Data operation method and data management server

Abstract

Embodiments of the invention provide a data operation method and a data management server. The method comprises the steps of obtaining an identifier and a data operation request of a lessee, wherein the data operation request is used for requesting to perform data operation on data of the lessee, and the data operation request comprises a first table name corresponding to the data; determining a second table name of a data table, corresponding to the data, in a first database according to the identifier of the lessee and the first table name, wherein the first database is used for storing data tables of multiple lessees, the data table of each lessee corresponds to the identifier of each lessee, and the lessee is one of the multiple lessees; replacing the first table name, corresponding to the data operation request, in a first SQL with the second table name; and requesting the first database to execute the first SQL after replacement, thereby finishing the data operation on the data. According to the method and the server, a complex SQL rewriting process can be avoided; the data operation efficiency can be improved; and data perturbation among the lessees can be prevented, so that data security is ensured.

Description

technical field [0001] The invention relates to the technical field of databases, in particular to a data operation method and a data management server. Background technique [0002] Based on the purpose of maximizing the sharing of resources, when storing data in a database, software as a service (Software as a service, referred to as SaaS) providers give priority to the shared database and shared data table mode, that is, to store multiple data in one data table in the database at the same time. The data of each tenant can be identified through the tenant identification field to identify the tenant to which the data belongs, which can achieve the highest degree of database resource sharing and minimum resource overhead, thereby supporting large-scale small and medium-sized enterprises to maximize revenue. In the prior art, there are huge differences in tenant customization fields in shared database and shared data table modes, which makes it very complicated to realize ten...

AI Technical Summary

Problems solved by technology

In the existing technology, there are huge differences in tenant customization fields in the shared database and shared data table mode, which makes it very complicated to realize tenant customization requirements

[0003] Existing technology solves the problem of complex data reading and writing while solving the flexible customization of tenants. However, this complexity should not be perceived by users. Users only need to know all the fields of the table (baseline fields and customized fields of tenants), A complex data storage model that does not need to be concerned with database extensions

In order to avoid the complexity of data processing by tenants and improve development efficiency, the existing data management model needs to analyze the data submitted by users in Structured Query Language (SQL) first, and then rewrite it according to the custom metadata translation for The SQL for the actual storage model, this step is generally more complicated, and this step must be repeated for each SQL request, which will consume a lot of performance resources and reduce system performance when there is a large amount of concurrency

In addition, since multiple tenants share data tables, in order to isolate tenant data, tenant identification (Identification, ID for short) will be added to filter when rewriting SQL, but this is only logical isolation, and the data is not physically isolated.

In this case, it is possible to bypass the filtering based on the tenant ID through attack methods such as SQL injection before adding the tenant ID for filtering, and data leakage interference between tenants occurs, so there are potential data security risks

Images