Skip to content
Tutorials Center
My learning

Topics

Lesson 2 of 39 ยท Getting Started

InfluxDB Versions and Editions

Understand the three InfluxDB generations, their query languages and terminology, the Core, Enterprise and Cloud editions, and how to choose one.

"InfluxDB" is not one product with one set of commands. It is three generations of software that share a name, a data model and a write format, but differ in storage engine, query languages and vocabulary. Many tutorials, forum answers and error messages you will find online refer to a specific generation, so knowing which is which saves hours of confusion.

This lesson draws the map. Later lessons say clearly when something applies only to one generation.

Product details change faster than a tutorial can. Treat the version numbers here as a guide to the families, and check the official documentation for the exact release you install. All InfluxDB commands in this course are read-only examples; the runnable blocks are labelled stand-ins.

The three generations

flowchart LR
  V1["InfluxDB 1.x<br/>InfluxQL<br/>databases + retention policies"] --> V2["InfluxDB 2.x<br/>Flux + InfluxQL<br/>buckets, orgs, tokens, UI"]
  V2 --> V3["InfluxDB 3.x<br/>SQL + InfluxQL<br/>Arrow, Parquet, DataFusion"]

Each generation kept line protocol, tags and fields, but changed the storage and the query story.

InfluxDB 1.x

The original open-source line. You work with databases and retention policies (a database can have several, and every measurement is written into one). The query language is InfluxQL, which looks like SQL with a few time-series twists such as GROUP BY time(5m). Writes go to /write?db=mydb. It is still widely deployed, so you will meet it in older systems and in many Grafana setups.

InfluxDB 2.x

Version 2 reorganised everything into organizations, buckets and API tokens, added a built-in web UI with a query builder and dashboards, scheduled tasks, checks and notifications, and introduced Flux, a functional scripting language for queries and data processing. InfluxQL remained available through a compatibility layer. A bucket combines a database name and a retention period in one object. Writes go to /api/v2/write?org=...&bucket=....

InfluxDB 3.x

Version 3 is a new engine written in Rust on top of Apache Arrow, Apache Parquet and the DataFusion query engine. It stores data in columnar Parquet files, which compress well and can live on object storage. Queries use SQL and InfluxQL; Flux is not part of the 3.x query engines. The vocabulary changes again: you write to a database that contains tables (what used to be measurements) with tags and fields as columns. A major design goal is to handle very high cardinality (many distinct series) without the performance cliff that 1.x and 2.x had.

Editions of InfluxDB 3

Edition Hosting Typical use
InfluxDB 3 Core Self-hosted, open source, single node Recent data, edge and development use
InfluxDB 3 Enterprise Self-hosted, commercial Clustering, high availability, longer history and compaction features
InfluxDB Cloud Serverless Managed, multi-tenant Pay for what you use
InfluxDB Cloud Dedicated / Clustered Managed or self-managed cluster Large dedicated workloads

The exact feature split between editions is decided by the vendor and evolves, so check the current comparison page before choosing.

Vocabulary cheat sheet

Idea 1.x 2.x 3.x
Place that holds data database bucket database
Group of points of one kind measurement measurement table
Expiry setting retention policy bucket retention period database retention period
Main query language InfluxQL Flux SQL
Access control users tokens (org-scoped) tokens
Write endpoint /write /api/v2/write /api/v3/write_lp (v1 and v2 endpoints kept for compatibility)

Line protocol stays the same across all three, which is why the line protocol lessons apply everywhere.

A decision helper (stand-in)

This small JavaScript function encodes the rule of thumb below. It is plain code, not an InfluxDB feature, but it makes the trade-offs concrete.

JS Which InfluxDB generation fits?

Press Run (or Ctrl + Enter) to see the console output here.

No output.

Same question, three dialects

Here is "the mean CPU usage per 5 minutes in the last hour" in each generation. These blocks are read-only: they need a running server.

TEXT
-- InfluxQL (1.x, also 2.x and 3.x)
SELECT mean("usage") FROM "cpu" WHERE time > now() - 1h GROUP BY time(5m)
TEXT
// Flux (2.x)
from(bucket: "metrics")
  |> range(start: -1h)
  |> filter(fn: (r) => r._measurement == "cpu" and r._field == "usage")
  |> aggregateWindow(every: 5m, fn: mean)
TEXT
-- SQL (3.x)
SELECT date_bin(INTERVAL '5 minutes', time) AS time, avg(usage)
FROM cpu
WHERE time > now() - INTERVAL '1 hour'
GROUP BY 1
ORDER BY 1

The three queries ask the same thing. The portable idea is a time filter followed by a time-bucketed aggregate. You can try that idea with SQLite, which has no date_bin, so the bucket is computed with arithmetic on epoch seconds:

SQL Five-minute buckets with epoch arithmetic (SQLite stand-in)

Press Run (or Ctrl + Enter) to see the result here.

How to choose

  • Starting fresh: pick InfluxDB 3 unless you specifically need Flux or a feature that only exists in 2.x.
  • Existing 2.x with many Flux tasks and dashboards: staying on 2.x is reasonable; migrate when there is a reason.
  • Existing 1.x: it keeps working. Plan the move, since 1.x receives little new development.
  • Hobby and learning: any version works; the concepts transfer. Use whichever installs most easily for you (Installing InfluxDB).

Check your understanding

Quick check ยท 4 questions

of 4 answered You scored / 4 ยท perfect!

  1. 1. Which query language is NOT available in the InfluxDB 3 query engines?

    InfluxDB 3 supports SQL and InfluxQL. Flux belongs to the 2.x line.
  2. 2. In InfluxDB 2.x, which object combines a database name with a retention period?

    A bucket is the named place that stores data and has a retention period.
  3. 3. What stays the same in all three generations?

    Line protocol is shared, which is why write examples are portable.
  4. 4. What did 1.x call what 3.x calls a table?

    A measurement in 1.x and 2.x corresponds to a table in 3.x.

Summary

  • InfluxDB 1.x uses databases, retention policies and InfluxQL.
  • InfluxDB 2.x adds organizations, buckets, tokens, a UI, tasks and Flux.
  • InfluxDB 3.x is a new Rust, Arrow and Parquet engine with SQL and InfluxQL; editions include Core, Enterprise and cloud offerings.
  • Line protocol, tags, fields and timestamps carry over between generations.
  • Pick 3.x for new projects unless you need Flux or already run 2.x.

Next: Installing InfluxDB