You ingest real-time vehicle telemetry into Cloud Datastore and need long-term, low-cost snapshots for point-in-time recovery or cloning in another environment, archived for a long time. Which two methods accomplish this?

The question asks for two methods to create long-term, low-cost snapshots of Cloud Datastore data for point-in-time recovery or cloning. "Use managed export, and store the data in a Cloud Storage bucket using Nearline or Coldline class" is correct because Cloud Datastore's managed export feature allows exporting data directly to Cloud Storage. Using Nearline or Coldline storage classes provides a low-cost solution for long-term archiving, which aligns with the requirement for "low-cost snapshots... archived for a long time." "Use managed export, and then import to Cloud Datastore in a separate project under a unique namespace reserved for that export" is also correct. After exporting data using managed export, it can be imported into another Datastore instance, potentially in a different project or namespace, effectively creating a clone or snapshot for recovery purposes. This directly addresses the "point-in-time recovery or cloning in another environment" requirement. The other options are incorrect: Importing to BigQuery is not a direct Datastore snapshot for point-in-time recovery or cloning within Datastore. Writing a custom application for streaming inserts to BigQuery is overly complex and not a standard snapshot method for Datastore. Storing data in Cloud Source Repositories is not suitable for large datasets or long-term, low-cost archiving of operational data.