topic Re: Your Compute session is disconnected in Data Engineering

Your Compute session is disconnected

Ojicletus — Thu, 28 May 2026 08:09:02 GMT

I am trying to use the trial version of Fabric. I get this whenever I try to load my dataset to a table or perform analysis using Spark. I get this error "InvalidHttpRequest

[TooManyRequestsForCapacity] [TooManyRequestsForCapacity] HTTP Response code 430: This Spark job can’t be run because you’ve hit a Spark compute or API rate limit. To proceed, cancel an active Spark job through the Monitoring hub, choose a larger capacity SKU, or try again later. For more visibility and control, go to Workspace settings → Job management (Job Concurrency & Queue Monitoring) to review running and queued Spark jobs, understand capacity contention, and take action as needed. [Learn more at 'https://go.microsoft.com/fwlink/?linkid=2356970&clcid=0x409']. HTTP status code: 430."

This is my first time using it.

Re: Your Compute session is disconnected

lbendlin — Thu, 28 May 2026 19:21:15 GMT

TooManyRequestsForCapacity - you are running too many requests in quick succession. See if you can space your requests out a bit more.

Re: Your Compute session is disconnected

arabalca — Thu, 28 May 2026 19:57:26 GMT

Hi @Ojicletus ,

This error means your Fabric capacity has reached the concurrent Spark jobs limit. This is a SKU/capacity-level throttling issue, not necessarily a problem with your code.

Learn more here: https://learn.microsoft.com/fabric/data-engineering/spark-job-concurrency-and-queueing

I would recommend reviewing the following:

1. Review and cancel active jobs

Go to Workspace Settings → Data Engineering/Science → Spark settings → Jobs and cancel any non-critical Spark jobs.

There are often "zombie" notebook sessions consuming Spark slots without doing anything useful.

https://learn.microsoft.com/fabric/data-engineering/job-concurrency-queue-monitoring

2. Activate High Concurrency (if applicable)

This can help improve Spark session utilization and reduce concurrency-related issues.

https://learn.microsoft.com/en-us/fabric/data-engineering/configure-high-concurrency-session-notebook

3. Review Spark compute configuration

In fact, I ran into the same error today and immediately thought of you.

In my case, I adjusted the Spark configuration. I had a Medium node running on an F4 capacity, and I changed it to Small, also adjusting the number of nodes. That resolved the issue for me.

Give it a try and let us know if it works for you: https://learn.microsoft.com/en-us/fabric/data-engineering/environment-manage-compute

Additional references that may help:

https://learn.microsoft.com/fabric/data-engineering/job-queueing-for-fabric-spark

https://learn.microsoft.com/fabric/data-engineering/autoscale-billing-for-spark-overview

https://learn.microsoft.com/fabric/data-engineering/troubleshoot-permissions-capacity#capacity-and-r...

If this response has been helpful, please don't forget to give it a Like and mark it as a Solution so other community members can find it easily.

Thank you!

Re: Your Compute session is disconnected

Olufemi7 — Fri, 29 May 2026 17:07:38 GMT

Hello @Ojicletus,
This looks like a Spark capacity issue rather than a dataset issue.
The TooManyRequestsForCapacity (HTTP 430) error means Fabric Spark compute is unavailable or a Spark/API rate limit has been hit.
Since you're using the trial version, try these quick checks:
1. Go to Monitoring hub and cancel any running or queued Spark jobs.
2. Check Workspace settings → Job management for queued jobs or capacity contention.
3. Disconnect the current compute session, reconnect it, then rerun the job after a short wait.
Docs:
Concurrency limits and queueing in Apache Spark for Microsoft Fabric

Re: Your Compute session is disconnected

carter_gray705 — Sat, 30 May 2026 04:03:23 GMT

This usually happens because the Fabric trial capacity is overloaded or has too many Spark jobs running. Check the Monitoring Hub for active or queued jobs and cancel any that are stuck. If nothing is running, wait a while and try again, as trial capacities often hit usage limits.