Spark Notebook termination after ~14 days (Eventstream Destination)

Hey Hamidr  I believe this is an expected behavior today in Microsoft Fabric.

When you set a Notebook as an Eventstream destination, Fabric spins up a managed Spark job.
That job runs continuously, consumes capacity and has an internal lifetime limit.

After around 14 days, Fabric terminates the job automatically (this is a platform safeguard).

At the moment there is no setting to make it run indefinitely.

I'd recommend you do lightweight transformations in Eventstream, then land data to Lakehouse (append) or Eventhouse / KQL database (as KQL table).

Then add scheduled or triggered notebooks to do upsert / merge logic for downstream consumptions.

Appreciate if you can 'Kudos' and/or 'Accept as Solution' if this answered your query.