stack_orchestrator/data/compose/docker-compose-grafana.yml

@@ -6,10 +6,16 @@ services:
     restart: always
     environment:
       GF_SERVER_ROOT_URL: ${GF_SERVER_ROOT_URL}
+      CERC_GRAFANA_ALERTS_SUBGRAPH_IDS: ${CERC_GRAFANA_ALERTS_SUBGRAPH_IDS}
     volumes:
       - ../config/monitoring/grafana/provisioning:/etc/grafana/provisioning
       - ../config/monitoring/grafana/dashboards:/etc/grafana/dashboards
+      - ../config/monitoring/update-grafana-alerts-config.sh:/update-grafana-alerts-config.sh
       - grafana_storage:/var/lib/grafana
+    user: root
+    entrypoint: ["bash", "-c"]
+    command: |
+      "/update-grafana-alerts-config.sh && /run.sh"
     ports:
       - "3000"
     healthcheck:

stack_orchestrator/data/config/monitoring/subgraph-alert-rules.yml

@@ -0,0 +1,64 @@
+apiVersion: 1
+groups:
+  - orgId: 1
+    name: subgraph
+    folder: SubgraphAlerts
+    interval: 30s
+    rules:
+      - uid: b2a9144b-6104-46fc-92b5-352f4e643c4c
+        title: subgraph_head_tracking
+        condition: condition
+        data:
+          - refId: diff
+            relativeTimeRange:
+              from: 600
+              to: 0
+            datasourceUid: PBFA97CFB590B2093
+            model:
+              datasource:
+                  type: prometheus
+                  uid: PBFA97CFB590B2093
+              editorMode: code
+              expr: ethereum_chain_head_number - on(network) group_right deployment_head{deployment=~"REPLACE_WITH_SUBGRAPH_IDS"}
+              instant: true
+              intervalMs: 1000
+              legendFormat: __auto
+              maxDataPoints: 43200
+              range: false
+              refId: diff
+          - refId: condition
+            relativeTimeRange:
+              from: 600
+              to: 0
+            datasourceUid: __expr__
+            model:
+              conditions:
+                  - evaluator:
+                      params:
+                          - 15
+                          - 0
+                      type: gt
+                    operator:
+                      type: and
+                    query:
+                      params: []
+                    reducer:
+                      params: []
+                      type: avg
+                    type: query
+              datasource:
+                  name: Expression
+                  type: __expr__
+                  uid: __expr__
+              expression: diff
+              intervalMs: 1000
+              maxDataPoints: 43200
+              refId: condition
+              type: threshold
+        noDataState: OK
+        execErrState: Alerting
+        for: 5m
+        annotations:
+          summary: Subgraph deployment {{ index $labels "deployment" }} is falling behind head by {{ index $values "diff" }}
+        labels: {}
+        isPaused: false

stack_orchestrator/data/config/monitoring/update-grafana-alerts-config.sh

@@ -0,0 +1,7 @@
+#!/bin/bash
+
+echo Using CERC_GRAFANA_ALERTS_SUBGRAPH_IDS ${CERC_GRAFANA_ALERTS_SUBGRAPH_IDS}
+
+# Replace subgraph ids in subgraph alerting config
+# Note: Requires the grafana container to be run with user root
+sed -i "s/REPLACE_WITH_SUBGRAPH_IDS/$CERC_GRAFANA_ALERTS_SUBGRAPH_IDS/g" /etc/grafana/provisioning/alerting/subgraph-alert-rules.yml

stack_orchestrator/data/stacks/monitoring/monitoring-watchers.md

@@ -113,21 +113,31 @@ Add the following scrape configs to prometheus config file (`monitoring-watchers
         labels:
           instance: 'ajna'
           chain: 'filecoin'
+
+  - job_name: graph-node
+    metrics_path: /metrics
+    scrape_interval: 30s
+    static_configs:
+      - targets: ['GRAPH_NODE_HOST:GRAPH_NODE_HOST_METRICS_PORT']
   ```
 
 Add scrape config as done above for any additional watcher to add it to the Watchers dashboard.
 
 ### Grafana alerts config
 
-Place the pre-configured watcher alerts rules in Grafana provisioning directory:
+Place the pre-configured alerts rules in Grafana provisioning directory:
 
   ```bash
+  # watcher alert rules
   cp monitoring-watchers-deployment/config/monitoring/watcher-alert-rules.yml monitoring-watchers-deployment/config/monitoring/grafana/provisioning/alerting/
+
+  # subgraph alert rules
+  cp monitoring-watchers-deployment/config/monitoring/subgraph-alert-rules.yml monitoring-watchers-deployment/config/monitoring/grafana/provisioning/alerting/
   ```
 
 Update the alerting contact points config (`monitoring-watchers-deployment/config/monitoring/grafana/provisioning/alerting/contactpoints.yml`) with desired contact points
 
-Add corresponding routes to the notification policies config (`monitoring-watchers-deployment/monitoring/grafana/provisioning/alerting/policies.yaml`) with appropriate object-matchers:
+Add corresponding routes to the notification policies config (`monitoring-watchers-deployment/config/monitoring/grafana/provisioning/alerting/policies.yml`) with appropriate object-matchers:
 
   ```yml
   ...
@@ -135,7 +145,7 @@ Add corresponding routes to the notification policies config (`monitoring-watche
       - receiver: SlackNotifier
       object_matchers:
         # Add matchers below
-        - ['grafana_folder', '=', 'WatcherAlerts']
+        - ['grafana_folder', '=~', 'WatcherAlerts|SubgraphAlerts']
   ```
 
 ### Env
@@ -149,6 +159,9 @@ Set the following env variables in the deployment env config file (`monitoring-w
   # Grafana server host URL to be used
   # (Optional, default: http://localhost:3000)
   GF_SERVER_ROOT_URL=
+
+  # List of subgraph ids to configure alerts for (separated by |)
+  CERC_GRAFANA_ALERTS_SUBGRAPH_IDS=
   ```
 
 ## Start the stack