Distributed database systems and methods of distributing and accessing data
AssigneeHuawei Technologies Co., Ltd.
InventorsYuanxi CHEN, Robin GROSMAN
Life of this patent
Estimating the term…
Abstract
A database system comprises a plurality of servers with data storage. The system holds first and second database tables. Records of the first database table are distributed among the servers according to values of a first distribution column. The servers can receive access requests and log workload data representative of the requests. Based on the workload data, the servers can determine a performance impact associated with columns of the tables and, upon determining that a performance impact associated with one of the columns is greater than a performance impact associated with the first distribution column, the servers can re-distribute the records among the plurality of servers according to values of the second column.
Claims20
A database system comprising:
- a plurality of servers each having interconnected data storage;
- first and second database tables, said first database table having records comprising a plurality of columns, said records distributed among said plurality of servers according to values of a first distribution column;
- said servers configured to:
- receive requests to access data in said first and second database tables and log workload data representative of said requests;
- based on said workload data, determine a performance impact associated with ones of said plurality of columns;
- upon determining that a performance impact associated with a second one of said columns is greater than a performance impact associated with said first distribution column, re-distribute said records among said plurality of servers according to values of said second column.
A method of re-distributing database tables among a plurality of servers comprising:
- receiving requests to access data in first and second database tables, said first database table having records comprising a plurality of columns, said records distributed among a plurality of servers according to values in a first column;
- writing workload data to a log, said workload data representative of said requests;
- based on said workload data, determining a performance impact associated with ones of said plurality of columns;
- upon determining that a performance impact associated with a second one of said columns is greater than a performance impact associated with said first distribution column, re-distributing said records among said plurality of servers according to values of said second column.
Dependencies are inferred from the claim text.