
+    env: dev
+  finalizers:
+    - resources-finalizer.argocd.argoproj.io
+spec:
+  project: default
+  sources:
+    - repoURL: 'https://opencost.github.io/opencost-helm-chart'
+      targetRevision: 1.38.1
+      helm:
+        releaseName: opencost
+        values: |
+          opencost:
+            prometheus:
+              internal:
+                serviceName: prometheus-kube-prometheus-prometheus
+                namespaceName: monitoring
+                port: 9090
+      chart: opencost
+  destination:
+    server: "https://kubernetes.default.svc"
+    namespace: monitoring
+  syncPolicy:
+    syncOptions:
+      - CreateNamespace=true
+      - ServerSideApply=true
+    automated:
+      selfHeal: true

observability/prometheus.yaml

@@ -0,0 +1,59 @@
+apiVersion: argoproj.io/v1alpha1
+kind: Application
+metadata:
+  name: prometheus
+  namespace: argocd
+  labels:
+    env: dev
+  finalizers:
+    - resources-finalizer.argocd.argoproj.io
+spec:
+  project: default
+  sources:
+    - repoURL: 'https://prometheus-community.github.io/helm-charts'
+      targetRevision: 57.2.0
+      helm:
+        releaseName: prometheus
+        values: |
+          grafana:
+            envFromSecret: grafana-oidc
+            additionalDataSources:
+              - name: loki
+                access: proxy
+                orgId: 1
+                type: loki
+                url: http://loki-gateway
+                jsonData:
+                  httpHeaderName1: X-Scope-OrgID
+                secureJsonData:
+                  httpHeaderValue1: '1'
+            grafana.ini:
+              server:
+                root_url: https://cnoe.localtest.me:8443/grafana
+                serve_from_sub_path: true
+              auth.generic_oauth:
+                enabled: true
+                name: grafana
+                allow_sign_up: true
+                auth_url: https://cnoe.localtest.me:8443/keycloak/realms/cnoe/protocol/openid-connect/auth
+                token_url: https://cnoe.localtest.me:8443/keycloak/realms/cnoe/protocol/openid-connect/token
+                api_url: https://cnoe.localtest.me:8443/keycloak/realms/cnoe/protocol/openid-connect/userinfo
+                scopes: openid email profile offline_access roles
+                role_attribute_path: contains(resource_access.grafana.roles[*], 'admin') && 'GrafanaAdmin' || contains(resource_access.grafana.roles[*], 'admin') && 'Admin' || contains(resource_access.grafana.roles[*], 'editor') && 'Editor' || 'Viewer'
+                allow_assign_grafana_admin: true
+                role_attribute_strict: true
+                auto_login: true
+                tls_skip_verify_insecure: true
+      chart: kube-prometheus-stack
+    - repoURL: cnoe://prometheus
+      targetRevision: HEAD
+      path: "manifests"
+  destination:
+    server: "https://kubernetes.default.svc"
+    namespace: monitoring
+  syncPolicy:
+    syncOptions:
+      - CreateNamespace=true
+      - ServerSideApply=true
+    automated:
+      selfHeal: true

