Horizontally scalable log aggregation system that stores labels instead of full-text indexes
AGPL-3.0
- Go
- Jsonnet
- Go Template

About Grafana Loki
Grafana Loki is a horizontally scalable, highly available, multi-tenant log aggregation system. It stores and queries logs cost effectively by skipping full text indexing, so teams keep large volumes without the index overhead of search-engine backends.
Loki stores compressed, unstructured logs and indexes only metadata labels for each log stream. It uses the same labels as Prometheus, supports Kubernetes Pod logs with scraped metadata, and has native support in Grafana. Alloy gathers logs and sends them to Loki, and LogCLI provides a command-line query interface.
Loki is part of the Grafana stack, where Alloy ships logs in and Grafana queries them out. It can run in single-host, no-dependencies mode or scale out across microservices, and it is licensed under AGPL-3.0-only with Apache-2.0 exceptions.
Key features
- Indexes log streams by labels, not full text
- Stores compressed, unstructured logs
- Supports Kubernetes Pod logs with scraped metadata
- Native support in Grafana
- LogCLI command-line interface for querying logs
Details
- First released
- 2018
- Licensing
- AGPL-3.0-only with Apache-2.0 exceptions
- Platforms
- Web · CLI
- Deployment
- self-hostable · docker
- Storage
- Compressed logs with metadata labels
- Scaling
- Horizontally scalable, highly available
