Geographically dispersed Spotfire Nodes and Servers in a single collective

The idea here is that by supporting geographically dispersed collectives we can gain several improvements:

• Ability to process data near the data source (e.g. if we have a large database in Europe and others in US it would be advantageous to process data near the source rather than move it across the Atlantic every time we process it).
• Transparent disaster recovery. If one data centre goes down the collective can still function, albeit with reduced resources.
• Ability to mix cloud hosted (e.g. AWS) and on-premise hosted servers for similar reasons.

Obviously this change would require some serious work on the data wrangling architecture and some intelligence around where data is located and where it can be best processed. As an interim measure we could at least use Analysis routing rules to move processing to Nodes based near databases.