Database and file system access method and apparatus

Abstract

Embodiments of the invention disclose a database and file system access method and apparatus. The method comprises the steps of determining a source database or source file system type corresponding to a table name and a keyword in an input standard SQL (Structured Query Language) statement according to metadata; when the source database or source file system type of the table name and the keyword in the standard SQL statement is an RDBMS (Relational Database Management System) or an MPP (Massive Parallel Processing) system, correspondingly transmitting the standard SQL statement in which the table name and the keyword are located to the RDBMS or the MPP system; when the source database or source file system type of the table name and the keyword in the standard SQL statement is a distributed computing Hadoop system or an NoSQL (Not Only Structured Query Language) database, analyzing the table name and the keyword, correspondingly generating an access statement capable of running in the Hadoop system or the NoSQL database, and correspondingly transmitting the access statement capable of running in the Hadoop system or the NoSQL database to the Hadoop system or the NoSQL database; and receiving an access processing result from the RDBMS, the MPP system, the Hadoop system or the NoSQL database.

Description

Technical field [0001] The invention relates to database management technology and file system access technology, in particular to a method and device for accessing a database and a file system. Background technique [0002] Currently, the existing big data processing platforms generally adopt a mashup architecture, generally including relational database management systems (RDBMS, Relational Database Management System), Massive Parallel Processing (MPP) systems, and non-relational databases (NoSQL, Not OnlySQL). ) And all or part of the components in the distributed computing Hadoop system. [0003] However, when accessing the above four databases or file systems, the required access methods are different. For example, commonly used RDBMS systems have similar storage methods and access methods, and data processing generally uses structured query languages ​​(SQL, Structured Query Language), supports the American National Standards Institute (ANSI, American National Standards Inst...

AI Technical Summary

Problems solved by technology

[0004] To sum up, in order to cope with different access methods corresponding to different databases or file systems, data users may need to write SQL statements, Shell scripts and Map / Reduce programs to process data at the same time, resulting in data users needing to learn different Access technology accesses data, which increases the technical threshold and is not easy to get started

[0005] At present, in order to solve the above-mentioned differences in access technologies of different databases and file systems, relational database data is usually migrated to the big data environment of the Hadoop system, so as to perform relatively unified data analysis, but migrating data requires additional input / output (IO, Input / Output) ports and space resources still cannot provide a method of accessing data across heterogeneous platforms that can make various databases and file systems transparent to users.

Images