Updated: Dec 2019

This is not an exhaustive list. I am no longer actively maintaining this list.

Open source

These are either time series databases or general-purpose databases that work well with time series. Some are layers on top of existing databases.

• Aerospike
  • High performance, in-memory, NoSQL
• Akumuli
  • Written in C++
  • Query language based on JSON over HTTP
  • Can be used as a server application or an embedded library
• Apache Apex
  • DataTorrent
• Apache Cassandra
  • Or Scylla, a much faster C++ implementation of Cassandra
  • Distributed, columnar database
  • Has a query language
• Apache Kudu (Incubating)
  • Columnar, part of the Hadoop stack
  • “fast analytics on fast data”
• Atlas by Netflix
  • Written in Scala
  • In memory
  • Stack language for queries
• Axibase Time Series Database
  • Visualizations, rules engine, forecasting
• Beringei by Facebook
  • In memory
  • Open source implementation of ideas presented in their Gorilla paper (link below)
• Blueflood
  • Built on Cassandra
  • Multi-tenant distributed database and metric processing system created by Rackspace
  • Apache 2.0 license
• Chronix
  • Built on Apache Lucene, Solr, and Spark
• CitusDB
  • Distributed Postgres (through an extension)
• ClickHouse
  • Distributed columnar database
  • Powers Yandex.Metrica (basically Russia’s Google Analytics)
• Cortex
  • Multitenant, horizontally scalable Prometheus as a Service
• CrateDB
  • Distributed SQL database
  • Fully searchable document oriented data store
  • Uses Presto for SQL, Elasticsearch and Lucene for storage
• Cube by Square
  • Built on MongoDB
• Cyanite
  • Compatible with the Graphite ecosystem
  • Stores data in Cassandra
• Dalmatiner
  • Built on ZFS and Riak Core
• Druid
  • Column-oriented open-source distributed data store written in Java
• Elasticsearch
  • Java & Lucene
  • Support for live time series resampling
  • Distributed data storage
• EventQL
  • Distributed, columnar database built for large-scale data collection and analytics workloads
  • Supports SQL
• FiloDB
  • Distributed, versioned, and columnar analytical database
  • Uses Spark SQL
• GridDB
  • Horizontally scalable NoSQL DB
• GridGain
  • In-memory data fabric
• Hawkular
  • Open source monitoring solution by Red Hat
  • Metrics storage uses Cassandra
• HBase
  • distributed database for very large tables
  • Related: Google Cloud BigTable (hosted)
• Heroic by Spotify
  • Based on Bigtable, Cassandra, and Elasticsearch
• InfluxDB
  • Written in Go
  • Clustering is a paid feature now
• IoTDB
  • Written in Java
• KairosDB
  • Rewrite of OpenTSDB
• M3DB
  • Distributed time series database using M3TSZ float64 compression
  • Developed at Uber
• Newts
  • Based on Cassandra
• OpenTSDB
  • Built on top of HBase, BigTable, or Cassandra
• Pinot
  • Realtime distributed OLAP datastore
  • Horizontally scalable
  • Used at LinkedIn
• Prometheus
  • Monitoring system and TSDB
  • Not distributed
  • Polling-based
• Riak TS
  • Query language
  • Apparently 10x faster than Cassandra (I don’t have any more details about this)
• Roshi by SoundCloud
  • Time-series event storage
  • Stateless, distributed layer on top of Redis and is implemented in Go
• SciDB
  • Multidimensional arrays
  • ACID
  • By Michael Stonebraker
• SiriDB
  • Written in C and focused on performance
  • Query language
• sonnerie
  • Written in Rust
  • Append only (no updates yet)
  • Simple TCP API
• Timely by the NSA
  • Backed by Accumulo
• TimescaleDB
  • Built on PostgreSQL (as an extension)
• Vulcan by DigitalOcean
  • Extends Prometheus adding horizontal scalability and long-term storage
  • Written in Go
• Warp 10
  • Geo Time Series database
  • Storage engine: standalone version uses LevelDB, distributed version uses HBase
  • Analytics environment: WarpScript (900+ functions)
  • From SenX
• Yuvi
  • In-memory storage engine for recent time series metrics data
  • Implemented in Java
  • Supports OpenTSDB metric ingestion and OpenTSDB queries

Proprietary or internal

These are either proprietary or internal, and not open source.

