NOTE: The PR for REP-5265 Resolved (which is a Minor version update) is associated with this Spike.
The distributed datastore is fundamentally broken in a highly dynamic environment. One possible solution would be to externalize and centralize the datastore into persistent storage of some sort. Another would be to replicate data across nodes (using something like Hazelcast, perhaps).
- Determine the level of effort required to update Repose to support the new datastore.
- Create stories to make the updates.
- Make sure we refer to it as the "remote datastore" instead of the "distributed datastore".