Skip to main content
Version: Next


HOTROD is a distributed data processing system, composed mostly of three components: Server, Agent, and Pipe.


The server component and central part of a deployment which manages all Agents and Pipes via a web-based UI, HTTP API, or CLI.

$> hotrod run server

The server component that manages a deployment:

  • keeps track of one or more Agents
  • keeps track of Pipes
  • provides user management and authentication


A data processing service that is managed by the Server and deployed by Agents.

Pipes are used to instruct data which is then processed by a Pipe definition — written using YAML formatted Pipe Language.


A substrate on which Pipes can be attached (deployed). Agents are logical identifiers created on the Server — each with a unique name and unique Agent ID.

$> hotrod run agent

The client component that manages Pipes:

  • queries the Server for new, removed and changed Pipes
  • uses that information to deploy, remove, or update Pipes
  • sends logs and metrics data to the Server


$> hotrod

A shell (bash, zsh, fish et al.) utility for various administration tasks. A wrapper around the HTTP API.



The web-based UI (BUI) for administration.