Inserting data into an in-memory distributed nodal database

• US 9,183,284 B2
• Filed: 03/06/2013
• Issued: 11/10/2015
• Est. Priority Date: 09/26/2007
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Barsness, Eric L., Darrington, David L., Randles, Amanda, Santosuosso, John M.
• Primary Examiner(s): Singh, Amresh
• Application Number:
  US13/786,581
• Publication Number:
  US 20130185302A1
• Time in Patent Office: 979 Days
• Field of Search: 707/737
• US Class Current: 1/1
• CPC Class Codes: G06F 16/254 Extract, transform and load [ETL] procedures, e.g. ETL data flows in data warehouses
• CPC Class Codes: G06F 16/285 Clustering or classification
• CPC Class Codes: G06F 16/35 Clustering; Classification

Abstract

A database loader loads data to an in-memory database across multiple nodes in a parallel computing system. The database loader uses SQL flags, historical information gained from monitoring prior query execution times and patterns, and node and network configuration to determine how to effectively cluster data attributes across multiple nodes. The database loader may also allow a system administrator to force placement of database structures in particular nodes.

Claims

1. A computer implemented method for pre-loading an in-memory database into memory of a plurality of compute nodes of a parallel computer system, the method comprising the steps of:

receiving a database structure to load into the in-memory database in the plurality of compute nodes;

determining an optimized data node mapping to cluster database attributes across multiple compute nodes, wherein clustering database attributes across multiple compute nodes comprises placing data corresponding to a first attribute for a plurality of records together on a first compute node of the computer system and placing data corresponding to a second attribute for the plurality of records together on a second compute node of the computer system;

clustering the database attributes based on a use flag in an SQL statement that indicates an anticipated access frequency of placing a record of the plurality of records in the in-memory database; and

loading the database structure into the in-memory database with the determined optimized data node mapping.

2. The computer implemented method of claim 1 further comprises the step of determining the data is accessed often.

3. The computer implemented method of claim 1 further comprising the step of determining if there is a force location for the database structure in the in-memory database as indicated by a system administrator input.

4. The computer implemented method of claim 1 wherein the database attributes are further clustered based on a record flag in the SQL statement that initiates placing a range of columns of the data that should be clustered together in the in-memory database.

5. The computer implemented method of claim 1 further comprising clustering the database attributes based on node information comprising node identification, timestamp, current utilization future utilization and availability.

6. The computer implemented method of claim 1 further comprising clustering the database attributes based on network information comprising network identification, timestamp, current utilization future utilization and availability.

7. The computer implemented method of claim 1 further comprising clustering the database attributes based on query information comprising query identification, network used, elapsed time, node list and priority.

8. A computer implemented method for pre-loading an in-memory database into memory of a plurality of compute nodes of a parallel computer system, the method comprising the steps of:

receiving a database structure to load into the in-memory database in the plurality of compute nodes;

determining an optimized data node mapping to cluster database attributes across multiple compute nodes, wherein clustering database attributes across multiple compute nodes comprises placing data corresponding to a first attribute for a plurality of records together on a first compute node of the computer system and placing data corresponding to a second attribute for the plurality of records together on a second compute node of the computer system;

clustering the database attributes based on a use flag in an SQL statement that indicates an anticipated access frequency of placing a record of the plurality of records in the in-memory database;

loading the database structure into the in-memory database with the determined optimized data node mapping;

determining the data is accessed often; and

determining if there is a force location for the database structure in the in-memory database as indicated by a system administrator input.

9. The computer implemented method of claim 8 wherein the database attributes are further clustered based on a record flag in the SQL statement that initiates placing a range of columns of the data that should be clustered together in the in-memory database.

10. The computer implemented method of claim 8 further comprising clustering the database attributes based on node information comprising node identification, timestamp, current utilization future utilization and availability.

11. The computer implemented method of claim 8 further comprising clustering the database attributes based on network information network identification, timestamp, current utilization future utilization and availability.

12. The computer implemented method of claim 8 further comprising clustering the database attributes based on query information comprising query identification, network used, elapsed time, node list and priority.