Yes, the approach to set it in the spark config you used is correct and according to best practices. Additionally, you can put it in cluster policy so it will be for all clusters.


My blog: https://databrickster.medium.com/