A Cache Memory Replacement Method with Failed Disk Priority

Abstract

The invention provides a VDF (Victim Disk First) based cache memory replacement method. The method comprises the following steps: calculating the weight of a bottom memory block in all local stacks, deleting memory blocks with the maximum / or minimum weight in all the local stacks, wherein the weight calculation methods are different for a victim disk and a non-victim disk; putting a disk block in an available memory block obtained after deletion. Since the probability that the weight of the memory block of the victim disk is maximum or minimum is high, the purpose of preferentially retainingthe memory disk of the victim disk is reached. In the method, I / O (Input / Output) request frequency for the whole live disk is reduced by reducing the directing frequency of the cache to the victim disk, and less I / O requests can improve the performance of a disk array.

Description

technical field [0001] The invention belongs to the field of computer storage, and in particular relates to a method for replacing a cache memory (Cache) of a failed disk first (Victim Disk First, VDF). This method is suitable for a failed disk in a disk array, and can reduce the number of I / Os for the failed disk and improve the performance and reliability of the disk array. Background technique [0002] Reliability and availability are important criteria to measure the quality of online storage services. For systems that provide online services for commercial and industrial applications, data loss and intermittent service often means economic losses. Redundant Array of Independent Disk (RAID) is one of the mainstream choices that can provide high reliability and high availability. In earlier RAID specifications, such as RAID-0 through RAID-5, they only provided single-disk fault tolerance. In recent years, various RAID systems with higher fault tolerance have gradually ...

AI Technical Summary

Problems solved by technology

[0004] Cache is widely used and many Cache algorithms are used to solve the problem of disk delay, but the commonly used Cache replacement algorithm only considers that in the error-free mode, if some disks fail in the disk array, the RAID will still fail here Working in the mode, use Spare-rebuilding mode and degraded mode. At this time, the loss cost of the failed disk is completely different from that of the surviving disk. If the missing data resides in the Among the disks, the RAID controller accesses the surviving disks and performs parity checks to reconstruct the invalid data. At this time, the number of I / O requests is determined by the RAID structure. For the commonly used LRU and LFU algorithms, the missing cost of one disk failure among n disks is n-1 (n is a positive integer greater than or equal to 2), at this time, for a cache miss of a failed disk, the remaining n-1 disks need to be reconstructed to obtain the data of the failed disk, and then the data block of the failed disk If the cache is full at this time, the memory block at the bottom of the LRU global stack (the stack stores all the memory blocks of all disks) is directly removed, regardless of whether the memory block belongs to the invalid disk. The memory blocks of the failed disk are removed, and the data still needs to be reconstructed when accessing the failed disk, and n-1 I / O requests are generated again, which greatly increases the number of I / Os

Images