Details
-
Wish
-
Status: Resolved
-
P2
-
Resolution: Fixed
-
None
Description
So far all work on the Spark portable runner has been in batch mode. This is intended as an uber-issue for tracking progress on adding support for streaming.
It might be advantageous to wait for the structured streaming (non-portable) runner to be completed (to some reasonable extent) before undertaking this, rather than using the DStream API. Since work on the structured streaming runner is blocked by SPARK-26655, we should implement this using DStreams instead.
Attachments
Issue Links
- links to