Hierarchical data optimization method and apparatus for hybrid disk

Abstract

The invention discloses a hierarchical data optimization method and apparatus for a hybrid disk. The method comprises the steps of classifying data according to data indexes; storing the classified data in corresponding storage devices of a hierarchical storage management system according to types; and when the data indexes are changed, performing upgrading migration or degrading migration on the data. According to the hierarchical data optimization method and apparatus for the hybrid disk, provided by the invention, the data distribution can be enabled to be more effective based on the hybrid disk; and the writing capability is ensured by utilizing the disk, and hotspot data is read by utilizing an SSD (Solid State Disk) to improve the reading capability, so that the read and write capabilities of the storage system are improved.

Description

technical field [0001] The invention belongs to the technical field of storage systems, and in particular relates to a hybrid disk layered data optimization method and device. Background technique [0002] In modern storage systems, due to the gradual increase in storage requirements, the capacity of hard disks has increased from TB to PB, and the speed of data processors and various links is getting faster and faster, but the speed of hard disk placement cannot keep up with it in time. The processing speed makes it a short board, which is easy to cause system downtime, suspended animation and other related states, and many application data require the support of the database, which greatly affects the performance of data processing. Traditional data storage is based on disks. Disks have large capacity, stability, and low cost, but the speed cannot keep up. In the frequent read and write access of large amounts of data, especially hot data, the defects are clearly manifested...

AI Technical Summary

Problems solved by technology

In addition, the frequent use of some calculation-type functions, such as custom functions and decode, will consume a lot of CPU time and increase the load on the system. The correct design method or optimization method needs to avoid the calculation process as much as possible, such as It is a good way to save the calculation results to the statistical table

In the OLTP environment, the disk subsystem's carrying capacity generally depends on its IOPS processing capacity, because in the OLTP environment, the disk physical read is generally dbfile sequential read, that is, single block read, but the number of reads is very frequent , if it is so frequent that the disk subsystem cannot bear its IOPS, there will be a big performance problem

Images