Explanation:
Correct Answer: B
Spark configurations set through the Clusters UI or the Clusters REST API are applied at the compute level. This means every notebook, job, or pipeline attached to that cluster inherits those settings. Note that changes to a cluster's Spark configuration require a restart to take effect.
Why the other options are incorrect:
RUNNING state, it must be restarted for the new attributes to take effect. This restart interrupts all active notebooks and jobs.spark.conf.set) are local to that specific notebook's SparkSession. They do not leak into or affect other notebooks sharing the same cluster.spark.conf.set or SET in SQL, the notebook-level value takes precedence for the life of that session.Ultimate access to all questions.
No comments yet.
When configuring Spark properties on the Databricks platform, which of the following statements regarding the scope and impact of those settings is accurate?
A
Modifying the Spark configuration of a running interactive cluster via the Databricks REST API allows the changes to take effect immediately without interrupting active workloads.
B
Spark configuration settings defined via the Clusters UI apply to the compute level, meaning every notebook or job attached to that cluster inherits those settings.
C
Any Spark configuration parameters defined within a specific notebook are automatically applied to all other Spark sessions sharing the same interactive cluster.
D
If a Spark configuration property is defined at both the cluster level and within a notebook session, the notebook-specific setting is ignored in favor of the cluster default.