The GizmoData Dispatch
December 2025 Edition
Merry Christmas, Happy Hanukkah, and Happy New Year from the GizmoData team!
As 2025 wraps up, we’re closing the year with big leaps for both GizmoSQL and GizmoEdge — and some exciting news heading into 2026.
🚀 GizmoSQL Updates
⚡ Bulk Ingestion Now Available
Our latest release introduces server-side Bulk Ingestion, allowing you to ship entire Arrow datasets directly into GizmoSQL server.
This breakthrough makes it effortless to push large datasets to DuckLake and other DuckDB-powered data lakes — no serialization or manual file staging needed.
We’re also contributing client-side improvements to the Apache Arrow ADBC codebase so that Go and Python clients can take full advantage of this endpoint. (See the “In the Works” section below for links!)
🧩 ADBC Scanner Extension from Query.Farm
Our brilliant friends at Query.Farm have built a DuckDB ADBC Scanner extension, which lets DuckDB query ADBC data sources directly — including GizmoSQL.
That means you can now reference GizmoSQL servers right from your local DuckDB CLI, or even chain multiple GizmoSQL instances together for federated queries!
👉 Try it yourself with our demo server and example here — running on our dime.
🪣 Upgraded to DuckDB v1.4.3
GizmoSQL now runs on DuckDB v1.4.3, which adds support for Azure Blob Storage writes — perfect for modern cloud workflows.
⚙️ Configurable Query Timeouts
You can now adjust the server’s query timeout both globally and per session via SQL:
-- Global (admins only)
SET GLOBAL gizmosql.query_timeout = 60;
-- Session-specific
SET gizmosql.query_timeout = 30;
🔉 Independent Log Levels
Admins can now set independent log levels for query and authentication logs, making it easier to tune verbosity without losing critical insight.
🧵 Better Concurrency Handling
We continue to tighten our concurrency architecture: shorter mutex locks, shared locks for read operations, and optimized thread coordination add up to dramatically improved throughput.
☸️ Kubernetes Operator by the Community
Big thanks to Nima Khoshini, who contributed a new GizmoSQL Kubernetes Operator!
This makes provisioning GizmoSQL instances on Kubernetes a breeze.
🏢 Introducing GizmoSQL Enterprise Edition
We’re thrilled to announce the upcoming GizmoSQL Enterprise Edition — a commercially licensed version that extends the already robust open-source GizmoSQL Community Edition with advanced enterprise features, including:
🔐 Role-Based Access Control (RBAC) with system, catalog, schema, and table/view-level privileges
🔑 Single Sign-On (SSO) for seamless authentication
📊 OpenTelemetry integration for observability and tracing
🧑💻 Session management and monitoring — admins can view or terminate sessions in real time
🤝 Enterprise support with SLAs and direct access to our engineering team
🧭 Roadmap influence — customers help shape future product direction
If your organization needs enterprise-grade support, compliance, and visibility for your DuckDB-powered (and DuckLake) workloads, GizmoSQL Enterprise Edition is built for you.
🌐 GizmoEdge Progress
GizmoEdge continues to evolve toward warp-speed distributed SQL — powered by a fleet of DuckDB workers that collaborate across devices and regions.
We’re currently seeking design partners interested in testing real-time, distributed query execution — especially for IoT use cases where you want to query your fleet without ETL.
If you want to join the early access program, reach out at gizmodata.com or reply to this newsletter!
🧠 In the Works: Contributing Upstream to Apache Arrow ADBC
We’re actively collaborating with the Apache Arrow team to improve the ADBC client libraries:
PR #3808 – Enables Go and Python clients to use GizmoSQL’s new Bulk Ingestion endpoint seamlessly.
PR #3810 – Fixes cursor cleanup when closing connections in the Python driver manager.
These improvements will make ADBC clients more robust, performant, and ready for enterprise use with GizmoSQL and beyond.
🎤 Upcoming Event
GizmoData Founder Philip Moore to Speak at the DuckDB Developer Meeting 2026
We’re excited to announce that Philip will be speaking about GizmoEdge at the DuckDB Developer Meeting on January 30, 2026.
If you’ll be in Amsterdam, stop by and say hi — and get a sneak peek at what’s next for distributed DuckDB compute.
💡 That’s a Wrap
2025 was the year we proved that open columnar data can scale from edge devices to the cloud — and 2026 will be the year we make it universal.
From all of us at GizmoData — thank you for your support, feedback, and enthusiasm throughout the year.
See you in Amsterdam 👋
💬 Call to Action
If your organization wants to:
Run GizmoSQL or GizmoEdge in production
Get enterprise-grade support for your data platform
Or explore modern data lakehouse architecture with the help of our team
👉 Contact us at gizmodata.com — we’d love to collaborate.