Skip to content
Open
Changes from 1 commit
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
18 changes: 18 additions & 0 deletions backend/README.md
Original file line number Diff line number Diff line change
Expand Up @@ -5,6 +5,24 @@
This directory contains code for the components that comprise the Kubeflow
Pipelines backend.

## Backend Component Overview

The Kubeflow Pipelines backend is composed of several cooperating
components that manage pipeline execution and lifecycle:

- **API Server** – Exposes REST and gRPC endpoints for managing
pipelines, runs, and experiments.
- **Driver** – Translates pipeline specifications into executable
workflow definitions.
Copy link

Copilot AI Feb 21, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The Driver description as "Translates pipeline specifications into executable workflow definitions" may be imprecise. Based on the code in backend/src/v2/driver/driver.go and backend/src/v2/cmd/driver/main.go, the driver appears to be responsible for orchestrating task execution (ROOT_DAG, DAG, CONTAINER types) and managing execution state with MLMD, rather than translating specifications.

The pipeline-to-workflow translation is typically done by the compiler component. Consider revising this description to something more accurate like "Orchestrates task execution and manages execution state" or "Prepares and manages the execution context for pipeline tasks".

Suggested change
- **Driver**Translates pipeline specifications into executable
workflow definitions.
- **Driver**Orchestrates task execution and manages execution state
with ML Metadata (MLMD).

Copilot uses AI. Check for mistakes.
- **Launcher** – Executes pipeline tasks within Kubernetes pods.
- **Persistence Agent** – Reconciles workflow state and updates
run status in the database.
- **Metadata Writer** – Handles interactions with ML Metadata (MLMD).
- **Cache Layer** – Manages execution result reuse when caching is enabled.

Understanding these components helps contributors identify where
new features, debugging logic, or lifecycle changes should be implemented.

This README will help you set up your coding environment in order to build and run the Kubeflow Pipelines backend. The KFP backend powers the core functionality of the KFP platform, handling API requests, workflow management, and data persistence.

## Prerequisites
Expand Down
Loading