StreamCore Inc. has an upstream data pipeline that writes files to Cloud Storage . These files are consumed by an Apache Spark job running on Dataproc in the us-central1 region. While data can be stored anywhere within the United States , the company requires a disaster recovery solution that can tolerate a maximum of 15 minutes of data loss (RPO = 15 mins) in the event of a catastrophic regional outage . It is also important to minimize latency when reading data for processing. What is the best approach?