observability/prometheus/manifests/grafana-config.yaml

@@ -0,0 +1,200 @@
+---
+apiVersion: v1
+kind: ConfigMap
+metadata:
+  name: grafana-config-job
+  namespace: keycloak
+data:
+  client-role-admin-payload.json: |
+    {"name": "admin"}
+  client-role-editor-payload.json: |
+    {"name": "editor"}
+  client-role-viewer-payload.json: |
+    {"name": "viewer"}
+  admin-role-assignment-payload.json: |
+    [
+      {
+        "id": "$ADMIN_ROLE_ID",
+        "name": "admin"
+      }
+    ]
+  roles-mapper-payload.json: |
+    {
+      "id":"$CLIENT_ROLES_MAPPER_ID",
+      "name": "client roles",
+      "protocol":"openid-connect",
+      "protocolMapper":"oidc-usermodel-client-role-mapper",
+      "config": {
+        "access.token.claim":"true",
+        "claim.name":"resource_access.${client_id}.roles",
+        "jsonType.label":"String",
+        "multivalued":"true",
+        "id.token.claim": "true",
+        "userinfo.token.claim": "true"
+      }
+    }
+  grafana-client-payload.json: |
+    {
+      "protocol": "openid-connect",
+      "clientId": "grafana",
+      "name": "Grafana Client",
+      "description": "Used for Grafana SSO",
+      "publicClient": false,
+      "authorizationServicesEnabled": false,
+      "serviceAccountsEnabled": false,
+      "implicitFlowEnabled": false,
+      "directAccessGrantsEnabled": true,
+      "standardFlowEnabled": true,
+      "frontchannelLogout": true,
+      "attributes": {
+        "saml_idp_initiated_sso_url_name": "",
+        "oauth2.device.authorization.grant.enabled": false,
+        "oidc.ciba.grant.enabled": false
+      },
+      "alwaysDisplayInConsole": false,
+      "rootUrl": "",
+      "baseUrl": "",
+      "redirectUris": [
+        "https://cnoe.localtest.me:8443/grafana/login/generic_oauth"
+      ],
+      "webOrigins": [
+        "/*"
+      ]
+    }
+
+---
+apiVersion: batch/v1
+kind: Job
+metadata:
+  name: grafana-config
+  namespace: keycloak
+spec:
+  template:
+    metadata:
+      generateName: grafana-config
+    spec:
+      serviceAccountName: keycloak-config
+      restartPolicy: Never
+      volumes:
+        - name: keycloak-config
+          secret:
+            secretName: keycloak-config
+        - name: config-payloads
+          configMap:
+            name: grafana-config-job
+      containers:
+        - name: kubectl
+          image: docker.io/library/ubuntu:22.04
+          volumeMounts:
+            - name: keycloak-config
+              readOnly: true
+              mountPath: "/var/secrets/"
+            - name: config-payloads
+              readOnly: true
+              mountPath: "/var/config/"
+          command: ["/bin/bash", "-c"]
+          args:
+            - |
+              #! /bin/bash
+              set -ex -o pipefail
+              apt -qq update && apt -qq install curl jq gettext-base -y
+              
+              curl -sS -LO "https://dl.k8s.io/release/v1.28.3//bin/linux/amd64/kubectl"
+              chmod +x kubectl
+              
+              echo "checking if we're ready to start"
+              set +e
+              ./kubectl get secret -n keycloak keycloak-clients &> /dev/null
+              if [ $? -ne 0 ]; then
+                exit 1
+              fi
+              set -e
+              
+              ADMIN_PASSWORD=$(cat /var/secrets/KEYCLOAK_ADMIN_PASSWORD)
+
+              KEYCLOAK_URL=http://keycloak.keycloak.svc.cluster.local:8080/keycloak
+
+              KEYCLOAK_TOKEN=$(curl -sS  --fail-with-body -X POST -H "Content-Type: application/x-www-form-urlencoded" \
+                --data-urlencode "username=cnoe-admin" \
+                --data-urlencode "password=${ADMIN_PASSWORD}" \
+                --data-urlencode "grant_type=password" \
+                --data-urlencode "client_id=admin-cli" \
+                ${KEYCLOAK_URL}/realms/master/protocol/openid-connect/token | jq -e -r '.access_token')
+                
+              set +e
+              
+              curl --fail-with-body -H "Authorization: bearer ${KEYCLOAK_TOKEN}"  "${KEYCLOAK_URL}/admin/realms/cnoe"  &> /dev/null
+              if [ $? -ne 0 ]; then
+                exit 0
+              fi
+              set -e
+              
+              echo "creating Grafana client"
+              curl -sS -H "Content-Type: application/json" \
+                -H "Authorization: bearer ${KEYCLOAK_TOKEN}" \
+                -X POST --data @/var/config/grafana-client-payload.json \
