A Scheduling Method for Database Unloading Data

Abstract

The invention discloses a scheduling method applicable to data unloading of a database. The method comprises the steps of obtaining configuration information of the database to be unloaded, and deploying parallel multi-processes used for data unloaded from the database; according to the configuration information of the database to be unloaded, establishing a first-in-first-out queue (FIFO) pipe file, and writing a preset number of line breaks into the FIFO pipe file, wherein each line break corresponds to a process; after each process reads a data line, writing a line break into the FIFO pipe file, and when no line break exists in the FIFO pipe file, completing data unloading of the database. By means of the scheduling method, data unloading can be accelerated, and rapid migration of database data is achieved.

Description

technical field [0001] The invention relates to the technical field of databases, in particular to a scheduling method for unloading data from a database. Background technique [0002] Data migration refers to the conversion of data in one data environment as needed, and import into another same or different data environment. Heterogeneous database migration is the need to transfer as much equivalent and lossless data as possible between data with different data structures, data components, and definitions in the database management system. [0003] In the process of migrating heterogeneous databases with a large amount of data on the existing Linux and Unix operating system platforms, because a single process performs data unloading, I / O resources cannot be fully utilized, making I / O the bottleneck of data unloading, and it is often impossible to Completing data unloading within the specified time window not only prolongs the data migration time, but also increases the cos...

AI Technical Summary

Problems solved by technology

[0003] In the process of migrating heterogeneous databases with a large amount of data on the existing Linux and Unix operating system platforms, because a single process performs data unloading, I / O resources cannot be fully utilized, making I / O the bottleneck of data unloading, and it is often impossible to Completing data unloading within the specified time window not only prolongs the data migration time, but also increases the cost of data migration, making data unloading often become the bottleneck of the entire data migration

If you want to unload data in parallel, you need to manually start multiple processes, which is inconvenient to operate

Images