Failure-Atomic Byte-Addressable R-tree for Persistent Memory

In this article, we propose Failure-atomic Byte-addressable R-tree (FBR-tree) that leverages the byte-addressability, persistence, and high performance of persistent memory while guaranteeing crash consistency. We carefully control order store cacheline flush instructions prevent any single instruction from making an FBR-tree inconsistent unrecoverable. also develop a non-blocking lock-free range query algorithm for FBR-tree. Since allows read transactions to detect ignore transient states, multiple can concurrently access tree nodes without using shared locks other write are changes them. Our study shows successfully reduces legacy logging overhead up 2.6x higher processing throughput than lock-based crabbing concurrency protocol.