Cluster Monitoring¶
This page describes monitoring options for the Launchpad cluster.
Overview¶
When enabled via Harmony/Terraform, the stack can include:
- Prometheus - Metrics collection from the cluster and applications
- Grafana - Dashboards and visualization
- Alertmanager - Alert routing and notifications
Configuration is typically done through Terraform variables (e.g. prometheus_enabled, grafana_enabled, alertmanager_config).
Accessing Grafana¶
If Grafana is enabled, it is usually exposed at a hostname such as grafana.<cluster-domain>. Check the Terraform outputs or ingress resources:
kubectl get ingress -A | grep grafana
Use the credentials configured in Terraform or stored in a Kubernetes secret (e.g. in the same namespace as Grafana).
Prometheus and Alertmanager¶
- Prometheus scrapes metrics from cluster and application targets. Queries can be run in Grafana or via the Prometheus UI if exposed.
- Alertmanager receives alerts from Prometheus and sends notifications (e.g. email, Slack). Configure receivers and routes in the Terraform/Alertmanager config.
Open edX Instance Metrics¶
Instance-level monitoring (e.g. LMS, Celery, Redis) may be documented under Instances → Monitoring. Ensure service monitors or scrape configs include the instance namespaces if you want them in cluster Prometheus.
Related Documentation¶
- Cluster Overview - Cluster operations
- Infrastructure Overview - Harmony and optional monitoring stack
- Cluster Configuration - Terraform and monitoring variables
- Instance Monitoring - Instance-level monitoring
See Also¶
- Backup - Backup and retention
- Restore - Restore procedures
- User Guides Overview - Task-based guides