• Google BigQuery
  • Managed data warehouse for analytics hosted on Google Cloud
  • BigQuery can do lots of things in addition to time series (also see RedShift)
• Infiniflux
  • Time series DBMS with SQL
• IRONdb
  • Scalable storage for a Graphite infrastructure. IRONdb is a new product by Circonus, who also created “Snowth” a few years ago (see below).
• kdb+ by Kx Systems
  • Very popular in the financial industry
• AWS Redshift
  • Managed data warehouse for analytics hosted on AWS
  • Redshift can do lots of things in addition to time series (also see BigQuery)
• Rocana (acquired by Splunk)
  • Proprietary columnar TSDB using Apache Lucene, Kafka, and HDFS
• eXtremeDB
  • Made for financial data
  • Columnar, ACID-compliant, SQL support
• Facebook Scuba
  • Fast, scalable, distributed, in-memory database
• quasardb
  • Distributed transactional key-value store with distributed secondary indexes and native time series support
  • Written in C++14
• SnappyData
  • fuses Apache Spark with a highly available, multi-tenanted in-memory database
  • OLTP + OLAP on streaming data
• TempoIQ
  • IoT platform
• VictoriaMetrics
  • Long-term remote storage for Prometheus

Things to look at for ideas

These are not exactly TSDBs, but are interesting resources to take a look at.

• Apache Arrow
  • Columnar in-memory format and API
• Apache Drill
  • “Schema-free SQL Query Engine for Hadoop, NoSQL and Cloud Storage”
• Apache Parquet
  • Columnar storage format for HDFS
• BTrDB paper [PDF]
  • “Optimizing Storage System Design for Timeseries Processing”
• Circonus Snowth
  • YouTube video about the design
• Facebook Gorilla paper [PDF]
  • Fast, scalable, in-memory TSDB
• Honeycomb’s columnar data store “Retriever”
  • YouTube video about the design
• Pulsar
  • Streaming SQL
• Square metrics query engine

Changelog

• 2016-04-09
  Initial version
  • Thanks to Csaba Csoma and Damian Gryski (@dgryski) for their contributions.
  • Added Apache Drill, Kudu (thanks Mark Papadakis)
• 2016-04-10
  Added Cityzen Data, Hawkular, Infiniflux, TempoIQ, kdb+
  • Thanks to @pganti in the comments
• 2017-01-30
  Added SciDB, SiriDB
  • Thanks to @Pranas and @ps22 in the comments
• 2017-04-05
  Added Akumuli, Atlas, Beringei, Chronix, Roshi, Timely, TimescaleDB, Vulcan;
  Ordered by name
  • Thanks to Damian Gryski (@dgryski) and Khalid Lafi (@LafiKL) for their contributions.
• 2017-09-03
  Added EventQL, eXtremeDB, IRONdb; reorganized sections.
• 2018-03-23
  Added ClickHouse, M3DB, Honeycomb’s data store, quasardb, Cortex, Cyanite, BigQuery, RedShift, CrateDB, Pinot; misc cleanup and expanded notes.
  • Thanks to Damian Gryski (@dgryski), gilles t (@akiragt) Misha Brukman (@MishaBrukman), Andrew Montalenti (@amontalenti), and KurtB, Andy Ellicott in the comments.
• 2018-09-13
  Added Yuvi, sonnerie. Updated M3DB link to official website.
• 2019-05-13
  Added GridDB and VictoriaMetrics.