
Ultimate access to all questions.
A web server sends click events to a Pub/Sub topic, with each event having an eventTimestamp attribute marking the click time. A Dataflow streaming job processes these events and sends the results to another Pub/Sub topic for the advertising department. Despite the Dataflow job's system lag being only 5 seconds and data freshness at 40 seconds, the advertising department is not receiving messages within 30 seconds of the click. What's the root cause of this delay, and how can it be resolved?
A
The advertising department is slow in consuming messages. Collaborate with them to improve consumption speed.
B
The Dataflow job processes messages in under 30 seconds but struggles with the Pub/Sub subscription backlog. Optimize the job or scale up workers to address this.
C
The Dataflow job's watermark isn't advancing because the eventTimestamp isn't factored into the windowing logic. Adjust the job's windowing logic to include eventTimestamp.
D
The web server's message push rate to Pub/Sub is insufficient. Coordinate with the web server team to enhance the push rate.