Onehouse Open Engines is a managed capability for deploying selected open-source compute engines against lakehouse tables. It is not a new query engine: Onehouse provides the deployment and platform layer, while Apache Flink, Trino, or Ray handles the workload. The launch was announced on April 17, 2025; current Onehouse documentation describes the engines’ roles, costs, support boundaries, and technical limitations.
What is Onehouse Open Engines?
Onehouse announced Open Engines on April 17, 2025, as a component of its cloud platform that automates deployment of open-source engines on Onehouse Compute Runtime. The announcement says these engines can connect to tables created or managed inside or outside Onehouse. The initial engines named were Apache Flink, Trino, and Ray. Onehouse’s launch announcement describes the product; it does not make Open Engines itself a new engine.
Onehouse’s current product page presents the service as a way to choose compute suited to a workload while working with lakehouse data. Onehouse describes deployment, scaling, cost management, and performance in promotional terms; those are vendor claims, not independent validation. See Onehouse’s Open Engines product page.
Which engines does Open Engines support?
Current Onehouse documentation lists three engines and associates each with a different workload. Onehouse’s Open Engines overview describes their present roles.
Recommended Free Tools
#1 Best Overall
| Engine | Documented use | Important behavior on Onehouse tables |
|---|---|---|
| Apache Flink | Stream processing | Onehouse documentation identifies it as a supported engine; Flink supports one external catalog at a time. |
| Trino | Fast SQL analytics | Read-only for Onehouse tables; supports one external catalog at a time. Some access-control features, including CREATE ROLE, are not yet supported. |
| Ray | AI, machine learning, and data science | Read-only for Onehouse tables. |
How does Open Engines work with lakehouse tables?
The basic idea is to deploy an engine for a particular workload and connect it to lakehouse data. Onehouse says its engines can read existing Onehouse tables, and that Onehouse-managed table services can be deployed on tables created with Open Engines, subject to product constraints. This does not mean every engine can write to every table, or that every external catalog and table format works without configuration.
Table format and table management
According to the current limitations documentation, tables created by Open Engines can only be viewed and managed by Onehouse in Apache Hudi format, and they must be external tables under an Observed Lake. Review Onehouse’s documented Open Engines limitations before planning table creation or management.
Rank #2
Catalogs and concurrent writers
Trino and Flink currently support only one external catalog each. For concurrent writers, lock-provider configurations must be added manually. These constraints matter when integrating an existing catalog or coordinating writes from multiple processes; confirm that your intended setup fits the documented configuration.
Read and access controls
Trino and Ray are read-only for Onehouse tables. The docs also say access-control features such as CREATE ROLE in Trino are not yet supported. Treat these as product constraints when assessing write paths and security requirements, not as details that deployment automation will necessarily solve.
Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchPC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Rank #3
What does Open Engines cost, and what support is included?
As described in Onehouse’s documentation accessed October 4, 2026, the company is offering Open Engines free for a limited time, with no Onehouse OCU charges for usage. Cloud-provider resource consumption remains billable. The offer’s duration and terms can change, so check the current pricing and usage notes before budgeting. “Free” therefore does not mean that running workloads has no cost.
Onehouse documents support as limited to infrastructure-level issues. Customers seeking full engine-level support are directed toward specialized compute-engine partners; the cited documentation does not name particular providers. If production operations depend on help debugging engine behavior, establish who owns that support before adopting the service.
Rank #4
What performance and savings has Onehouse claimed?
In its April 17, 2025 launch announcement, Onehouse stated that Onehouse Compute Runtime could provide 2x to 30x query acceleration and reduce customer cloud-infrastructure bills by 20 to 80 percent. These are Onehouse’s claims, not independently validated benchmarks or guaranteed outcomes. The sources cited here do not establish that an individual workload will achieve either range. Measure performance and total cost with your own data, workload, and cloud-provider charges.
The announcement also mentioned $1,000 in free credits for 30 days as part of a launch-era test-drive invitation. That was a historical offer, not evidence of a current promotion.
When should you consider Open Engines?
Open Engines may suit teams that want Onehouse to manage deployment of a supported engine for a workload against lakehouse tables, and whose format, catalog, write, and support needs fit the documented constraints. Before choosing it over self-managed deployment or another managed offering, compare:
- Workload fit: whether Flink, Trino, or Ray covers the stream-processing, analytics, or AI/ML task.
- Data behavior: required read/write operations and Apache Hudi or other format requirements.
- Integration: external catalog needs, including the one-catalog limit for Trino and Flink.
- Operations and security: deployment and scaling responsibilities, concurrency configuration, and required access-control features.
- Total cost: any Onehouse charges under the current offer plus cloud-provider resource usage.
- Support: who will handle engine-level troubleshooting if infrastructure support is insufficient.
The Onehouse materials describe the product and its current constraints, but do not provide a neutral comparative benchmark against self-managed engines or competing managed services.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
