InfluxDB expands the possibilities of time series data, making challenging time series use cases, like network monitoring, observability and tracing, easier to implement than ever before.

100x faster real-time queries

A new purpose-built columnar time series database that combines a hot, compressed in-memory datastore and a cold, low-cost object store for the fastest write and query performance.

native-hl

Unlimited time series volume

Eliminating cardinality limits makes it easier to support use cases that use metrics, events, logs, and traces. Write data with infinite cardinality and slice-and-dice on any dimension without sacrificing performance.

021-petals-hl

SQL support

The new DataFusion-based query engine provides more options than ever to query data – API, InfluxQL, and SQL.

Built on modern open source formats

The new InfluxDB Engine is built in the Rust programming language and comprises of 4 major components:

  • Apache Parquet files for on disk storage
  • A compressed in-memory data store
  • Apache Arrow for data in-memory operations between components
  • The DataFusion query engine
IOx Architecture Diagram
fish icon

To learn more about this next generation of InfluxDB, watch Paul Dix’s InfluxDays 2022 session.