Influxdb

What it is

InfluxDB is an open source time series database, rewritten in Rust on top of Apache Arrow, Apache Parquet, and Apache DataFusion (the 3.x line, sometimes referred to as InfluxDB IOx). This is a fundamental architectural break from the original Go/TSM engine used in 1.x and 2.x — the storage format, query engine, and columnar processing model are all different. It's built for ingesting and querying high-cardinality time series data: metrics, events, sensor data, observability telemetry.

Who builds it and why

Built and maintained by InfluxData, the commercial company behind InfluxDB since its inception. The Rust/Arrow/Parquet/DataFusion rewrite was a deliberate bet to fix cardinality limits and query performance issues that plagued the original storage engine, and to align with the broader Arrow-based data ecosystem (so InfluxDB data plays natively with Parquet-based lakes, DataFusion-powered query tools, and Arrow Flight for transport). 693 contributors on GitHub indicates activity beyond a single vendor's core team, though InfluxData drives the roadmap and employs the majority of maintainers.

Production readiness signal

31,596 GitHub stars and a commit as recent as 2026-06-30 show active, ongoing development — this isn't a stalled project. Latest release is v3.10.0, suggesting a maturing but still fast-iterating major version line. CNCF maturity is not publicly available (InfluxDB is not a CNCF project; it's independently governed by InfluxData). Apache-2.0 licensing means no rug-pull risk on the open source core, though note that InfluxData has shifted licensing terms across versions historically (1.x/2.x had OSS and enterprise splits) — check which specific edition/tier you're deploying, since not all 3.x capabilities are OSS.

Who should use this

• Teams building observability, IoT, or industrial telemetry pipelines with high write throughput and high cardinality tags.
• Organizations already invested in the Arrow/Parquet/DataFusion ecosystem who want a time series store that speaks the same columnar language natively.
• Teams that outgrew InfluxDB 1.x/2.x cardinality limits and need better compression and query performance at scale.
• Anyone needing SQL and InfluxQL/Flux-adjacent querying over time series data without standing up a full data warehouse.

Who should NOT use this

• Teams needing a battle-tested, decade-proven storage engine today — the 3.x Rust rewrite is comparatively young versus the old TSM engine; if you need maximum stability, 1.x/2.x or an alternative may be safer short-term.
• Shops without Rust or Arrow ecosystem familiarity who just want to point Grafana at something and move on — the operational model and tooling maturity around 3.x is still catching up to 1.x-era simplicity.
• Anyone needing a general-purpose relational or document database — this is purpose-built for time series, not transactional workloads.
• Teams that need CNCF-governed, vendor-neutral guarantees — InfluxDB is single-vendor controlled, which matters for some procurement/compliance requirements.
• Small-scale use cases where SQLite, Postgres with TimescaleDB, or even a Prometheus box already does the job — InfluxDB's complexity isn't worth it below a certain data volume.

Alternatives

• TimescaleDB — Postgres extension for time series, if you want SQL and existing Postgres tooling/ops muscle memory.
• Prometheus — better fit for pull-based metrics monitoring at moderate scale, with a massive ecosystem of exporters and alerting.
• ClickHouse — general-purpose columnar OLAP store that many teams use for time series workloads at large scale, with a more battle-tested track record.

Pricing

Core InfluxDB 3.x is open source under Apache-2.0 — free to self-host with no licensing cost. InfluxData also sells InfluxDB Cloud (managed, usage-based pricing) and Enterprise tiers with additional clustering/support features. Exact enterprise/cloud pricing: not publicly available — check InfluxData's site for current quotes.