1. How does stream processing differ from batch processing?
A. Stream processing handles data continuously, often in real-time
B. Stream processing only works with static data sets
C. Stream processing always requires manual scheduling of jobs
D. There is no difference between them
2. What is a windowed aggregation in stream processing?
A. Processing each data point independently without grouping
B. Grouping and aggregating events within a specific time interval
C. An offline technique used only after the stream has ended
D. A method to replicate data to multiple clusters
3. Why is fault tolerance an important consideration in stream processing?
A. Because all streaming systems run on a single server
B. To ensure that event data is neither lost nor double-processed when failures occur
C. It only matters for batch systems
D. Data in streams is never critical and can be discarded
4. How can exactly-once semantics be implemented in a distributed stream processing system?
A. By simply restarting the system whenever a failure is detected
B. By using idempotent writes, checkpointing, and careful coordination protocols (e.g., two-phase commit)
C. It’s not possible to have exactly-once semantics; at best, you get at-least-once
D. By disabling all retries and only processing the first event