Abstract-While virtual machine (VM) migration is allowing data centers to rebalance workloads across physical machines, the promise of a maximally utilized infrastructure is yet to be realized. Part of the challenge is due to the inherent dependencies between VMs comprising a multi-tier application, which introduce complex load interactions between the underlying physical servers. For example, simply moving an overloaded VM to a (random) underloaded physical machine can inadvertently overload the network. We introduce AppAware-a novel, computationally efficient scheme for incorporating (1) inter-VM dependencies and (2) the underlying network topology into VM migration decisions. Using simulations, we show that our proposed method decreases network traffic by up to 81% compared to a well known alternative VM migration method that is not application-aware.

No abstract

No abstract

Newly virtualized network functions (like firewalls, routers, and intrusion detection systems) should be easy to consume. Despite recent efforts to improve their elasticity and high availability, network functions continue to maintain important flow state, requiring traditional development and deployment life cycles. At the same time, many cloud-scale applications are being rearchitected to be stateless by cleanly pushing application state into dedicated caches or backend stores. This state separation is enabling these applications to be more agile and support the so-called continuous deployment model. In this paper, we propose that network functions should be similarly redesigned to be stateless. Drawing insights from different classes of network functions, we describe how stateless network functions can leverage recent advances in low-latency network systems to achieve acceptable performance. Our Click-based prototype integrates with RAMCloud; using NAT as an example network function, we demonstrate that we are able to create stateless network functions that maintain the desired performance.

No abstract