Skip to content

Node Operators

This section covers the official Koinos node composed of microservices and distributed by koinos/koinos. Choose what the node must do before enabling optional services. The standard starting point is a node without block production, with private local JSON-RPC for health checks.

The operating procedures were verified on 2026-07-25 against koinos/koinos commit 8216746. That revision is a recorded commit, not a new immutable release. See Configuration before installing or upgrading.

Choose what the node will do

Purpose Services Public exposure Main risk Start here
Standard node Required services; optional private JSON-RPC P2P 8888/tcp Disk growth and synchronization Run a Koinos node
Public API node Standard node plus selected API services HTTPS through a reverse proxy Abuse, capacity, and incorrect port exposure Run a public API node
Block producer Standard node plus block_producer P2P; APIs need not be public Hot key and irreversible PoB actions Block production

Do not use all as a generic quick start

The upstream all profile includes block_producer. Enable only the profiles required for the selected role.

  • Standard node quick start


    Prepare an Ubuntu LTS host, select a release or commit, start only the required services, and verify synchronization and P2P activity.

    Run a Koinos node

  • Requirements


    Plan CPU, RAM, SSD capacity, bandwidth, time synchronization, and disk headroom for the selected services.

    Plan capacity

  • Networks


    Keep mainnet, the current public testnet, and legacy Harbinger data, identities, keys, and chain IDs separate.

    Choose a network

  • Public API node


    Enable JSON-RPC, REST, and gRPC intentionally, with loopback bindings, clear reverse-proxy requirements, TLS, and external verification.

    Publish APIs

  • Security


    Apply least-service deployment, binding and firewall controls, proxy protections, host hardening, and key separation.

    Secure the node

  • Operations


    Monitor, update, roll back, back up, restore, reindex, resync, and collect useful incident evidence.

    Operate the node