A distributed, Prometheus-compatible, real-time, in-memory time series database designed for massive scalability and low-latency operational metrics.
FiloDB is a distributed, real-time time series database compatible with Prometheus, designed to handle massive-scale operational metrics ingestion and querying. It solves the scalability limitations of traditional monitoring systems by offering a sharded, in-memory architecture that supports millions of time series with low latency.
DevOps engineers, SREs, and platform teams managing large-scale monitoring infrastructure who need a scalable, self-hosted alternative to Prometheus for high-cardinality metrics.
Developers choose FiloDB for its native PromQL support combined with superior scalability, efficient histogram storage, and real-time performance, making it ideal for replacing or augmenting Prometheus in environments with millions of time series.
Distributed Prometheus 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.
Supports PromQL and integrates with Grafana, allowing drop-in replacement for existing monitoring stacks without changing query tools.
Sharded architecture is designed to ingest millions of time series across a cluster, with distributed querying for high-cardinality metrics.
First-class histogram support uses a compressed columnar format, improving query speed and reducing wire overhead compared to Prometheus' bucket-per-series model.
Data is immediately queryable after ingestion, with performance optimized for dashboards and alerting, as highlighted in the real-time ingestion features.
Built for dual-datacenter operation with no single point of failure, using Akka Cluster for peer-to-peer coordination and recoverability.
Requires multiple pre-requisites like Java, Cassandra, Kafka, and Rust, with a multi-step setup process that can be daunting for new users.
The README admits that FiloDB currently supports only about 60% of PromQL, which may limit query capabilities for advanced monitoring use cases.
Relies on external systems such as Cassandra and Kafka, adding to maintenance burden and operational complexity in production environments.
Dataset column definitions are immutable after creation, as noted in the configuration section, making it difficult to adapt to evolving data models.