The previous articles in this series have focused on how to deploy the application to IBM pureapplication System. After you deploy your application, you need to consider how your application stores and accesses the data. The relational database management system provides a standard data storage and retrieval interface, an advanced security model around data access, and concurrent support for multiple users accessing stored data concurrently.
If you use Pureapplication system, you can deploy an application in the form of an IBM DB2 software using an enterprise relational database management system (RDBMS). This article describes DB2, different ways to include DB2 in pureapplication system, and processes and best practices for deploying DB2 databases and enterprise applications together in the Pureapplication system.
DB2 in the Pureapplication System
IBM DB2 is an RDBMS product that supports different operating systems, such as Linux, UNIX, Windows, and z/OS. Specifically, the Pureapplication System contains DB2 Linux, UNIX, and Windows (LUW) versions of the product.
As of April 2012, the latest version of DB2 LUW is DB2 10.1. This new version introduces adaptive compression, further improve the existing data compression technology, the introduction of continuous data intake, without interruption of continuous data access and availability of the data table can be filled, but also introduced a time travel query, so you can view and easily report the data in the past a point in time. You can also find additional enhancements in DB2 10.1, including General query optimization and performance improvements, as well as improvements in security and workload management. The Pureapplication System contains DB2 10.1 and DB2 9.7, which prevent applications from being validated on newer versions.
The DB2 LUW 10.1 family's product portfolio contains different versions, from DB2 express-c (a free DB2 version with Community support) to the Advanced Enterprise Edition (with multiple additional features in accordance with the standard). However, because all DB2 versions of the DB2 LUW family use the same underlying database engine, any applications developed for a DB2 LUW version are applicable to all DB2 LUW versions. The DB2 virtual system model is available in the DB2 Express and Enterprise editions with different add-ons, while the DB2 database workload pattern provides you with a DB2 Enterprise version.
Select a database for applications in Pureapplication System
In the Pureapplication System context, you can deploy or configure a database in a variety of ways. Because IBM DB2 software is integrated within Pureapplication System, the use of DB2 as a database for a deployed application does not involve additional costs, reducing overhead and other licensing tracking mechanisms. This intrinsic advantage of the DB2 in Pureapplication System reduces the total cost of ownership for the platform. Furthermore, the unified nature of the DB2 in the Pureapplication System supports the application and adherence to the best practices and expertise that focus on integration when applications use DB2 as a database backend service.
DB2 Virtual System mode
Like other applications running on IBM pureapplication system, DB2 can be used as a DB2 virtual system mode or as part of a DB2 database workload pattern. As described in previous articles in this series, the DB2 virtual system model supports more flexible control and configuration of the middleware environment. Currently, multiple images can be deployed as DB2 virtual system mode:
DB2 Enterprise (primary node for high availability disaster recovery (HADR) features)
DB2 Enterprise (auxiliary node for HADR)
DB2 Express (main node for HADR)
DB2 Express (main node for HADR)
DB2 Database Workload Pattern
In addition to the available DB2 virtual system patterns, DB2 database workload patterns can be found in pureapplication system, where some configuration and best practices are applied to specific contexts. The deployment of the DB2 database workload pattern is simple, and it is flexible to change some configuration parameters in the database tier. For more information about the database workload standard features of the DB2 database workload pattern, see the Database Schema section. Currently, there are two different types of database workload patterns available for DB2:
IBM Transaction Database Schema
IBM Data mart Model
The IBM transaction database schema is used to accommodate departmental online transaction processing (OLTP) applications that do not require a high level of database optimization. This database workload pattern includes automatic configuration for departmental OLTP deployments, Virtual machine deployment tuning templates, and database backup schedules. The DB2 Enterprise version is used in this mode, and storage optimization is enabled for data compression.
Some of the features provided by the IBM Data mart model are essential for data mart infrastructures that equip and manage data-centric applications in Pureapplication System. The IBM Data mart model is tuned for the unique I/O throughput required by the data mart workload, which includes data compression and data mobility tools that are designed to help drive business with much-needed information. The DB2 Enterprise version is used in this mode, and storage optimization is enabled for data compression. This pattern contains SQL warehouse tools for creating and modifying physical patterns, control flows, and data flows for the Target data mart.
DB2 SQL Compatibility features
For users who have not previously used IBM DB2 software as a database solution, Pureapplication System is a good example of a production deployment that evaluates DB2 with an existing enterprise application. In addition to the advantages described earlier in using DB2 in Pureapplication system, the DB2 virtual system and database workload patterns also allow you to open SQL compatibility mode to help migrate applications written for use with other competing database software. When this feature is turned on, native SQL written for other competitive databases can be natively compiled in the DB2 engine without the use of slow-running simulation software. In addition, a compatible data concurrency model can be used, and DB2 contains tools that are compatible with existing scripts and personal skills, simplifying the transition to DB2.
Using remote databases outside the Pureapplication System
In some use cases, you may need to use an application deployed within the Pureapplication system to access and link the database systems on the remote system. For example, possible performance (perhaps performance) and other conditions classify a particular database workload into a mission-critical database class, so a dedicated physical system is required to host such a database system.
When you define a virtual application pattern, you can attach an existing remote database component to the schema. These databases will be located outside the Pureapplication System, and the configuration properties of these components will define the connection parameters for that remote database.