[GEODE-3163] Spike: Check turning all backups into snapshots

• Type: Sub-task
• Status: Open
• Priority: Major
• Resolution: Unresolved
• Component/s: regions
• Assignee: Unassigned
• Reporter: Fred Krone
• Created: 30/Jun/17 20:14
• Updated: 25/Jul/17 17:31

Description

The core of this is a way to turn a backup file into some kind of "recovery" file that can be easily imported into any type of cluster. As an example we could turn a backup into a snapshot which is importable into any region.

This idea came up as we were thinking of how to restore a backup to a cluster of a different configuration.

-Check what the level of effort is to convert backups to snapshots
-How CPU intensive is this process? Can cluster be benchmarked during normal operations vs normal + snapshot conversion is running.
How long does it take to complete snapshot conversion on x size cluster? 2 TB?