feat(repo): Added metrics for publisher and dashboards

FuelLabs · Jan 22, 2025 · 1c7f41e · 1c7f41e
1 parent fc5ead3
commit 1c7f41e
Show file tree

Hide file tree

Showing 21 changed files with 1,918 additions and 108 deletions.
diff --git a/Makefile b/Makefile
@@ -225,13 +225,12 @@ bench:
 
 run-publisher: NETWORK="testnet"
 run-publisher: MODE="dev"
-run-publisher: PORT="4000"
-run-publisher: TELEMETRY_PORT="9001"
+run-publisher: PORT="9001"
 run-publisher: NATS_URL="localhost:4222"
 run-publisher: EXTRA_ARGS=""
 run-publisher: FROM_HEIGHT="0"
 run-publisher: check-network
-	@./scripts/run_publisher.sh
+	@./scripts/run_publisher.sh --mode $(MODE) --network $(NETWORK) --port $(PORT) --from-height $(FROM_HEIGHT) --extra-args $(EXTRA_ARGS)
 
 run-publisher-mainnet-dev:
 	$(MAKE) run-publisher NETWORK=mainnet MODE=dev FROM_HEIGHT=0
@@ -285,7 +284,7 @@ run-webserver-testnet-profiling:
 # ------------------------------------------------------------
 
 # Define service profiles
-DOCKER_SERVICES := nats docker cockroach
+DOCKER_SERVICES := nats docker cockroach monitoring
 
 run-docker-compose: PROFILE="all"
 run-docker-compose:

diff --git a/cluster/docker/.dockerignore b/cluster/docker/.dockerignore
@@ -0,0 +1 @@
+/target
diff --git a/cluster/docker/docker-compose.yml b/cluster/docker/docker-compose.yml
@@ -8,10 +8,13 @@ services:
     restart: always
     ports:
       - 4222:4222
+      - 8222:8222
     volumes:
       - ./nats-config/core.conf:/etc/nats/nats.conf
       - ./nats-config/accounts.conf:/etc/nats/accounts.conf
     command:
+      - -m
+      - '8222'
       - --name=fuel-streams-nats
       - --js
       - --config=/etc/nats/nats.conf
@@ -24,6 +27,7 @@ services:
       - all
       - cockroach
     image: cockroachdb/cockroach:v23.2.2
+    container_name: cockroachdb
     ports:
       - "26257:26257"
       - "8080:8080"
@@ -36,5 +40,106 @@ services:
       timeout: 5s
       retries: 5
 
+  prometheus:
+      profiles:
+        - all
+        - monitoring
+      image: prom/prometheus:latest
+      container_name: prometheus
+      restart: always
+      volumes:
+        - ./monitoring/prometheus/:/etc/prometheus/
+        - prometheus_data:/prometheus
+      command:
+        - '--config.file=/etc/prometheus/prometheus.yml'
+        - '--storage.tsdb.path=/prometheus'
+        - '--web.console.libraries=/usr/share/prometheus/console_libraries'
+        - '--web.console.templates=/usr/share/prometheus/consoles'
+        - '--storage.tsdb.retention.time=200h'
+        - '--web.enable-lifecycle'
+      ports:
+        - 9090:9090
+
+  node-exporter:
+    profiles:
+      - all
+      - monitoring
+    image: prom/node-exporter:latest
+    container_name: node-exporter
+    restart: always
+    volumes:
+      - /proc:/host/proc:ro
+      - /sys:/host/sys:ro
+      - /:/rootfs:ro
+    command:
+      - '--path.procfs=/host/proc'
+      - '--path.sysfs=/host/sys'
+      - --collector.filesystem.ignored-mount-points
+      - "^/(sys|proc|dev|host|etc|rootfs/var/lib/docker/containers|rootfs/var/lib/docker/overlay2|rootfs/run/docker/netns|rootfs/var/lib/docker/aufs)($$|/)"
+    ports:
+      - 9100:9100
+    depends_on:
+      - prometheus
+    deploy:
+      mode: global
+
+  cadvisor:
+    profiles:
+      - all
+      - monitoring
+    image: gcr.io/cadvisor/cadvisor
+    container_name: cadvisor
+    restart: always
+    volumes:
+      - /:/rootfs:ro
+      - /var/run:/var/run:rw
+      - /sys:/sys:ro
+      - /var/lib/docker/:/var/lib/docker:ro
+    ports:
+      - 9200:8080
+    depends_on:
+      - prometheus
+    deploy:
+      mode: global
+
+  alertmanager:
+    profiles:
+      - all
+      - monitoring
+    image: prom/alertmanager:latest
+    container_name: alertmanager
+    restart: always
+    volumes:
+      - ./monitoring/alertmanager/:/etc/alertmanager/
+    command:
+      - '--config.file=/etc/alertmanager/config.yml'
+      - '--storage.path=/alertmanager'
+    depends_on:
+      - prometheus
+    ports:
+      - 9093:9093
+
+  grafana:
+    profiles:
+      - all
+      - monitoring
+    image: grafana/grafana:latest
+    container_name: grafana
+    user: '472'
+    restart: always
+    environment:
+      GF_INSTALL_PLUGINS: "grafana-clock-panel,grafana-simple-json-datasource,grafana-worldmap-panel,grafana-piechart-panel"
+    volumes:
+      - grafana_data:/var/lib/grafana
+      - ./monitoring/grafana/provisioning/:/etc/grafana/provisioning/
+    env_file:
+      - ./monitoring/grafana/config.monitoring
+    ports:
+      - 3000:3000
+    depends_on:
+      - prometheus
+
 volumes:
   cockroach_data:
+  prometheus_data:
+  grafana_data:
diff --git a/cluster/docker/monitoring/alertmanager/config.yml b/cluster/docker/monitoring/alertmanager/config.yml
@@ -0,0 +1,11 @@
+route:
+  receiver: 'slack'
+
+receivers:
+  - name: 'slack'
+
+#       slack_configs:
+#           - send_resolved: true
+#             username: '<username>'
+#             channel: '#<channel-name>'
+#             api_url: '<incomming-webhook-url>'
diff --git a/cluster/docker/monitoring/grafana/config.monitoring b/cluster/docker/monitoring/grafana/config.monitoring
@@ -0,0 +1,9 @@
+GF_SECURITY_ADMIN_USER=admin
+GF_SECURITY_ADMIN_PASSWORD=admin
+GF_USERS_ALLOW_SIGN_UP=false
+GF_SERVER_DOMAIN=myrul.com
+GF_SMTP_ENABLED=true
+GF_SMTP_HOST=smtp.gmail.com:587
+[email protected]
+GF_SMTP_PASSWORD=mypassword
+[email protected]