Mainframe > Administrator > Systems Management

Jump-Start Enterprise Modernization Efforts

Optim provides an integrated database-development environment.

The tighter everyone’s belt gets, the more the topic of conversation centers on reducing costs and improving ROI. Yet economic downturns can provide real opportunity for competitive differentiation, and enterprise modernization efforts can help your organization pull ahead of the pack. If you’re looking at Java technologies as part of your modernization process, or if you’re already using modern platforms such as Java, you’re probably being asked to come up with ways to save additional money and to get by with the same or even fewer staff and hardware resources. In this article, I’ll overview how integrated data management can accelerate the delivery of applications, upgrades and enhancements to support business growth and resilience. At the same time, this approach enables more efficient workloads, thereby reducing CPU usage and hardware expenses.

Accelerate high-quality solution delivery

A development environment that “understands” the database can help developers deliver better-quality, higher-performing software, faster. Java developers, who may have limited database-programming skills, can improve their database skills and develop better performing software with IBM Optim Development Studio (previously known as Data Studio Developer) into their Java-development environment. Optim Development Studio provides an integrated database-development environment and supports the DB2 family, Informix Dynamic Server and Oracle Database. This means skills are easily transferred across different OSs and database-management systems. It points out SQL errors while developers are editing and lets users run SQL without requiring a test program, which encourages Java database coding best practices and reduces time-consuming errors.

In addition, problem determination and impact analysis (which can be difficult or impossible with SQL that is generated from frameworks) becomes much easier with the capability to visualize linkages among the Java source code, the SQL (generated or not), and the database objects themselves—whether they are coded manually or generated from frameworks (see Figure 1). This allows database administrators (DBAs) and developers see the relationship among Java queries and database queries. They can assess the impact of change (and let’s face it, things are always changing)—whether the change is in the database schema, the application or both, and respond more quickly and with less risk to changing requirements.

Better performance equals less cost

So I’ve summarized agility and responsiveness to change, but it’s performance that has a direct relationship to IT costs—whether for software licensing costs on the mainframe or whether more hardware has to be purchased to support a growing application. Especially when times are tough, it’s important in every stage of the data-application lifecycle to squeeze the most out of existing resources. Even early in development or during application maintenance, developers can use Optim Development Studio to find potential SQL performance “hotspots” quickly and easily. Optim displays elapsed time costs, the frequency the queries are run and other cost information (see Figure 1). Additionally, developers can do first-level query tuning themselves using advice from Optim Query Tuner (available for DB2 Linux, UNIX, Windows and z/OS at this time)—which is easily invoked from within several natural points of Optim Development Studio. This makes query turning a more natural part of a developer’s day and helps them build their skill set. At the very least, developers will be able to do first-level analysis of potential performance problems on their own—reducing their dependency on the DBA and speeding time to delivery.

To reduce costs and lost business down the line, application testing is critical at every stage of the development process. Developers can create small test-data subsets from their own desktops using the copy and paste capabilities in Optim Development Studio. Alternatively, they can use Optim Development Studio to create scripts that can be input into Optim Test Data Management solutions for creating larger subsets of data for performance and stress testing, or for any test data in which data privacy rules must be adhered to.

Let’s move on to where performance really counts: production. