All-flash storage server system data management method and related components

Abstract

The invention discloses an all-flash storage server system data management method. According to the method, after the simplified volume receives the host IO, the grainSize granularity splitting is notcarried out any more, so the data is placed in the disk as much as possible according to a bulk data mode, continuous physical addresses are obtained and continuously stored in the disk space, the disk reading and writing frequency of the data is reduced, dicing processing is only carried out during metadata processing, and the mapping relation between actual physical addresses and logic addresses of placed disks is convenient to maintain. According to the method, the times of reading and writing the disk by data are effectively reduced, the read-write performance of the large IO is improved,and the time delay during data read-write is effectively reduced, so that the performance bottleneck during large data processing is broken through, the read-write modes of different sizes of data can be dealt with, the performance requirements of an all-flash system for different loads are met, and the performance of a storage system is integrally improved. The invention further provides a full-flash storage server system data management device and equipment and a readable storage medium, which have the above beneficial effects.

Description

technical field [0001] The present application relates to the field of electronic technology, in particular to an all-flash storage server system data management method, device, equipment and a readable storage medium. Background technique [0002] All-flash storage can provide customers with higher performance, lower latency, more flexible data storage services and efficient and flexible active-active, disaster recovery, backup and other solutions, which well meet the requirements of OLTP / OLAP for large and medium-sized databases. , virtualization, big data analysis and other application requirements, widely applicable to finance, communications, energy, media assets, medical care, education and other industries. [0003] The backend of the all-flash storage system is an SSD disk. Unlike traditional mechanical disks, SSD disks do not require the mechanical seek time of traditional disks, and their performance is greatly improved, especially the random read and write perform...

AI Technical Summary

Problems solved by technology

Therefore, the all-flash storage system has great advantages in the reading and writing of small blocks of IO. At the same time, in order to meet the needs of features such as online deduplication and compression, the data block granularity (grainSize) of the general back-end storage space is small (such as 8K). In business processing, there are a large number of large data blocks in addition to small data blocks, and the host IO will be split according to grainSize granularity and metadata management during data processing, which will cause large blocks of data to be split into small blocks, resulting in disk access The number of times increases, and the host IO time is extended

Images