+                  ${KEYCLOAK_URL}/admin/realms/cnoe/clients
+              
+              CLIENT_ID=$(curl -sS -H "Content-Type: application/json" \
+                -H "Authorization: bearer ${KEYCLOAK_TOKEN}" \
+                -X GET ${KEYCLOAK_URL}/admin/realms/cnoe/clients | jq -e -r  '.[] | select(.clientId == "grafana") | .id')
+              
+              CLIENT_SCOPE_GROUPS_ID=$(curl -sS -H "Content-Type: application/json" -H "Authorization: bearer ${KEYCLOAK_TOKEN}" -X GET  ${KEYCLOAK_URL}/admin/realms/cnoe/client-scopes | jq -e -r  '.[] | select(.name == "groups") | .id')
+              curl -sS -H "Content-Type: application/json" -H "Authorization: bearer ${KEYCLOAK_TOKEN}" -X PUT  ${KEYCLOAK_URL}/admin/realms/cnoe/clients/${CLIENT_ID}/default-client-scopes/${CLIENT_SCOPE_GROUPS_ID}
+              
+              GRAFANA_CLIENT_SECRET=$(curl -sS -H "Content-Type: application/json" \
+                -H "Authorization: bearer ${KEYCLOAK_TOKEN}" \
+                -X GET ${KEYCLOAK_URL}/admin/realms/cnoe/clients/${CLIENT_ID} | jq -e -r '.secret')
+
+              # Add Grafana roles to client
+              curl -sS -H "Content-Type: application/json" \
+                -H "Authorization: bearer ${KEYCLOAK_TOKEN}" \
+                -X POST --data @/var/config/client-role-admin-payload.json \
+                ${KEYCLOAK_URL}/admin/realms/cnoe/clients/${CLIENT_ID}/roles
+              
+              curl -sS -H "Content-Type: application/json" \
+                -H "Authorization: bearer ${KEYCLOAK_TOKEN}" \
+                -X POST --data @/var/config/client-role-editor-payload.json \
+                ${KEYCLOAK_URL}/admin/realms/cnoe/clients/${CLIENT_ID}/roles
+              
+              curl -sS -H "Content-Type: application/json" \
+                -H "Authorization: bearer ${KEYCLOAK_TOKEN}" \
+                -X POST --data @/var/config/client-role-viewer-payload.json \
+                ${KEYCLOAK_URL}/admin/realms/cnoe/clients/${CLIENT_ID}/roles
+
+              export ADMIN_ROLE_ID=$(curl -sS -H "Content-Type: application/json" \
+                -H "Authorization: bearer ${KEYCLOAK_TOKEN}" "${KEYCLOAK_URL}/admin/realms/cnoe/clients/${CLIENT_ID}/roles/admin" | jq -r '.id')
+              
+              # Assign admin role to user1
+              USER1_USERID=$(curl -sS -H "Content-Type: application/json" \
+                -H "Authorization: bearer ${KEYCLOAK_TOKEN}" "${KEYCLOAK_URL}/admin/realms/cnoe/users?lastName=one" | jq -r '.[0].id')
+
+              envsubst < /var/config/admin-role-assignment-payload.json | curl -k -sS -H 'Content-Type: application/json' \
+                -H "Authorization: bearer ${KEYCLOAK_TOKEN}" \
+                -X POST --data @- \
+              ${KEYCLOAK_URL}/admin/realms/cnoe/users/${USER1_USERID}/role-mappings/clients/${CLIENT_ID}
+              
+              # Add role to token
+              CLIENT_SCOPE_ROLES_ID=$(curl -sS -H "Content-Type: application/json" \
+                -H "Authorization: bearer ${KEYCLOAK_TOKEN}" \
+                -X GET ${KEYCLOAK_URL}/admin/realms/cnoe/client-scopes | jq -e -r  '.[] | select(.name == "roles") | .id')
+              
+              export CLIENT_ROLES_MAPPER_ID=$(curl -sS -H "Content-Type: application/json" \
+                -H "Authorization: bearer ${KEYCLOAK_TOKEN}" \
+                -X GET ${KEYCLOAK_URL}/admin/realms/cnoe/client-scopes/${CLIENT_SCOPE_ROLES_ID}/protocol-mappers/models | jq -e -r  '.[] | select(.name == "client roles") | .id')
+
+              cat /var/config/roles-mapper-payload.json | envsubst '$CLIENT_ROLES_MAPPER_ID' | curl -sS -H "Content-Type: application/json" \
+                -H "Authorization: bearer ${KEYCLOAK_TOKEN}" \
+                -X PUT --data @- \
+                ${KEYCLOAK_URL}/admin/realms/cnoe/client-scopes/${CLIENT_SCOPE_ROLES_ID}/protocol-mappers/models/${CLIENT_ROLES_MAPPER_ID}
+              
+              ./kubectl patch secret -n keycloak keycloak-clients --type=json \
+                -p='[{
+                "op" : "add" ,
+                "path" : "/data/GRAFANA_CLIENT_SECRET" ,
+                "value" : "'$(echo -n "$GRAFANA_CLIENT_SECRET" | base64 -w 0)'"
+                },{
+                  "op" : "add" ,
+                  "path" : "/data/GRAFANA_CLIENT_ID" ,
+                  "value" : "'$(echo -n "grafana" | base64 -w 0)'"
+                }]'

