Wide-area cooperative storage with CFS
SOSP '01 Proceedings of the eighteenth ACM symposium on Operating systems principles
Making gnutella-like P2P systems scalable
Proceedings of the 2003 conference on Applications, technologies, architectures, and protocols for computer communications
Gossip-Based Computation of Aggregate Information
FOCS '03 Proceedings of the 44th Annual IEEE Symposium on Foundations of Computer Science
ATEC '04 Proceedings of the annual conference on USENIX Annual Technical Conference
Characterization and measurement of TCP traversal through NATs and firewalls
IMC '05 Proceedings of the 5th ACM SIGCOMM conference on Internet Measurement
Journal of Systems and Software
Chunkyspread: Heterogeneous Unstructured Tree-Based Peer-to-Peer Multicast
ICNP '06 Proceedings of the Proceedings of the 2006 IEEE International Conference on Network Protocols
A comparison of structured and unstructured P2P approaches to heterogeneous random peer selection
ATC'07 2007 USENIX Annual Technical Conference on Proceedings of the USENIX Annual Technical Conference
An adaptive dynamic load balancing for large scale distributed and virtual simulations
VECIMS'09 Proceedings of the 2009 IEEE international conference on Virtual Environments, Human-Computer Interfaces and Measurement Systems
Efficient dynamic itinerary and memory allocation for mobile agents
Proceedings of the International Conference on Advances in Computing, Communications and Informatics
The XtreemOS Resource Selection Service
ACM Transactions on Autonomous and Adaptive Systems (TAAS) - Special Section: Extended Version of SASO 2011 Best Paper
Replica-aided load balancing in overlay networks
Journal of Network and Computer Applications
| Hi-index | 0.00 |
Random peer selection is commonly used to provide load-balancing in decentralized P2P systems. This paper addresses two practical concerns with using random peer selection for load balancing. The first has to do with heterogeneous peer selection mechanisms that accommodate differences in capacities of different peers. These heterogeneous selection mechanisms are not parameter-free: Each peer assumes knowledge of a particular selection-parameter, where correctly setting the parameter requires the knowledge of the global distribution of peer capacities. In this paper, we present a method that realizes parameter-free random peer selection by automatically computing the required parameter at each peer, thus making the selection primitive easier to employ by P2P applications. A second common problem addressed by this paper is that of ensuring good load-balance in heavily loaded P2P systems, where new requests that result in more load still need to be accommodated. We give a method that estimates the overall utilization in the network, and adaptively determines the number of attempts needed to probabilistically accommodate each new request. We implement both these enhancements over the Swaplinks peer selection algorithm, and show using experimental evaluations that they work as intended.