Selective backups allow the backup of specific collections or databases.

This is a FR to allow the backup (and restore) of only a part of a collection, where the subset of data to be included in the backup is defined by a specific query (similar to what the --query option provides to mongodump and to mongoexport).

Use case:

• Reducing the amount of data in production, by archiving old data. Example: keep the data of the last 12 or 18 months, and "archive" all the remaining data.

Assuming this FR is implemented, the archive process could would as follow:

• Identify the specific set of data to be archived, using a query

• Use this feature to take a "filtered or partial" backup - and verify the backup

• Delete from production the data returned by the query used in previous steps

The backup can be used to restore the missing data from production, if it will be needed.

For this specific use case, if the restore will be needed to recover the archived data, it will be performed over the same cluster (and over the original collection), not on a different cluster.

Thanks,