KairosDB

Free plan APILinuxSelf-hostedWeb

What KairosDB is

KairosDB is a free, open-source distributed time-series database built on Cassandra. It accepts metric data through REST, Telnet, Graphite plaintext or pickle, and plugins. Its REST API retrieves data by metric name and time range, with optional tag filters, aggregation, and downsampling. Built-in aggregators include minimum, maximum, sum, count, and mean. Documentation describes H2 as the default in-memory datastore and Cassandra as an alternative for persistent storage. The system supports retention policies and horizontal scaling, and Guice-based plugins can add listeners, datastores, protocol handlers, and system monitors. Command-line tools import and export data, while a separate Java client can send metrics and query data points. A default installation serves a browser query page intended mainly for development. KairosDB runs with Java 1.8 or later and supports Linux and self-hosted deployment. It is licensed under Apache License 2.0. The Cassandra row width defaults to about three weeks and cannot be changed after schema creation without risk of data loss.

Who it suits

KairosDB may suit teams that need a self-hosted time-series database with several ingestion methods, query aggregation, and extensibility. Its development-oriented browser interface and Java requirement are worth considering when planning deployment.

What it does well

Accepts data through REST, Telnet, Graphite, and plugins.
Queries support tag filters, aggregation, and downsampling.
Supports retention policies and horizontal scaling.
Command-line tools import and export data.
Licensed under Apache License 2.0.

What to know first

Runs with Java 1.8 or later.
The browser query page is primarily for development.
Cassandra row width cannot safely be changed after schema creation.
The default Cassandra row width is about three weeks.

Limits

H2 is an in-memory datastore by default, while persistent storage uses Cassandra. Cassandra row width defaults to about three weeks and cannot be changed after schema creation without risking data loss.

Getting started

Install Java 1.8 or later
Deploy KairosDB on Linux, Kubernetes, or another self-hosted environment
Choose H2 for in-memory storage or configure Cassandra for persistence
Set the Cassandra row width while creating the schema
Send metrics through REST, Telnet, Graphite, or plugins
Use the REST API or Java client to query data

Verdict

KairosDB offers multiple ingestion paths and query features for time-series data, with plugin options and persistent storage through Cassandra. Plan the Cassandra schema carefully because its row-width setting cannot be changed later without risking data loss.

Questions

Is KairosDB free?

Yes. KairosDB is open source and licensed under Apache License 2.0.

Which ingestion methods are supported?

Data can arrive through REST, Telnet, Graphite plaintext, Graphite pickle, or plugins.

What storage backends are available?

The documentation describes H2 as the default in-memory datastore and Cassandra as an alternative for persistent storage.

Can KairosDB scale horizontally?

Yes. KairosDB supports horizontal scaling and retention policies.

What runtime does KairosDB require?

KairosDB runs with Java 1.8 or later.

Facts from KairosDB's own pages

Aggregators: Available aggregation functions include min, max, sum, count, and mean. · kairosdb.github.io, 11 Oct 2026
Data collection: It accepts data through Telnet, REST, Graphite, and plugins. · kairosdb.github.io, 5 Oct 2026
Data ingestion: Data can be submitted through Telnet, REST, Graphite, or plugins. · kairosdb.github.io, 11 Oct 2026
Deployment: The documentation says KairosDB runs under Kubernetes and is mostly stateless, while Cassandra can be harder to run efficiently there. · kairosdb.github.io, 11 Oct 2026
Extensibility: Guice-based plugins can add data point listeners, data stores, and protocol handlers. · kairosdb.github.io, 11 Oct 2026
Import and export: The server provides command-line tools to import and export data. · kairosdb.github.io, 5 Oct 2026
Java client: KairosDB provides a Java client library using HttpClient to send metrics and query the server. · kairosdb.github.io, 11 Oct 2026
License: The repository identifies the license as Apache License 2.0. · github.com, 11 Oct 2026
Notable configuration limit: The Cassandra row width can be changed when the schema is created, but not afterward. · kairosdb.github.io, 11 Oct 2026
Notable limit: The Cassandra row width defaults to about three weeks and cannot be changed after the schema is created without risking data loss. · kairosdb.github.io, 5 Oct 2026
Product: KairosDB is a distributed time-series database built on Cassandra. · kairosdb.github.io, 5 Oct 2026
Purpose: KairosDB is a fast, distributed, scalable time-series database built on Cassandra. · github.com, 11 Oct 2026
Queries: The REST API supports querying by metric and time range, with optional tags, aggregation, averaging, min and max calculations, and downsampling. · kairosdb.github.io, 11 Oct 2026
Requirements: The installation guide says KairosDB runs with Java 1.8 or later. · kairosdb.github.io, 5 Oct 2026
Runtime requirement: KairosDB runs with Java 1.8 or later. · kairosdb.github.io, 11 Oct 2026
Security: The KairosDB pages reviewed do not state security certifications or compliance claims. · kairosdb.github.io, 5 Oct 2026
Storage backends: KairosDB can use an in-memory H2 database or Cassandra to store data points. · kairosdb.github.io, 11 Oct 2026
Support: The project README directs users to a KairosDB discussion group for questions and community involvement. · github.com, 5 Oct 2026
Web interface: A default install serves a browser-based query page that the documentation describes as designed primarily for development. · kairosdb.github.io, 11 Oct 2026

Where KairosDB stands on Freedom251

Time-Series Databases#20 of 36