deepflow

What it is

DeepFlow is an open-source observability project that provides zero-code data collection using eBPF for metrics, distributed tracing, request logs and function profiling, with SmartEncoding for full-stack correlation and efficient storage.

How it works

The project description states zero-code data collection via eBPF, supports Wasm plugins for private protocols, and offers continuous profiling with low overhead. It integrates as storage for Prometheus, OpenTelemetry, SkyWalking and Pyroscope, and provides SQL/PromQL/OLTP APIs for data access. It also mentions SmartEncoding to reduce storage needs and enable scalable tagging.

Getting started

Quick start notes indicate there are three editions: Community, Enterprise, and Cloud. The Community edition comprises the core components (Agent and Server) with deployment docs available at the linked docs site. The README references a complete deployment guide for the Community Edition and a Community Demo with login credentials.

Quoted commands or steps for setup are not present in the truncated README excerpt provided, but there are references to deployment documentation and compile instructions for the agent.

Recent releases

Latest releases include:

• v7.1 (2026-03-25)
• v7.1.8 (2026-02-10)
• v7.1.7 (2026-01-28)
• v7.1.6 (2025-12-30)
• v7.1.5 (2025-11-19)

Traction

Stars: 4210 Forks: 477 Open issues: 264

Behind the repo

DeepFlow is developed as an open-source project and provides multiple editions (Community, Enterprise, Cloud) with a focus on zero-code observability and integration with common stacks.

Caveats

License: Apache-2.0 Created: 2022-01-16 Last push: 2026-08-04 Language: Go Topics include apm, gpu, kubernetes, llm, opentelemetry, wasm, zero-code

License and age information are stated; no further issues or open PR details are provided in the excerpt.