HyperbyteDB

A sharded time-series database, built so tag cardinality stops being a ceiling.

Speaks InfluxDB v1 line protocol and InfluxQL, so switching costs you one URL. Open source.

GitHub

A time-series database without a cardinality ceiling

The problem is tags, not volume

Time-series databases rarely fall over because of how many points you write. They fall over because of how many distinct series those points create. Add a container ID, a device serial, a player session, and the series count multiplies — a database comfortable at a hundred thousand series starts taking seconds to answer queries that took milliseconds.

InfluxDB v1 was built for lower-cardinality workloads, and it is very good at them. The problem is that nothing about your business stays low-cardinality, and the exits are all bad: drop the tags you needed, downsample data you'll want next quarter, or move up to a tier where replication stops being optional.

Change one URL

Your monitoring, IoT, observability or analytics stack already speaks InfluxDB v1 line protocol and InfluxQL. HyperbyteDB is a full drop-in replacement:

• Same `/write` and `/query` endpoints
• Same InfluxQL you know (aggregates, derivatives, moving averages, GROUP BY time(), subqueries, regex measurements, and more)
• Same clients and agents (Telegraf, Grafana, etc.)
• Continuous Queries
• High Availability Replication

Point them at HyperbyteDB and go. The documentation covers the endpoints in detail.

The numbers, and where they came from

We lead with the sustained figure because it is the one that survives a capacity plan. The 99M is real too — it is the storage engine with nothing in front of it. Between the two sit a network, a parser, a series index and a durable write-ahead log, and that gap is where the honest number lives.

Where the other 98% goes · How the engine reaches 99M · Reproduce it yourself

Sharding

HyperbyteDB runs today as a single node, and a single node has a cardinality limit however good the storage engine is. Sharding splits series across nodes — that is what turns cardinality from a hard limit into a capacity decision. It lands this year, and when it does we will publish the same kind of measurement we published for ingest.

What changes for you

Tag what you actually want to tag. Columnar storage and a ClickHouse execution engine handle wide WHERE clauses and high-cardinality tags without the series explosion that degrades InfluxDB v1.

Replication without a sales call. Master-master clustering where every node accepts writes, in the open-source binary rather than behind a licence tier.

Durable by default. Writes land in a RocksDB write-ahead log before acknowledgement, then flush to columnar storage in the background.

No rewrite. The migration is a URL change, not a project.

Self-hosted or managed

HyperbyteDB Cloud runs it for you. Early access is open to a small number of design partners.

Solutions

Optimizing Energy Systems with HyperbyteDB

Real-time monitoring and analytics for smart grids, renewables, and utility operations The energy sector is undergoing massive transformation — from renewable integration and smart metering to predictive maintenance and demand-response programs. These initiatives generate vast amounts of time series data that must be ingested, stored, and analyzed at scale with high

Hyperbyte CloudAustin

Powering Real-Time Financial Analytics with HyperbyteDB

How a high-performance time series database transforms trading, risk management, and market intelligence Financial markets move at lightning speed. Every millisecond counts when processing tick data, monitoring portfolios, detecting anomalies, or running complex risk calculations. Traditional databases often buckle under the volume, velocity, and cardinality of financial time series data.

Hyperbyte CloudAustin

Observability and Monitoring with HyperbyteDB

High-cardinality metrics at scale — without the pain of traditional time series databases Modern applications are distributed, dynamic, and complex. Engineering and SRE teams need deep visibility into system health, application performance, user experience, and infrastructure behavior — in real time. This generates enormous volumes of time series data: metrics, events, and

Hyperbyte CloudAustin

Unlocking Scalable IoT Insights with HyperbyteDB

How a modern time series database powers the Internet of Things revolution In today’s connected world, IoT devices generate massive volumes of time-stamped data. From smart factories and agricultural sensors to fleet management and smart cities, organizations need a robust solution to ingest, store, and analyze this data in real

Hyperbyte CloudAustin