Let me be very clear at the outset that I am new to cloud technology. All I know is cloud servers are easily scalable (i.e. adding physical hardware is transparent to applications and users) and unlike in older days application don't need to worry about scalability.
Now what does this mean to a TCP server application running on Azure? Can it assume that it is going to receive 500K TCP handles? (Unlike in older days TCP server application needed to be developed considering "Max Concurrent Connections" C10K limit)