Key value data indexing method of workload self-adaptive single-layer LSMT

The invention discloses a key value data indexing method of a workload self-adaptive single-layer LSMT. According to the invention, a traditional log structured merge tree (Log-Structured-Merge Tree,LSMT) is optimized, multi-layer design and fixed memory table capacity design are removed, and design of a single-layer LSMT and a dynamic capacity memory table is introduced. According to the invention, the method comprises the steps: writing the writing operations into a log file on a storage device in sequence, and then modifying a memory table; when the size of the memory table reaches the capacity limit, converting the memory table into a read-only memory table, and combining the read-only memory table into a single-layer LSMT structure on the storage device in a background thread. On thebasis, the method can automatically optimize the storage structure according to the key value read-write distribution in the workload. According to the indexing method, read-write amplification of the storage device can be reduced simultaneously, system throughput is improved, and the service life of the storage device is prolonged. Meanwhile, self-adaptive optimization is carried out for the workload, so the system performance is further improved.