We've got an existing commercial OpenEdge application with many customers, each of which has its own database and its own state-free appserver. We've been providing web service access to the application for awhile now, with a separate web service deployed for each appserver broker. We're finding that performance is decent with that configuration, but that maintaining that many web service instances is taking more resources (machine and human) than we'd like.
The multi-tenancy options that are available in OpenEdge 11 have raised the question of whether it would be advisable to try to figure out a way to have one web service instance provide access to all of the different customers. The obvious first idea was to put all of the customers' data in one big multi-tenant database with one state-free appserver broker and one web service instance, but for various reasons that is not practical.
Is there a good scalable approach for configuring OpenEdge web services/appservers such that multiple applications with their own databases can be securely accessed through one web service instance?