This adds support for automated backups on a mongo db node to an EBS volume.
This has been used successfully on a client's mongo cluster. The way it works is to have a mongo node designated as the `MONGO_BACKUP_NODE`. On this node is a cron job that periodically launches the backup script. This backup script does roughly the following:
mounts an ebs volume
deletes all the files
dumps the mongo database with mongodump
unmounts the ebs volume
creates a snapshot of the volume
prunes old snapshots of the volume
This allows for flexible, consistent backups, versioned with ebs snapshots on aws.
*JIRA tickets*: TBD
*Merge deadline*: None
1. set appropriate vars for the `MONGO_BACKUP_*` vars found in
2. deploy a mongo node on aws
3. verify that the backup script runs on the `MONGO_BACKUP_NODE` instance.
[ ] @itsyed
[ ] edX reviewer[s] TBD