LoginSign Up
← Back to all sparks
Q

QuestDB

DEVOPS
Velocity7.5

High-performance time-series database for real-time analytics

questdb.io ↗

QuestDB 10.0 ships QWP — a binary protocol unifying ingestion and Arrow reads — while Enterprise 4.0 adds cold storage and hot failover.

time-seriesarrowbinary-protocolcold-storageicebergdata-lakehouse
◆Current state
QuestDB landed two major releases back to back: 10.0 (open source) and Enterprise 4.0. The central innovation in both is QWP — a binary, columnar wire protocol that handles both data writes and Apache Arrow reads over a single client connection. Enterprise 4.0 layers cold storage on top, aging partitions to object storage while keeping them queryable, and adds hot failover without a restart. The benchmarks are real: 33M rows/s against ILP, 500M rows streamed to Arrow in 2.3 seconds.
◆Where it's heading
QuestDB is positioning itself as the time-series database that fits into modern lakehouse architectures without requiring ETL. QWP + Arrow + cold storage + Iceberg compatibility all point toward a single goal: a streaming time-series engine that speaks natively to Parquet-based analytics stacks. The blog post explaining Parquet/Iceberg positioning is the clearest signal — the team is selling to data engineers building lakehouses, not just ops engineers running dashboards.
◆Prediction
The cold storage Iceberg integration will deepen with catalog support and query pushdown for Spark and Trino. QWP client libraries will ship for major languages beyond the initial Python/Java coverage.

◆Recent moves

  1. 12d ago

    Parquet and Iceberg: how a table format builds on a file format

    An educational post explaining how Parquet (file format) and Iceberg (table format) relate, and how QuestDB's cold storage integrates with an Iceberg lakehouse. Supporting content for the Enterprise 4.0 positioning; no product changes.

    View source ↗
  2. 22d ago

    QuestDB Enterprise 4.0: cold storage, QWP, and hot failover

    ⚡ SPARK

    QuestDB Enterprise 4.0 ships three major capabilities: cold storage that tiers aging partitions to object storage while keeping them queryable, QWP for faster ingestion and Arrow reads, and hot failover that doesn't require a restart. The lakehouse integration and HA story both make a meaningful enterprise jump.

    View source ↗
  3. 1mo ago

    QWP: QuestDB's own binary wire protocol for ingestion and queries

    A detailed technical post on QWP — QuestDB's new binary, columnar wire protocol. One client handles both writes (ingestion) and reads (SQL → Arrow), with built-in failover. Supporting content for the 10.0/Enterprise 4.0 launch.

    View source ↗
  4. 1mo ago

    Streaming 500 million rows into Apache Arrow in 2.3 seconds

    QWP benchmark: 500 million rows streamed to Apache Arrow in 2.3 seconds with a naive parallel reader. The post includes a careful accounting of measurement artifacts, making the numbers credible. Benchmarks ClickHouse and TimescaleDB for comparison.

    View source ↗
  5. 1mo ago

    QuestDB 10.0: QWP, one binary streaming protocol for writes and Arrow reads

    ⚡ SPARK

    QuestDB 10.0 is the open-source release anchoring the Enterprise 4.0 launch: QWP ships as the primary protocol, live views land in beta, and agent-driven notebooks are added. This release provides the storage and protocol foundation that the Enterprise cold storage tier builds on.

    View source ↗
  6. 1mo ago

    Introducing QuestDB's new binary ingestion protocol: QWP

    QWP vs ILP benchmark: QWP reaches 33 million rows per second on a single machine, approximately 3.6x faster than ILP over a network, even at one million series. Supporting content for the QWP protocol launch in 10.0.

    View source ↗