This is a dedicated watch page for a single video.
As part of your Google Cloud setup, you're establishing a new pipeline to stream IoT data from Cloud Pub/Sub through Cloud Dataflow to BigQuery. Upon previewing the data, you've noticed that approximately 2% of it seems to be corrupted. How should you adjust the Cloud Dataflow pipeline to filter out this corrupted data?