The requirement is to calculate the average temperature over the last 5 minutes, but only when a new sensor reading arrives. Which windowing function should you use in the Eventstream Group By transformation?

• A. Tumbling Window
• B. Hopping Window
• C. Sliding Window
• D. Session Window

Answer: (C)

To get an up-to-date average over the most recent five minutes and have it recomputed with every new sensor reading, you need a moving, time-based window that advances as data arrives. A sliding window fits this scenario: it keeps a 5-minute window of data and updates the result each time a new reading comes in, so the average reflects the last five minutes of activity at the moment of the new event.

Tumbling windows split data into non-overlapping 5-minute intervals and emit only at the end of each window, so you won't get an update with every single reading. Hopping windows create overlapping windows that advance at fixed gaps, which can produce outputs on each hop rather than exactly when a new reading arrives. Session windows depend on idle gaps between events and aren’t fixed to a 5-minute lookback. Hence, the sliding window is the best match for this requirement.