You need to migrate thousands of on-prem Apache Spark jobs to Google Cloud using managed services, minimize code changes, and move quickly. What is the migration approach?

Moving your data to Cloud Storage and running jobs on Dataproc is the most effective approach for migrating thousands of on-prem Apache Spark jobs to Google Cloud while minimizing code changes and moving quickly. Dataproc is a fully managed Apache Spark and Hadoop service, allowing you to run existing Spark jobs with minimal modifications. Cloud Storage provides a scalable and durable object storage solution for your data, seamlessly integrating with Dataproc. Converting Spark scripts to SQL for BigQuery would require significant code rewriting, which goes against the goal of minimizing code changes. Rewriting jobs in Apache Beam for Dataflow would also necessitate extensive code changes. Copying data to Compute Engine disks and managing jobs directly on instances would involve significant operational overhead and negate the benefits of managed services.