eBPF-powered observability and APM with AI-guided root cause analysis
- Stars7.8k
- Forks385
- Open Issues116
Apache-2.0
- Go
- Vue
- JavaScript

About Coroot
Coroot is an open source observability and APM tool that brings metrics, logs, traces, and profiles together in one place. It cuts down manual investigation by turning that telemetry into actionable insights, including automated root cause analysis and SLO-based alerting.
It collects everything automatically with eBPF, so even legacy or third-party services are covered with zero code changes. On top of the data it builds a service map, predefined inspections, distributed tracing, log pattern clustering, logs-to-traces correlation, continuous profiling, and deployment tracking, and can attribute cloud costs to applications across AWS, GCP, and Azure.
Coroot follows an open-core model: the community edition is Apache-2.0 and self-hostable, running as a Docker container or in any Kubernetes cluster, while a separate enterprise edition is offered alongside it. It uses ClickHouse for fast log search.
Key features
- Automatic collection of metrics, logs, traces, and profiles via eBPF
- eBPF instrumentation with zero code changes
- Service map, predefined inspections, and SLO-based alerting
- Distributed tracing and continuous profiling
- Log pattern clustering, logs-to-traces correlation, ClickHouse search
Details
- First released
- 2022
- Instrumentation
- eBPF · zero code changes
- Deployment
- self-hostable · docker
- Telemetry
- Metrics · logs · traces · profiles
- Storage
- ClickHouse
- Model
- Open core
