Best practice/model for scaling connector things?
Hi. Is there a best practice for how many connector Things should be created for a given system and given number of concurrent users/mashups that may be using that said connector?
For example, let's say I have 400 concurrent users/mashups that all connect to a business system via a Microsoft SQL jdbc connection. This connector thing has 50 services that call stored procedures on that SQL server. At any given point of time, a user may interact with the mashup to cause a data pull or at a minimum, we have refresh widgets that cause the mashups to refresh every minute. So I'd say it's a 80% chance that services on this connector gets called at the same time.
Do we create multiple SQL connection things to spread the load? Or can I bump the maximum connections on the MS SQL Template to 400?
And how does the inherit ThingWorx architecture/engine beneath support this, i.e. multi-treading?
Thanks!
Michael
