Dashboards

Grafana dashboards are pre-configured for monitoring Laminar pipelines and infrastructure.

Datasources

Grafana has three pre-configured datasources:

Name	Type	Endpoint	Purpose
Laminar-Metrics	Prometheus	`laminar-engine:8004`	Laminar engine metrics
GrepTimeDB-Metrics	Prometheus	`greptimedb:4000/v1/prometheus`	Container/node/K8s metrics
GrepTimeDB-Logs	MySQL	`greptimedb:4002`	Log queries

Dashboard Provisioning

Dashboards are provisioned from ConfigMaps with the label grafana_dashboard: "1".

The Grafana sidecar watches all namespaces for ConfigMaps with this label and automatically loads them.

Adding Custom Dashboards

apiVersion: v1
kind: ConfigMap
metadata:
  name: my-custom-dashboard
  namespace: laminar-monitoring
  labels:
    grafana_dashboard: "1"
  annotations:
    grafana_folder: "Custom"
data:
  my-dashboard.json: |
    {
      "title": "My Dashboard",
      "panels": [...]
    }

Built-in Dashboards

Dashboards are provisioned from the laminar-dashboards ConfigMap in the "Laminar" folder.

Laminar Logs

UID: laminar-logs Datasource: GrepTimeDB-Logs (MySQL)

Displays application logs from GrepTimeDB.

Panel	Type	Query
Errors (1h)	Stat	`SELECT COUNT(*) FROM logs WHERE level = 'error' AND greptime_timestamp > NOW() - INTERVAL 1 HOUR`
Warnings (1h)	Stat	`SELECT COUNT(*) FROM logs WHERE level = 'warn' AND greptime_timestamp > NOW() - INTERVAL 1 HOUR`
Total Logs (1h)	Stat	`SELECT COUNT(*) FROM logs WHERE greptime_timestamp > NOW() - INTERVAL 1 HOUR`
Recent Logs	Table	Last 100 logs with timestamp, level, target, message
Error Logs (24h)	Table	Last 50 ERROR level logs with fields

Laminar Metrics

UID: laminar-metrics Datasource: GrepTimeDB-Metrics (Prometheus)

Displays Laminar engine metrics.

Panel	Type	Query
Active Pipelines	Time Series	`laminar_controller_active_pipelines`
Slot Utilization	Time Series	`laminar_controller_registered_slots` and `laminar_controller_free_slots`
Registered Nodes	Stat	`laminar_controller_registered_nodes`
Total Slots	Stat	`laminar_controller_registered_slots`
Free Slots	Stat	`laminar_controller_free_slots`
Active Pipelines	Stat	`laminar_controller_active_pipelines`
Worker Bytes	Time Series	`sum(laminar_worker_bytes_recv) by (operator_name)`
Worker Messages	Time Series	`sum(laminar_worker_messages_recv) by (operator_name)`

Kubernetes Cluster Overview

UID: k8s-cluster-overview Datasource: GrepTimeDB-Metrics (Prometheus)

High-level view of the Kubernetes cluster.

Panel	Type	Query
Total Nodes	Stat	`count(kube_node_info)`
Running Pods	Stat	`sum(kube_pod_status_phase{phase="Running"})`
Pending Pods	Stat	`sum(kube_pod_status_phase{phase="Pending"})`
Failed Pods	Stat	`sum(kube_pod_status_phase{phase="Failed"})`
Cluster CPU Usage	Time Series	`1 - avg(rate(node_cpu_seconds_total{mode="idle"}[5m]))`
Cluster Memory Usage	Time Series	`sum(node_memory_total_bytes) - sum(node_memory_available_bytes)`
Pods by Phase	Time Series	`sum by (phase) (kube_pod_status_phase)`
Pod Restarts (1h)	Table	Top 10 pods by restart count

Kubernetes Node Resources

UID: k8s-node-resources Datasource: GrepTimeDB-Metrics (Prometheus) Variables: $node (multi-select, default: All)

Detailed resource metrics per node.

Panel	Type	Query
Node CPU Usage	Time Series	`1 - avg by(node) (rate(node_cpu_seconds_total{mode="idle"}[5m]))`
Node Memory Usage	Time Series	`1 - (node_memory_available_bytes / node_memory_total_bytes)`
Node Disk Usage (/)	Time Series	`1 - (node_filesystem_free_bytes / node_filesystem_total_bytes)`
Node Network I/O	Time Series	`rate(node_network_receive_bytes_total[5m])`
Node Load Average	Time Series	`node_load1`, `node_load5`, `node_load15`
CPU Requests / Allocatable	Time Series	CPU requests vs allocatable ratio
Memory Requests / Allocatable	Time Series	Memory requests vs allocatable ratio
Node Resource Requests vs Allocatable	Table	Summary table with CPU/Memory requests, allocatable, available, and percentages

Kubernetes Pod Resources

UID: k8s-pod-resources Datasource: GrepTimeDB-Metrics (Prometheus) Variables: $namespace, $pod (multi-select, default: All)

Per-pod resource metrics.

Panel	Type	Query
Pod CPU Usage vs Requests	Time Series	`sum by(pod) (rate(container_cpu_usage_seconds_total[5m]))`
Pod Memory Usage vs Limits	Time Series	`sum by(pod) (container_memory_working_set_bytes)`
Container Restarts (1h)	Time Series	`sum by(pod, container) (increase(kube_pod_container_status_restarts_total[1h]))`
Pod Status	Table	Current pod phase by namespace and pod
Pod Network I/O	Time Series	`rate(container_network_receive_bytes_total[5m])`

Access

Desktop Mode

Grafana is available at: http://localhost:3001

Credentials: laminar / laminar

Kubernetes

Grafana is available at: https://grafana.{baseDomain}

Credentials: laminar / laminar

Creating Custom Dashboards

Using Grafana UI

Navigate to Grafana → Dashboards → New Dashboard
Add panels using the available datasources
Save and export as JSON

Using ConfigMaps (Kubernetes)

Export dashboard JSON from Grafana
Create a ConfigMap with the grafana_dashboard: "1" label
Apply to the cluster

kubectl create configmap my-dashboard \
  --from-file=dashboard.json \
  -n laminar-monitoring
 
kubectl label configmap my-dashboard grafana_dashboard=1 \
  -n laminar-monitoring