An Efficient Distributed Index Method for Cloud Computing

To improve the performance and maintenance cost of existing index methods，this paper presents a distributed multi access entrance B+ tree index method，which features:①efficient parallel interval queries，and②and low maintenance cost for index structure. Our scheme relies on four key techniques to achieve the advantages mentioned above:(1)It provides a routing table for each leaf node of a distributed B+ tree，so a interval search can be completed from any leaf node to break through the bottleneck caused by the root node in a distributed B+ tree;(2)It constructs a balanced binary tree at different levels of a node，and selects the relevant nodes for its routing table;(3)It useslayer-by-layer transitivity of a B+ tree node to split information to sense the node splitting position and only updates routing information within the subtree of a corresponding particle at the time of nodesplitting;(4)It uses the balanced structure of B+ trees to only update single route information for relevant nodes at the time of node splitting，without adjusting the route tables of a large area.Our scheme has been implemented and evaluated.Performance results are encouraging.