Data storage method applicable for media streaming service

Abstract

The invention provides a data storage method applicable for media streaming service, and relates to the field of a file system. The data storage method comprises the following steps of numbering all storage disks in sequence, and recording the disk information of each storage disk into the same disk information file; dividing each storage disk into multiple equal-sized disk blocks, and numbering the disk blocks in sequence according to the starting offset positions of the disk blocks from small to large; selecting the storage disks by media files based on an algorithm M=(N-1)%C, in which N represents the number of the media file, M represents the number of the disk storing the media file, and C represents the total number of the storage disks; not only working out the quantity of the required disk blocks, but also selecting a required quantity of disk blocks by the media files; storing the media files into the selected disk blocks, and recording the media file information into content information files. According to the data storage method, the large-concurrency stream service of a stream service node is supported, the storage cost is reduced, and the dynamic update of hot contents and the available normal stream service of the stream service node under the abnormal conditions of the disks are supported.

Description

technical field [0001] The invention relates to the field of file systems, in particular to a data storage method suitable for streaming media services. Background technique [0002] At present, most of the Internet adopts the deployment method of CDN (Content Delivery Network, content distribution network). The media content in the Internet is stored and maintained by the central content library, and provides large concurrent streaming services through the streaming service nodes set up throughout the network. Therefore, the streaming service node only needs to store some hot media content, and the non-existent media content can be obtained from the central content library. [0003] Due to the large amount of media file data and the limited storage capacity of streaming service nodes, it is necessary for streaming service nodes to support not only dynamic updates of hot content, but also large concurrent streaming services; and the current standard file system does not sup...

AI Technical Summary

Problems solved by technology

Since the read and write bandwidth of a single storage disk is limited, in the case of large concurrent streaming services, if all streaming services access the same disk, read and write delays or failures will occur, affecting the quality of some streaming services

[0004] In addition, although some file systems can still provide streaming services when the disk is abnormal, these file systems generally require data redundancy, which not only cannot meet the large amount of concurrent disk IO reads and writes, but also has high storage costs.

Images