Fast file system scheduling method and device

Abstract

The invention provides a fast file system scheduling method. The method includes the following steps: S100) acquiring a use rate of at least one already-allocated hard-disk space, judging whether a to-be-written file can be stored on continuous positions of the already-allocated hard-disk space, if yes, executing a step S200, and otherwise, executing a step S300; S200) storing the to-be-written file at the hard-disk space continuous-positions which are at an end of a most recently written file and are unassigned to chain tables, or storing the same at the starting positions which are of the hard-disk space and are unassigned to chain tables, and establishing a file chain table; and S300) assigning another at least partially physically-continuous extended hard-disk space of preset fixed capacity from a hard-disk pool for a virtual machine, writing the file into starting positions of the extended hard-disk space, and establishing a file chain table. The method has the advantages of avoiding frequent application of fragmented small-amount hard-disk spaces through limiting a size of each time of capacity expansion of a hard disk, and thus enabling fragmentation phenomena of the hard disk to be alleviated.

Description

technical field [0001] The invention relates to the field of computer technology, in particular to a fast file system scheduling method and device. Background technique [0002] In computer technology, with the rapid development of virtualization (Virtualization) technology and network technology, it has been widely used by various academic or commercial institutions to manage different hardware resources (such as central processing unit, server, memory, etc.) and hard disk resources). After the virtualization technology abstracts and converts the above hardware resources, it integrates physically independent hardware resources with different device parameters to form a unified hardware resource platform for easy management, so that the above hardware platforms meet isolation and scalability , stability and security and other technical indicators. After the above hardware resources are virtualized, different types of hardware resources jointly form various resource pools (...

AI Technical Summary

Problems solved by technology

However, multiple virtual machines or multiple clients editing shared files may make the file link too long, making it difficult to read related files

In addition, because the file link method adopts the data structure of the linked list to maintain files, each part of the file is stored in different physical locations on the hard disk; therefore, when multiple virtual machines or hard disks set in the network cloud run for a period of time, Ikegami will inevitably generate a large number of hard disk fragments that are difficult to recycle

In addition, when the file cannot be stored, the virtual machine will expand the hard disk capacity with an unfixed size, which further leads to fragmentation of the hard disk and makes it difficult to organize and recycle.

Images