MANAGING DATABASE

• US 20150278394A1
• Filed: 03/26/2015
• Published: 10/01/2015
• Est. Priority Date: 03/27/2014
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Li, Li, Liu, Liang, Qu, Junmei, Yin, Wen Jun, Zhuang, Wei
• Granted Patent: US 9,940,406 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/213 with details for schema evolution support
• CPC Class Codes: G06F 16/221 Column-oriented storage; Management thereof
• CPC Class Codes: G06F 16/2322 using timestamps
• CPC Class Codes: G06F 16/24554 Unary operations; Data partitioning operations
• CPC Class Codes: G06F 16/2474 Sequence data queries, e.g. querying versioned data
• CPC Class Codes: G06F 16/86 Mapping to a database

Abstract

The present invention discloses a method and system for managing a database. According to embodiments of the present invention, there is provided a method for managing a database, each item of data in the database being associated with a timestamp and a data point, the timestamps being used as row keys for rows of a table in the database, the method comprising: obtaining a behavior characteristic of a user based on a previous data access to the database by the user; partitioning columns in the table into column families based on the obtained behavior characteristic and system configuration of the database; and causing data in the database to be stored in respective column families at least in part based on the associated data point. There is further disclosed a corresponding system.

Claims

1. A method for managing a database, each item of data in the database being associated with a timestamp and a data point, the timestamps being used as row keys for rows of a table in the database, the rows defined on one of two dimensions of the table, the method comprising:

obtaining a behavior characteristic of a user based on a previous data access to the database by the user;

partitioning columns in the table into column families based on the obtained behavior characteristic and system configuration of the database, the columns defined on the other of the two dimensions of the table; and

causing data in the database to be stored in respective column families at least in part based on the associated data points.

2. The method according to claim 1, wherein the behavior characteristic indicates a time span of the previous data access, and the system configuration indicates a size of blocks in a file system associated with the database,

and wherein partitioning columns in the table into column families comprises:

generating at least one logical table from the table, the number of rows in the at least one logical table determined based on the time span; and

determining the number of columns included in each of the column families at least in part based on the number of rows in the at least one logical table and the size of the blocks, such that the number of the blocks occupied by data in the column family is minimized.

3. The method according to claim 2, wherein the behavior characteristic further indicates a data type of the previous data access, and wherein the number of columns included in each of the column families is further determined based on the data type.

4. The method according to claim 2, wherein data in the table of the database is partitioned into a plurality of regions to store, the method further comprising:

determining the number of the column families included in the at least one logical table based on a size of the region and the size of the blocks, such that the number of the regions occupied by data in the at least one logical table is minimized.

5. The method according to claim 2, further comprising:

maintaining an index for the table in the database, the index mapping identifications of the data points to the at least one logical table and the column families.

6. The method according to claim 1, wherein causing data in the database to be stored in respective column families at least in part based on the associated data points comprises:

causing data for the correlated data points to be stored in a same column family.

7. The method according to claim 1, wherein the database is a Hadoop database (HBase).

8. A system for managing a database, each item of data in the database being associated with a timestamp and a data point, the timestamps being used as row keys for rows of a table in the database, the rows defined on one of two dimensions of the table, the system comprising:

a behavior characteristic obtaining unit configured to obtain a behavior characteristic of a user based on a previous data access to the database by the user;

a column family partitioning unit configured to partition columns in the table into column families based on the obtained behavior characteristic and system configuration of the database, the columns defined on the other of the two dimensions of the table; and

a data storage managing unit configured to cause data in the database to be stored in respective column families at least in part based on the associated data points.

9. The system according to claim 8, wherein the behavior characteristic indicates a time span of the previous data access, and the system configuration indicates a size of blocks in a file system associated with the database,

and wherein the column family partitioning unit comprises:

a logical table generating unit configured to generate at least one logical table from the table, the number of rows in the at least one logical table determined based on the time span; and

a column number determining unit configured to determine the number of columns included in each of the column families at least in part based on the number of rows in the at least one logical table and the size of the blocks, such that the number of the blocks occupied by data in the column family is minimized.

10. The system according to claim 9, wherein the behavior characteristic further indicates a data type of the previous data access, and wherein the number of columns included in each of the column families is further determined based on the data type.

11. The system according to claim 9, wherein data in the table of the database is partitioned into a plurality of regions to store, the system further comprising:

a column number determining unit configured to determine the number of the column families included in the at least one logical table based on a size of the region and the size of the blocks, such that the number of the regions occupied by data in the at least one logical table is minimized.

12. The system according to claim 9, further comprising:

an index maintaining unit configured to maintain an index for the table in the database, the index mapping identifications of the data points to the at least one logical table and the column families.

13. The system according to claim 8, wherein the data storage managing unit is configured to cause data for the correlated data points to be stored in a same column family.

14. The system according to claim 8, wherein the database is a Hadoop database (HBase).