Description

We regularly get confusion from customers as to why some older files still exist in their snapshot storage.  In their mind, they are envisioning a world where the snapshot replaces everything (full backup), but in truth, S3 does an incremental, so older blocks that have been unchanged will remain while newer blocks keep coming in.  We self clean up old blocks as snapshots fall off, but in some cases, certain blocks may not go away for a while.

The issue/risk for customers in OM is that they also have to manage their S3 and filesystem stores, which means they often have direct access to these endpoints.  In some cases, customers try to reduce storage costs by expiring these blocks based on time outside of ops manager verses allowing us to manage.  This ultimately ensures corrupted snapshots with restores.  

We need a page or section on or around https://docs.opsmanager.mongodb.com/v4.4/tutorial/manage-s3-blockstore-storage/index.html that better explains how this works.  The ultimate goal here is to prevent accidental foot shooting and gives support a simple way to point and educate customers on the behavior.  

Scope of changes

Impact to Other Docs

MVP (Work and Date)

Resources (Scope or Design Docs, Invision, etc.)