Date Added: Mar 2012
Distributed computing is being used in several fields to solve many computation intensive problems. There has been recent interest in developing Peer-To-Peer (P2P) networks that can support ad-hoc queries over key ranges. A solution is to use range partitioning of data across the peer nodes. If the data and query distributions are uniform, nodes will have equal loads. However, if the data and/or execution are skewed, the network will develop hot-spots with high query traffic for a few nodes. Load balancing thus becomes a critical requirement in such a system. The P2P environment imposes three significant challenges for developing a load-balanced, range-partitioned system, that is, scale, dynamism and decentralization.