ACM SIGACT News
Strategies for Traffic-Aware VM Migration
UCC '13 Proceedings of the 2013 IEEE/ACM 6th International Conference on Utility and Cloud Computing
| Hi-index | 0.00 |
This paper initiates the study of self-adjusting networks (or distributed data structures) whose topologies dynamically adapt to a communication pattern $\sigma$. % (i.e., an ever changing "traffic matrix''). We present a fully decentralized self-adjusting solution called \Splay Net. A \Splay Net\ is a distributed generalization of the classic splay tree concept. It ensures short paths (which can be found using local-greedy routing) between communication partners while minimizing topological rearrangements. We derive an upper bound for the amortized communication cost of a \Splay Net\based on empirical entropies of $\sigma$, and show that \Splay Nets\ have several interesting convergence properties. For instance, \Splay Nets\ features a provable online optimality under special requests scenarios. % and multicast tree scenarios We also investigate the optimal static network and prove different lower bounds for the average communication cost based on graph cuts and on the empirical entropy of the communication pattern $\sigma$. % which may be of independent interest. From these lower bounds it follows, e.g., that \Splay Nets\ are optimal in scenarios where the requests follow a product distribution as well. Finally, this paper shows that in contrast to the Minimum Linear Arrangement problem which is generally NP-hard, the optimal static tree network can be computed in polynomial time for any guest graph, despite the exponentially large graph family. We complement our formal analysis with a small simulation study on a Facebook graph.