Runtime updates in Microsoft Fabric have always been about keeping you on the latest libraries, security patches, and platform improvements. But we've heard the feedback — surprise updates that reach production without a testing window can be disruptive. Now, we're introducing Fabric Runtime Release Channels, a new feature that gives you control over when and how Spark runtime updates reach your workloads and an opportunity for you to validate your workloads with these changes early.
Our solution focus
Spark runtime updates can include library upgrades, dependency changes, security patches, and even operating system updates. While every update goes through rigorous internal validation, no test matrix can cover every customer scenario. Previously, there was no way for you to preview these changes before they became the default. That changes today.
How it works
Every Spark runtime now has two release channels:
- Default channel — The current production-grade runtime that all users run automatically.
- Early access channel — A production-grade early access channel containing the next set of updates, available for you to test before they become the default.
The process is simple and cyclical: we publish upcoming changes to the early access channel, and then you get a validation window to test your workloads. Once that window closes, the early access channel is promoted to the new default. A fresh early access channel is then introduced with the next round of changes, and the cycle continues.
Figure: The Fabric Runtime Release Channel Lifecycle through publication validation, promotion, and refresh.
Get started in minutes
Opting in takes just two Spark configuration properties:
spark.fabric.pools.skipStarterPools=true
spark.computeConf.runtime.releaseChannel=earlyAccess
Figure: Setting up Spark properties to use Early Access Release Channel.
Set these in an Environment item, attach it to your notebook or Spark Job Definition, and your session runs on the early access channel. When you're done testing, remove the setting or set it to default — no permanent changes, no risk.
Track what's changing
Release notes are now published per channel in the Spark Runtime Releases and Updates repository. You can compare default and early access notes side by side to know exactly what's coming. Additionally, you can identify the exact VHD name by running this command in your session:
print("VHD Name: " + spark.conf.get("spark.synapse.vhd.name", ""))
Next steps
You now have greater control over your Fabric Spark Runtime experience. Try the early access channel today and open a support ticket if you face any issue. For step-by-step setup instructions, configuration details, and information on tracking VHD versions, explore the full documentation: Fabric Runtime Release Channels.
Microsoft Employee