As public could offerings get more numerous and a growing number of enterprises embrace cloud-based IT solutions, the shortcomings of deploying traditional network gear and architectures in virtualized environments become more apparent. In scenarios where (virtual) end-systems are dynamically provisioned, dimensioned, and moved, the network services they use - ranging from mere connectivity to sophisticated application layer communication support - need to feature a similar level of flexibility, configurability and elasticity. These principles inspired the design of heleos, a software platform for powering elastic virtual network services that can be dynamically provisioned on general purpose hosts and modified as needed to match the changes in traffic requirements as well as the number and location of (virtual) end-systems. This paper reports on experiments demonstrating the scalability benefits of heleos-powered services and dismissing concerns on additional latency that might stem from their distributed architecture.