Liferay PaaS Experience: Metrics
Grafana is the metrics interface for Liferay PaaS Experience environments. Liferay provisions and manages the full metrics stack: Google Managed Prometheus (GMP) collects metrics from the cluster, and Grafana provides pre-built dashboards for visualizing them. No additional monitoring infrastructure is required.
Pre-Built Dashboards
Grafana includes a set of pre-built dashboards in the Liferay folder, which you reach from Grafana’s dashboard list. The Central Hub dashboard opens by default when you log in, and links directly to the Kubernetes Workloads, Backups, Databases, and Load Balancers dashboards. Builds and Secrets are dashboards specific to the managed PaaS Experience platform. They don’t exist in Liferay’s self-hosted observability chart, because a self-hosted team already has direct visibility into builds and secrets through its own tooling.
| Dashboard | What It Shows |
|---|---|
| Central Hub | Links to Argo CD, Argo Workflows, your GitOps Git repository, the alerting list, the Kubernetes Workloads, Backups, Databases, and Load Balancers dashboards, the product documentation, and the Liferay support portal. It links to each area rather than reproducing its data inline. |
| Builds | Build ID and status, a link to the commit that triggered the build, and searchable, downloadable build logs |
| Kubernetes Workloads | Liferay pod and container metrics, including JVM and search metrics |
| Databases | Database CPU usage, memory usage, disk usage, and instance status |
| Load Balancers | Request rate, latency (P95), ingress and egress throughput, and response status breakdown |
| Backups | Backup run counts, success and failure rates |
| Secrets | A read-only list of your secrets and an audit trail of create, update, and delete operations performed on them. You manage secrets through Argo CD; see Managing Secrets. |
Metrics Sources
The metrics stack collects from several sources in the cluster:
| Source | What It Collects |
|---|---|
Liferay JVM (port 12345) | JVM thread counts, Hikari connection pool activity, Tomcat session metrics, and other runtime metrics exposed via the JMX Prometheus agent |
| cAdvisor | Container CPU and memory usage at the container runtime level |
| kube-state-metrics | Kubernetes object state: deployment readiness, pod restart counts, resource requests and limits |
| Node Exporter | Node-level CPU, memory, disk, and network metrics |
| Envoy | Ingress request rates, latency, and error rates from the Envoy Gateway |
The scrape interval for all sources is one minute. Liferay manages the metrics collection configuration for these sources as part of the platform.