A high-performance time-series database optimized for modern hardware, supporting both metrics and events with efficient compression.
Akumuli is a time-series database optimized for modern hardware, designed to capture, store, and process time-series data in real-time. It uses a novel LSM and B+tree hybrid data structure with efficient compression, supporting both metrics and events. The database provides high-throughput ingestion, crash safety, and fast queries without dependency on database cardinality.
Developers and engineers building monitoring systems, IoT applications, or any platform requiring high-performance time-series data storage and real-time processing.
Akumuli offers superior compression performance, high write speeds, and flexible deployment options as a server or embedded library. Its design for modern SSDs/NVMe and support for tools like Grafana and Prometheus make it a robust choice for scalable time-series data management.
Time-series database
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
Akumuli's compression algorithm outperforms Gorilla, enabling many queries to run without decompressing data, as stated in the README.
Benchmarks show up to 16.1M writes/sec on high-end hardware, with millions of writes/sec on multi-core systems, making it ideal for real-time ingestion.
Built with crash safety and recovery mechanisms to ensure data durability, providing reliability for critical monitoring systems.
Can be used as a server or embedded library with a simple JSON/HTTP API, offering versatility for different application architectures.
Queries are executed lazily and read speed doesn't depend on database cardinality, enabling efficient range scans and joins on large datasets.
Lacks support for updates and deletes, which are only planned for future versions, restricting dynamic data management in current use.
Clustering and replication features are not implemented, limiting scalability for distributed setups without external workarounds.
No Windows support, and ARM architectures lack continuous integration, as noted in the supported platforms, which may hinder deployment flexibility.