single node Cluster CPU not fully used

Data Engineering

Join discussions on data engineering best practices, architectures, and optimization strategies within the Databricks Community. Exchange insights and solutions with fellow data engineers.

Hello community,

I use a cluster (Single node: Standard_F64s_v2 · DBR: 14.3 LTS (includes Apache Spark 3.5.0, Scala 2.12)) for a job. In this job I didn't use spark multiprocessing. Instead I use this single node cluster as a VM and use python multiprocessing to finish my job.

Before 2024-08-06 04:30 (CET) I can use the CPU of the single node cluster relative full. And the job can be finished less than one hour. But after 2024-08-06 04:30 (CET) the CPU of the single node cluster can not be fully used, it remains under 25%. Even though I didn't make any changes.

So does anyone know why this is happening? Is there any changes made around 2024-08-06 04:30 (CET) on Data Bricks? some update or something like that?

Thanks for any advice and really appreciate for any help.

Best regards,

Narsu