Explanation:
In a stream-static join, the static table is materialized exactly once at the start of the streaming query. Every subsequent micro-batch joins incoming stream records against that same in-memory snapshot. Updates made to the underlying static table on storage are not picked up until the query is stopped and restarted.
Ultimate access to all questions.
No comments yet.
When performing a join between a streaming DataFrame and a static Delta table in Databricks, which of the following statements is true regarding the behavior of the join?
A
The streaming query joins every micro-batch against the snapshot of the static Delta table that was present at the moment the job was initialized.
B
The checkpoint directory is responsible for tracking version changes and updates made to the static Delta table during the streaming job's execution.
C
Spark maintains state information for every unique key in the join within the checkpoint folder to support stateful stream-static joins.
D
Static Delta tables cannot be used in stream-static joins because Spark cannot guarantee exactly-once processing for the static side of the join.