Q: 8
You are designing a data processing pipeline. The pipeline must be able to scale automatically as load
increases. Messages must be processed at least once, and must be ordered within windows of 1
hour. How should you design the solution?
Options
Discussion
C/D? Practice exams hit on Dataflow windowing a lot. Check the official guide and do hands-on labs to be sure.
B tbh
Its D, Pub/Sub for ingestion and Dataflow for windowed ordering is the combo that auto-scales. Kafka is a common distractor here.
Is "must be ordered within windows" strictly referring to event time or processing time? If it's event time windows, then D fits, but if strict end-to-end ordering is needed across all messages, other options might be better.
Be respectful. No spam.
