BVLSM: Write-Efficient LSM-Tree Storage via WAL-Time Key-Value Separation

Abstract

Modern data-intensive applications increasingly store and process big-value items, such as multimedia objects and machine learning embeddings, which exacerbate storage inefficiencies in Log-Structured Merge-Tree (LSM)-based key-value stores. This paper presents BVLSM, a Write-Ahead Log (WAL)-time key-value separation mechanism designed to address three key challenges in LSM-Tree storage systems: write amplification, poor memory utilization, and I/O jitter under big-value workloads. Unlike state-of-the-art approaches that delay key-value separation until the flush stage, leading to redundant data in MemTables and repeated writes. BVLSM proactively decouples keys and values during the WAL phase. The MemTable stores only lightweight metadata, allowing multi-queue parallel store for big value. The benchmark results show that BVLSM significantly outperforms both RocksDB and BlobDB under 64KB random write workloads. In asynchronous WAL mode, it achieves throughput improvements of 7.6x over RocksDB and 1.9x over BlobDB.