Operations

Running nodes in production.

Servers and services

• Multi-node server setup — OS, firewall, data disks and ports for several nodes on one server
• systemd service template — run nodes and tools as services
• Configure journald — keep logs from filling the disk
• Caddy reverse proxy — HTTPS in front of RPC, REST and gRPC

Node configuration

• Pruning — limit disk usage of application state
• Transaction indexer on PostgreSQL — index events into a SQL database

Monitoring

• Node monitoring with Prometheus — Node Exporter and CometBFT metrics
• PANIC for Cosmos — Telegram alerts for missed blocks and lost peers