Peak time-series performance
QuestDB is the world's fastest growing open-source time-series database. It offers massive ingestion throughput, millisecond queries, powerful time-series SQL extensions, and scales well with minimal and maximal hardware. Save costs with better performance and efficiency.
Performance
- Columnar storage
- SIMD-optimized queries
- Ingest 4M rows/s per node
- Don’t worry about cardinality
- Data partitioned by time
- Compressed data in Parquet
Developer experience
- Open source under Apache 2.0
- SQL and time-series joins
- Open formats: Parquet & Arrow
- InfluxDB Line Protocol API
- REST and Postgres APIs
- Grafana native plugin
Enterprise ready
- Hot & cold read replicas
- Decoupled storage/compute
- Role-based access control
- Multiple availability zones
- Query Parquet via Object Storage
- Premium support SLA
Explore our live QuestDB demo and data
Query over 1.2 billion rows in milliseconds
Try QuestDB demo in your browser
QuestDB leaps your team forward
Leadership through open formats
Globally distributed, hyper-fast next generation database
- Open formats
- Leverages existing open formats. No vendor lock-in.
- Apache Parquet
- Enhanced compression and encoding, for ingress or egress.
- Super read/write
- In-memory processing, combined with SQL.
- Full stream
- Stream market data in from feeds or sensors, apply Parquet on read
- Direct to Parquet?
- Bypass QuestDB ingest, query Parquet directly from the object store
- Versatile ecosystem
- Diverse clients connect to your data, app, AI and ML frameworks
Simple, high performance SQL
Easily adopted, time-series optimized SQL
SELECT timestamp, symbol, priceFROM tradesWHERE timestamp IN '2024-06-21;1M';
-- Search timeSELECT timestamp, symbol, priceFROM tradesWHERE timestamp IN '2024-06-21;1M';
SELECT timestamp, symbol, avg(price)FROM tradesSAMPLE BY 5m FILL(LINEAR);
-- Slice timeSELECT timestamp, symbol, avg(price)FROM tradesSAMPLE BY 5m FILL(LINEAR);
SELECT timestamp, symbol, priceFROM tradesLATEST ON timestamp PARTITION BY symbol;
-- Navigate timeSELECT timestamp, symbol, priceFROM tradesLATEST ON timestamp PARTITION BY symbol;
SELECT timestamp, bid_price, ask_priceFROM bidsASOF JOIN asks;
-- Merge timeSELECT timestamp, bid_price, ask_priceFROM bidsASOF JOIN asks;
Peak performance time-series
Hyper ingestion
For massive volumes of time-series data, trust a specialized time-series database. Timestamps are first-class, and blistering throughput is sustained even through the highest data cardinality. Deduplication and out-of-order ingestion also work right out-of-the-box.
Using QuestDB, we deliver time series datasets with real-time market data from a top 10 blockchain. This outperforms a legacy cloud-based data platform at a fraction of the cost, reducing TCO by >90%.
Community love for QuestDB
Category user reviews on G2 rate QuestDB with the highest user satisfaction. Visit our vibrant Slack to chat with the community, the QuestDB core developers, and our technical AI-bot, to find out why.
Better on a Pi, best on full hardware
Max performance, min hardware
Compute costs keep rising. Keep them low with an efficient solution. QuestDB runs well on low hardware, within high cardinality data and with deduplication and out-of-order indexing. On just a 4GB Raspberry Pi, QuestDB outperforms competitors on robust hardware. Have robust hardware? QuestDB scales up to many millions of rows per second.
QuestDB is the first database that scales to the volume and cardinality of data that Beeks Analytics is capable of generating.
QuestDB is a time series database truly built by developers for developers. We found that QuestDB provides a unicorn solution to handle extreme TPS while also offering a simplified SQL programming interface.