observability/prometheus/manifests/grafana-external-secret.yaml

@@ -0,0 +1,20 @@
+apiVersion: external-secrets.io/v1beta1
+kind: ExternalSecret
+metadata:
+  name: keycloak-oidc
+  namespace: monitoring
+spec:
+  secretStoreRef:
+    name: keycloak
+    kind: ClusterSecretStore
+  target:
+    name: grafana-oidc
+  data:
+    - secretKey: GF_AUTH_GENERIC_OAUTH_CLIENT_ID
+      remoteRef:
+        key: keycloak-clients
+        property: GRAFANA_CLIENT_ID
+    - secretKey: GF_AUTH_GENERIC_OAUTH_CLIENT_SECRET
+      remoteRef:
+        key: keycloak-clients
+        property: GRAFANA_CLIENT_SECRET

observability/prometheus/manifests/grafana-ingress.yaml

@@ -0,0 +1,32 @@
+apiVersion: networking.k8s.io/v1
+kind: Ingress
+metadata:
+  name: grafana-ingress
+  namespace: monitoring
+  annotations:
+    nginx.ingress.kubernetes.io/backend-protocol: HTTP
+    nginx.ingress.kubernetes.io/rewrite-target: /grafana/$2
+    nginx.ingress.kubernetes.io/use-regex: 'true'
+spec:
+  ingressClassName: nginx
+  rules:
+    - host: cnoe.localtest.me
+      http:
+        paths:
+          - path: /grafana(/|$)(.*)
+            pathType: ImplementationSpecific
+            backend:
+              service:
+                name: prometheus-grafana
+                port:
+                  number: 80
+    - host: localhost
+      http:
+        paths:
+          - path: /grafana(/|$)(.*)
+            pathType: ImplementationSpecific
+            backend:
+              service:
+                name: prometheus-grafana
+                port:
+                  number: 80

observability/promtail.yaml

@@ -0,0 +1,31 @@
+apiVersion: argoproj.io/v1alpha1
+kind: Application
+metadata:
+  name: promtail
+  namespace: argocd
+  labels:
+    env: dev
+  finalizers:
+    - resources-finalizer.argocd.argoproj.io
+spec:
+  project: default
+  sources:
+    - repoURL: 'https://grafana.github.io/helm-charts'
+      targetRevision: 6.16.0
+      helm:
+        releaseName: promtail
+        values: |
+          config:
+            clients:
+              - url: http://loki-gateway/loki/api/v1/push
+                tenant_id: 1
+      chart: promtail
+  destination:
+    server: "https://kubernetes.default.svc"
+    namespace: monitoring
+  syncPolicy:
+    syncOptions:
+      - CreateNamespace=true
+      - ServerSideApply=true
+    automated:
+      selfHeal: true

ref-implementation/keycloak/manifests/keycloak-config.yaml

@@ -100,7 +100,7 @@ data:
   user-user1.json: |
     {
         "username": "user1",
-        "email": "",
+        "email": "user1@noreply.com",
         "firstName": "user",
         "lastName": "one",
         "requiredActions": [],
@@ -113,7 +113,7 @@ data:
   user-user2.json: |
     {
         "username": "user2",
-        "email": "",
+        "email": "user2@noreply.com",
         "firstName": "user",
         "lastName": "two",
         "requiredActions": [],