Method and device for data back-to-source, and edge server

Abstract

Embodiments of the invention disclose a method and a device for data back-to-source, and an edge server. The method comprises the following steps that a first edge server receives a plurality of downloading requests sent by front-end equipment, combines data intervals in the downloading requests to obtain a total data interval, covers each segment data of the total data interval as needed segment data after determining combination, obtains needed segment data which is not locally stored by the first edge server to back-end equipment, and transmits data which is requested to download by the downloading requests in the needed fragment data to the corresponding front-end equipment for each downloading request. The embodiments of the invention can combine the downloading requests for one to-be-downloaded file, which are sent by a plurality of front-end equipment, cover each needed fragment data of the total data interval after determining combination, and only download the needed fragment data which does not exist locally from the back-end equipment, so that the consumption of data back-to-source about the network bandwidth can be reduced for a condition of back-to-source of partial data of one file.

Description

technical field [0001] The present invention relates to the field of data transmission, in particular to a method, device and edge server for returning data to a source. Background technique [0002] At present, the data storage network generally adopts a layered topology, in which: one or several central storage server clusters are used as the central storage to store all the data, and these storages do not serve the client; the central storage is connected to the edge server, and the central storage Instead of serving clients, the edge server provides services to clients. Since it is impossible for the edge server to store all the data, there may be situations where the client device requests data from the edge server but the server does not have corresponding data. At this point, the edge server needs to request data from the back-end central storage and forward the requested data to the client device. The process in which the edge server requests data from the back-end ...

AI Technical Summary

Problems solved by technology

[0003] In the existing technology, if the client only needs the data of a certain part of a file, the edge server will still download all the data of the file, store all the data of the file, and then forward the part that the user needs to the user. This will download more file data that the user does not need, resulting in a waste of bandwidth; secondly, the edge server will not store the part of the file data that the user needs in the complete file data, and when a client requests this part of the data again, Then the edge server needs to re-return the data to the source, which will also cause a waste of bandwidth; again, if multiple users request the same data at the same time, the edge server will perform multiple data back-to-source operations to the back-end central storage , returning the same data to the source will cause low efficiency and waste of bandwidth; finally, in the existing data back-to-source technology, the edge server requests data from each client according to the order in which the requests arrive at the edge server. Perform data back-to-source without distinguishing the priority of each client's data request. If the first request is a low-priority request, and the subsequent request is a high-priority request, the first low-priority request will be Preempting bandwidth and affecting user experience

Images