Last updated: March 16, 2026
Grafana
URL: http://109.199.120.120:3000
Credentials: admin / coderz123
Grafana is the central visualization layer of the Coderz Stack. It connects to multiple data sources — Prometheus (metrics) and Loki (logs) — and provides pre-built dashboards for every component in the stack.
What It Does
- Displays real-time and historical metrics from all services
- Provides searchable log viewers (via Loki)
- Fires alerts when thresholds are exceeded
- Sends alert notifications via email
Pre-Provisioned Dashboards
All dashboards are automatically loaded when the stack starts — no manual setup needed.
| Dashboard | Description |
|---|
| Coderz Stack Home | All-in-one overview: CPU, RAM, Disk, top Docker containers |
| API Traffic Monitor | High-traffic API visibility: per-endpoint req/min, success %, 4xx/5xx, latency P95/P99, live error logs — built for 1M+ req/min |
| Server Overview | Node Exporter: CPU by mode, load average, RAM, disk, network, processes, FDs, TCP |
| Docker Containers | cAdvisor: per-container CPU, memory, network I/O |
| Container Logs | Loki: live log stream from all Docker containers |
| .NET API Full Stack | Prometheus request metrics + PostgreSQL stats + live request log table |
| Prefect Flows | Loki-based: flow completions, failures, alert log, live flow logs |
| Alerts Status | CPU/RAM/Disk stat panels + timeseries alert history |
| Kubernetes Overview | k3s pods, nodes, deployments via kube-state-metrics |
Alert Rules
Grafana evaluates these alert rules continuously:
| Alert | Condition | Evaluation Period |
|---|
| High CPU | CPU usage > 80% | 5 minutes |
| High RAM | Memory usage > 85% | 5 minutes |
| High Disk | Disk usage > 90% | 5 minutes |
| Container Count Low | Running containers < 6 | 2 minutes |
aboodm7med1995@gmail.com via Postfix SMTP.
Data Sources
| Source | Type | URL |
|---|
| Prometheus | Prometheus | http://prometheus:9090 |
| Loki | Loki | http://loki:3100 |
Adding a New Dashboard
- Go to Dashboards → New → New Dashboard
- Add panels using PromQL (for metrics) or LogQL (for logs)
- Save the dashboard
- To persist it across restarts, export as JSON and place in
/opt/coderz/configs/grafana/provisioning/dashboards/
Useful PromQL Queries
# CPU usage %
100 - (avg by(instance) (rate(node_cpu_seconds_total{mode="idle"}[5m])) * 100)
# RAM usage %
(1 - node_memory_MemAvailable_bytes / node_memory_MemTotal_bytes) * 100
# Disk usage %
(1 - node_filesystem_free_bytes / node_filesystem_size_bytes) * 100
# HTTP request rate (.NET API)
rate(http_requests_received_total[1m])
# HTTP error rate
rate(http_requests_received_total{code=~"5.."}[1m])