An Azure real-time analytics service designed for mission-critical workloads.
Hey Dhruv Singla,
Thankyou for reaching microsoft Q&A! At first glance the “estimated arrival time” formula can look more complicated than it really is. Let’s break it down:
- Documentation says: estimatedArrival = lastEventArrivalTime + (now – lastEventArrivalWallClock) watermark (when no new events) = estimatedArrival – lateArrivalTolerance
- But notice that lastEventArrivalTime + (now – lastEventArrivalWallClock) is mathematically equal to “now” (assuming system clocks stay in sync).
- So when there’s truly no incoming event, “estimated arrival time” effectively becomes the current wall-clock time, and the watermark becomes: currentTime – lateArrivalTolerance
In other words, yes—it boils down to the same thing. The doc spells it out in that way so the service can reliably re-compute watermarks (e.g. after a failover or when you have multiple substreams), but in steady-state with no new data it’s just “now minus your late-arrival window.”
Hope that clears it up!
Reference list
- Time handling in Azure Stream Analytics (watermark definition) https://learn.microsoft.com/azure/stream-analytics/stream-analytics-time-handling#how-time-progresses-in-azure-stream-analytics
- Illustrated example of watermarks https://learn.microsoft.com/azure/stream-analytics/stream-analytics-time-handling#illustrated-example-of-watermarks
- Events dropped when arriving outside the allowed time policy https://docs.microsoft.com/azure/stream-analytics/stream-analytics-time-handling#events-dropped-when-arriving-outside-the-allowed-time-policy
Note: This content was drafted with the help of an AI system. Please verify the information before relying on it for decision-making.