Restore Data from Glacier Lifecycle to S3

[nateiam]

As a Herd Consumer I want to restore a Bdata to S3 that had been archived to Glacier so I can access the data.

Existing restore endpoint currently restores from old Glacier 0d bucket. This story updates restore to work with data archived by the new approach as implemented in #324

Acceptance Criteria

• Restore is triggered by existing BData Restore POST endpoint.
  • Only change to endpoint is additional input for restore duration. Default if value not provided by user is configurable and defaults to 30d.
• Herd uses new AWS 'Bulk' restore tier, see AWS documentation
• Herd checks at configurable interval for completion of restore
• Once Glacier restore for all S3 Objects (Herd Storage Files) is complete, the following should occur
  • Herd validates metadata against restored files in original S3 locations
  • Herd modifies metadata to reflect completed restore and availability of data
• Glacier Storage is removed
  • Availability and DDL services behave as expected after data is restored