computing limit exceed by using auto-ml (free trial version)

[juan007105]

To test the auto-ml feature, I created a new dataflow in the workspace and trained it. The results were satisfactory, so I scheduled the dataflow for refresh.

Today, a BI error was detected, and upon investigation, the following message was found: "ErrorMessage":"Dataflow refresh failed because your organization's Fabric compute capacity has exceeded its limits. Try again later. Learn more at https://aka.ms/capacitymessages"

I have identified three problematic dataflows in the FABRIC CAPACITY METRICS, and I'm contemplating how to resolve this situation.

Currently, the dataflows in this workspace are unusable, and I'm unable to update any related BI materials.

• Can I use them normally if I only delete or do not update the problematic dataflows?
• Will this issue resolve itself if I wait?
• I am currently using a PRO account (along with a FREE TRIAL VERSION); would upgrading to a PPU license solve the problem?
• Should I abandon this workspace and create a new one?
• When using the AUTO-ML feature, what aspects should I look into to lift this restriction?

Is there anyone who can help with these questions?

[lbendlin · Accepted solution]

- Learn about Overages and Burndowns, The capacity throttling will resolve itself after 24 hours unless you keep accruing CU debt.

- consider enabling Autoscale  (comes at additional cost!)

- Abandoning the workspace will resolve nothing - what would "help"  is moving the workspace to another, non-overloaded capacity. Of course this assumes you have that option.