Open-source observability platform for searching and visualizing logs, traces, metrics, session replays and errors on top of ClickHouse.
- Stars
- 1,758
- Forks
- 93
- Open issues
- 28
- Last commit
- 22 hours ago
What is it?
Traceway is an observability platform built natively on OpenTelemetry. It brings logs, traces, metrics, exceptions, session replay for web and Flutter, LLM/AI tracing, and on-call paging into one system. Applications send data over OTLP/HTTP to Traceway directly, so no separate Collector or vendor-specific SDK is required for backends.
It can be self-hosted with docker compose, using SQLite, DuckDB, or ClickHouse with PostgreSQL as storage, or embedded inside a Go application using SQLite. Exceptions are fingerprinted and grouped into ranked issues, and minified JavaScript, iOS, Android and Flutter crashes are symbolicated back to original source locations using uploaded source maps, dSYMs or R8 mappings.
The project also ships an agent-oriented read-only CLI and agent skills for Claude Code, Cursor and Codex, with opt-in write access for agents. It is MIT licensed and includes SLO tracking, alerting integrations, and a fleet overview across projects.
Key features
- 01Native OTLP/HTTP ingest for logs, traces and metrics from any OpenTelemetry SDK.
- 02Exception grouping by SHA-256 fingerprint, with source map support for webpack, esbuild and Vite.
- 03Per-endpoint latency percentiles, throughput and error rate.
- 04Session replay for web applications and Flutter.
- 05LLM cost, token and conversation tracking via gen_ai attributes.
- 06On-call rotations, escalation policies and paging via email, Slack, Pushover, Telegram or SMS.
- 07SLO objectives with error budgets and burn-rate alerts.
Use cases
- Debugging production exceptions in a web or mobile app with replay and symbolicated stack traces.
- Consolidating logs, traces and metrics from microservices without running a separate Collector.
- Monitoring LLM-powered features for cost, latency and conversation content.
- Running a self-hosted observability stack on a single container with SQLite.
Quick start
- 1
Clone the repository
git clone https://github.com/tracewayapp/traceway - 2
Enter the directory and start with Docker Compose
cd traceway && docker compose up -d
Things to know
- Self-hosting the full stack requires running Docker and external databases unless you pick the SQLite or DuckDB image.
- Profiling is marked experimental in the README.
In collections
Topics
FAQ
- What is traceway?
- Open-source, OpenTelemetry-native observability platform combining logs, traces, metrics, exceptions, session replay and on-call paging, self-hostable with Docker. Traceway is an observability platform built natively on OpenTelemetry. It brings logs, traces, metrics, exceptions, session replay for web and Flutter, LLM/AI tracing, and on-call paging into one system. Applications send data over OTLP/HTTP to Traceway directly, so no separate Collector or vendor-specific SDK is required for backends.
- Is traceway free to use? What license does it have?
- traceway is open source under the MIT license. Check the license terms before using it in commercial products.
- Is traceway still maintained?
- Actively maintained. The last commit to traceway was on Nov 6, 2026.
- What can I use traceway for?
- Debugging production exceptions in a web or mobile app with replay and symbolicated stack traces. Consolidating logs, traces and metrics from microservices without running a separate Collector. Monitoring LLM-powered features for cost, latency and conversation content. Running a self-hosted observability stack on a single container with SQLite.
- How do I install traceway?
- Clone the repository: git clone https://github.com/tracewayapp/traceway Enter the directory and start with Docker Compose: cd traceway && docker compose up -d
- Where can I get traceway?
- The source code and installation instructions are on GitHub: https://github.com/tracewayapp/traceway
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
Open-source fullstack monitoring platform combining session replay, error monitoring, logs and traces, self-hostable via Docker.
Agentless, open-source real-time observability system unifying metrics, logs, alerting, notifications and status pages, with AI-assisted analysis.
An OpenTelemetry Collector distribution with programmable, expression-based pipelines for metrics, logs, traces and profiles, built with Prometheus integration in mind.
Self-hosted observability backend that accepts logs, metrics, traces and profiles and serves them through Loki, Prometheus, Tempo and Pyroscope compatible APIs.
An open-source observability database that stores metrics, logs, and traces in one columnar engine on object storage, queryable with SQL and PromQL.