As we continue to place a greater emphasis on improving durability, and availability in the server, there will be an increasing need to obtain production snapshots from internal Atlas clusters to test new features. Currently the process for retrieving these snapshots is unclear, so we should document the steps involved and possibly explore working with the Cloud teams to develop a more automated solution. 

The idea for needing this comes from SPM-4201 which is testing upgrade/downgrade using internal atlas snapshots, but has been facing some issues actually acquiring the snapshots.