Dashboard Configuration

SLOzy ships with Prometheus and Grafana for observability. This page covers how to configure, provision, and customise dashboards.

Grafana Provisioning

The docker-compose.yml mounts provisioning directories for automatic dashboard loading:

volumes:
  - ./grafana/provisioning:/etc/grafana/provisioning
  - ./grafana/dashboards:/var/lib/grafana/dashboards

Grafana loads datasources and dashboards from these mounted paths at startup. To add a new dashboard, place a JSON file in grafana/dashboards/ and restart Grafana.

Datasource Provisioning

Create grafana/provisioning/datasources/prometheus.yaml:

apiVersion: 1
datasources:
  - name: Prometheus
    type: prometheus
    access: proxy
    url: http://prometheus:9090
    isDefault: true

Dashboard Provisioning

Create grafana/provisioning/dashboards/slozy.yaml:

apiVersion: 1
providers:
  - name: SLOzy
    type: file
    options:
      path: /var/lib/grafana/dashboards

Recommended Dashboards

1. SLO Burn-Rate Overview

Track error budget consumption and burn rate for all SLOs.

Panel	Query	Description
SLO Status	slo_status	Current status per SLO (OK / warning / critical)
Error Budget	1 - (slo_error_budget_consumed)	Remaining error budget percentage
Burn Rate	rate(slo_error_budget_consumed[5m])	Current rate of budget consumption
Time to Exhaustion	slo_error_budget_remaining / rate(slo_error_budget_consumed[1h])	Hours until budget depleted

2. Application Performance

Monitor request rate, latency, and error rate for the SLOzy API.

Panel	Query	Description
Request Rate	rate(http_requests_total[5m])	Requests per second
Latency P95	histogram_quantile(0.95, rate(http_request_duration_seconds_bucket[5m]))	95th percentile latency
Latency P99	histogram_quantile(0.99, rate(http_request_duration_seconds_bucket[5m]))	99th percentile latency
Error Rate	rate(http_requests_total{status=~"5.."}[5m]) / rate(http_requests_total[5m])	Percentage of 5xx responses

3. Database

Panel	Query	Description
Active Connections	pg_stat_database_numbackends	Current connections to slozy_db
Cache Hit Ratio	rate(pg_stat_database_blks_hit[5m]) / (rate(pg_stat_database_blks_hit[5m]) + rate(pg_stat_database_blks_read[5m]))	Buffer cache efficiency

4. Host & Containers (pre-installed)

A comprehensive infrastructure dashboard showing host and container metrics from node_exporter and cAdvisor. Located at grafana/dashboards/host-overview.json.

Panel	Query	Description
CPU Usage	100 - avg(rate(node_cpu_seconds_total{mode="idle"}[$__rate_interval])) * 100	Host CPU utilisation gauge
Memory Usage	(1 - node_memory_MemAvailable_bytes / node_memory_MemTotal_bytes) * 100	Host RAM utilisation gauge
Disk Usage	(1 - node_filesystem_free_bytes{mountpoint="/"} / ...) * 100	Root partition usage gauge
Uptime	time() - node_boot_time_seconds	Host uptime
Container CPU	sum(rate(container_cpu_usage_seconds_total{...} [$__rate_interval])) by (container_label_com_docker_compose_service)	Per-container CPU
Container Memory	sum(container_memory_usage_bytes{...}) by (container_label_com_docker_compose_service)	Per-container memory
Container Network	sum(rate(container_network_receive_bytes_total{...} [$__rate_interval])) by (container_label_com_docker_compose_service)	Per-container network traffic

Alerting from Grafana

Grafana can send alerts to Slack, email, Telegram, or PagerDuty. Configure notification channels in:

• UI: Configuration > Notification Channels
• Provisioning: grafana/provisioning/notifiers/

Example Slack notifier:

apiVersion: 1
notifiers:
  - name: Slack
    type: slack
    uid: slack-notifier
    settings:
      url: https://hooks.slack.com/services/YOUR/WEBHOOK/URL

Access

• Development: Grafana runs at http://localhost:3001 (default credentials admin / admin)
• Production: Expose Grafana behind Nginx with TLS and authentication. Consider using Grafana's built-in OAuth or LDAP integration.

The SLOzy application itself does not serve dashboards — all visualisation is done through Grafana using Prometheus as the data source. To customise the dashboards, edit the JSON files in grafana/dashboards/ and redeploy.