This is a dedicated watch page for a single video.
Scenario: You are running a streaming Cloud Dataflow pipeline. Your engineers have developed a new version of the pipeline with an updated windowing algorithm and triggering strategy. You need to update the running pipeline while ensuring that no data is lost. Question: How can you update the running Cloud Dataflow pipeline with the new version while preventing data loss?