Log management method of distributed persistent memory transaction system

Abstract

The invention relates to the technical field of persistent memory storage, in particular to a log management method for a distributed persistent memory transaction system, and the method comprises a master copy node, a slave copy node and coordinator threads. The method comprises the steps that the master copy node and the slave copy node allocate an independent persistent log area to each coordinator thread in a cluster; the master copy node and the slave copy node provide log writing interfaces for the coordinator thread, call a CPU persistence instruction, and update log area metadata atomically; the master copy node executes the content of the persistent log area and recovers the junk log items, plays back the contents of the persistent log areas of the different coordinator threads inparallel from the replica node through the independent background thread group, and recycles the junk log items. According to the log management method of the distributed persistent memory transaction system, efficient log writing performance, multi-core extensible capability and recovery capability after crash can be provided at the same time.

Description

technical field [0001] The invention relates to the technical field of persistent memory storage, in particular to a log management method of a distributed persistent memory transaction system. Background technique [0002] Persistent memory (Persistent Memory, PM) is a new type of storage device, which supports directly using the CPU's load (load) and store (store) instructions to read and write to it at a byte granularity. Common persistent memory devices include Resistive RAM (RRAM), Phase Change Memory (Phase Change Memory, PCM), and Spin-Torque Transfer RAM (STT-RAM). Intel also released the first generation of Intel Optane DC Persistent Memory in 2019. Persistent memory has extremely high read and write bandwidth and extremely low read and write latency, which brings opportunities for building new storage systems. But at the same time, persistent memory has problems such as asymmetric read and write performance and low lifespan. Persistent memory changes the volatil...

AI Technical Summary

Problems solved by technology

First, the characteristics of the existing new hardware are not fully utilized: the persistence overhead of persistent memory is high, and CPU persistence instructions need to be called frequently; when using RDMA, it brings multiple unnecessary data copy overhead

Second, the multi-core scalability is poor, which is reflected in two aspects: first, multiple coordinator threads need to coordinate when writing logs to the master replica node and the slave replica node, and the requests are serialized; secondly, when the log content is executed from the replica node , due to the correct sequence of log playback, it is difficult to execute in parallel

Images