Data modification method and device based on storage service and storage server

Abstract

The embodiment of the invention discloses a data modification method and device based on a storage service and a storage server. The data modification method comprises the following steps: analyzing amodification operator carried in each data modification request concurrently reported for a storage server; in the process of sequentially operating each modification operator, modifying each modification operator; and updating the to-be-modified data pointed by the modification operator in the storage server into an operation result of the modification operator on the to-be-modified data, wherein the to-be-modified data is the data which is pointed by the modification operator in the storage server and is stored at the current operation moment of the modification operator. According to the technical scheme provided by the embodiment of the invention, the modification operators are operated in sequence, so that modification conflicts for the same storage data under high concurrency are avoided, and accurate data modification of the storage service under high concurrency is realized, and the high concurrency performance of data modification on the storage server is improved on the basis of ensuring the accuracy of data modification based on the storage service.

Description

technical field [0001] The embodiment of the present invention relates to the technical field of video storage, and in particular to a data modification method, device and storage server based on storage services. Background technique [0002] With the rapid development of Internet technology, there will be a large amount of resource data for storage. Common storage services will provide different business parties with functions such as query, modification and deletion of stored data, so that the business parties can use the stored data normally; At this time, the modification operation of the business party on the stored data usually requires the business server where the business party is located to read the corresponding stored data from the storage server first, and then execute the relevant calculations under this modification locally on the business server. And write the operation result into the storage server again as the modification result, so as to realize the mod...

AI Technical Summary

Problems solved by technology

[0004] At this time, when performing high-concurrency data modification operations based on mutexes, there is an operation of adding exclusive locks to the stored data, while other business parties cannot operate, which makes it necessary to sequentially modify the data under high concurrency. Execute the modification operations of various business parties, and continuously interact with the storage server to store data operations. There is a huge data operation overhead, which reduces the performance of high-concurrency modification of stored data; at the same time, high-concurrency data modification is performed based on optimistic locking. During the operation, the version verification will be performed on the write operation submitted for the final modification. The operation version used in this modification is inconsistent with the current version of the storage server. If there is a modification conflict, the current version needs to be used to re-execute the operation of this modification. And check it again. At this time, during the high-concurrency data modification process, the current version of the storage server will be continuously updated, which will lead to a sharp increase in the modification conflict rate under high concurrency, making the business side perform frequent re-calculation and modification operations, which greatly Reduced the performance of highly concurrent modification of stored data

Images