DATA REPLICATION METHOD IN A POOL OF DATA SERVERS BASED ON SCHEDULER

PURPOSE: A scheduler-based data replication method is provided to improve the speed and performance of data processing and secure a massive storage system's availability by scheduling a data processing process based on the status of the data servers which engage in a data replication process. C...

Full description

Saved in:
Bibliographic Details
Main Authors HUR, SUNG JIN, WON, HEE SUN, LEE, SANG MIN
Format Patent
LanguageEnglish
Korean
Published 04.07.2013
Subjects
CALCULATING
COMPUTING
COUNTING
ELECTRIC DIGITAL DATA PROCESSING
PHYSICS
Online AccessGet full text

Cover

More Information
Summary:PURPOSE: A scheduler-based data replication method is provided to improve the speed and performance of data processing and secure a massive storage system's availability by scheduling a data processing process based on the status of the data servers which engage in a data replication process. CONSTITUTION: The necessary information is prepared for data replication processing. A request is issued in order to duplicate the items including the necessary data on a data replication list on order. The data replication request is inputted in a processing list and the data replication process proceeds. Upon a processing completion signal, the request is deleted from a queue and a meta database is renewed reflecting the result of the completed data replication process. [Reference numerals] (801) Obtain a target number of copies; (802) Obtain the value of a bandwith to be copied; (803) Obtain the number of the data servers in good shape from a data server management table; (804) Set the number of the data servers in good shape as a value of copiable semaphore; (805) Obtain the version information in a disk to make a copy request; (806) Increase the copy request value by one; (807) Store the increased copy request value in the disk; (810) Is a copy queue with a high priority empty?; (820) Is a copy queue with a low priority empty?; (821) Set the target copy number as the target copy number of a copy processor; (822) Remove first file inode metadata from the copy queue with a low priority; (830) Set the target copy number as two; (831) Remove first file inode metadata from the copy queue with a high priority; (840) Place the file inode matadata on the copy processing list; (841) Make a request to inode-copy the files on the copy processing list; (850) Wait for the results; (851) Remove the processed files from the copy processing list and the queues; (852) Update the metadata of the copied inodes; (860) Sleep in a certain period of time; (AA) Start; (BB, CC) Yes; (DD, EE) No
Bibliography:Application Number: KR20110142138