1. In a backup scenario, the snapshot is used to allow normal I/O operation on the share.
2. At the start of the snapshot, there are metadata pointers created on it that mark the original data.
No actual data is being copied.
3. When the backup starts, it uses that metadata to access the original share and get the data.
4. When a piece of data on the original share is changed while the backup task is running, this piece is copied to the snapshot BEFORE the change to the original takes place. The backup task then takes the original data directly from the snapshot. This is to ensure consistency.
