Skip to content

grafana-setup.md

Latest commit

 

History

History
145 lines (103 loc) · 4.61 KB

grafana-setup.md

File metadata and controls

145 lines (103 loc) · 4.61 KB

Grafana Setup for Torii Metrics

This guide explains how to set up Grafana to visualize Torii indexer metrics using Docker Compose.

Prerequisites

  • Docker and Docker Compose installed
  • Torii running with metrics enabled
  • Access to the Torii metrics endpoint (default: localhost:9200)

Quick Start

  1. Start Torii with metrics enabled:

    torii --metrics --metrics.addr 0.0.0.0 --metrics.port 9200

  2. Start Grafana and Prometheus:

    docker-compose -f docker-compose.grafana.yml up -d

  3. Access Grafana:

    • Open http://localhost:3000 in your browser
    • Login with username: admin, password: admin
    • Navigate to the "Torii Indexer Overview" dashboard

Architecture

The setup includes:

  • Prometheus (port 9090): Scrapes metrics from Torii
  • Grafana (port 3000): Visualizes metrics with pre-configured dashboards

Dashboard Overview

The "Torii Indexer Overview" dashboard provides comprehensive monitoring of:

Indexer Engine Metrics

  • Fetch Duration: Time taken to fetch data from the blockchain
  • Process Duration: Time taken to process fetched events
  • Fetch/Process Rates: Success and error rates for operations
  • Backoff Delays: Current backoff delays when errors occur
  • Events Processed: Rate of events processed by contract type

Fetcher Metrics

  • RPC Duration: Time taken for individual RPC calls
  • RPC Request Rates: Success and error rates for RPC requests
  • Throughput: Events fetched and blocks processed rates
  • Error Rates: Combined error rates across all operations

Configuration

Prometheus Configuration

The Prometheus configuration (grafana/prometheus.yml) scrapes metrics from Torii:

scrape_configs:
  - job_name: 'torii-indexer'
    static_configs:
      - targets: ['host.docker.internal:9200']
    scrape_interval: 5s

Grafana Provisioning

Grafana is automatically configured with:

  • Prometheus as the default data source
  • Pre-built dashboards in the "Torii" folder
  • Admin user with password "admin"

Customization

Adding New Dashboards

  1. Create new dashboard JSON files in grafana/dashboards/
  2. Restart Grafana: docker-compose -f docker-compose.grafana.yml restart grafana

Modifying Data Sources

Edit grafana/provisioning/datasources/prometheus.yml to add or modify data sources.

Changing Grafana Settings

Modify environment variables in docker-compose.grafana.yml:

environment:
  - GF_SECURITY_ADMIN_PASSWORD=your_password
  - GF_USERS_ALLOW_SIGN_UP=false
  - GF_SERVER_ROOT_URL=http://your-domain.com

Troubleshooting

Grafana Can't Connect to Prometheus

  1. Ensure Prometheus is running: docker-compose -f docker-compose.grafana.yml logs prometheus
  2. Check Prometheus targets: http://localhost:9090/targets
  3. Verify Torii metrics endpoint: http://localhost:9200/metrics

No Data in Dashboards

  1. Verify Torii is running with metrics enabled
  2. Check that metrics are being exposed: curl http://localhost:9200/metrics
  3. Ensure Prometheus is scraping successfully: http://localhost:9090/targets

Dashboard Not Loading

  1. Check Grafana logs: docker-compose -f docker-compose.grafana.yml logs grafana
  2. Verify dashboard files are in the correct location
  3. Restart Grafana to reload provisioning

Metrics Reference

Indexer Engine Metrics

Metric Type Description
torii_indexer_fetch_duration_seconds Histogram Time taken to fetch data
torii_indexer_process_duration_seconds Histogram Time taken to process events
torii_indexer_fetch_total Counter Total fetch operations
torii_indexer_process_total Counter Total process operations
torii_indexer_backoff_delay_seconds Gauge Current backoff delay
torii_indexer_events_processed_total Counter Events processed by type
torii_indexer_errors_total Counter Errors by operation

Fetcher Metrics

Metric Type Description
torii_fetcher_rpc_duration_seconds Histogram RPC call duration
torii_fetcher_rpc_requests_total Counter RPC requests by status
torii_fetcher_events_fetched_total Counter Total events fetched
torii_fetcher_blocks_to_fetch_total Counter Blocks queued for fetching
torii_fetcher_errors_total Counter Fetcher errors by operation

Production Considerations

  • Change default Grafana admin password
  • Configure proper authentication and authorization
  • Set up persistent storage for Grafana data
  • Configure alerting for critical metrics
  • Use external Prometheus for production deployments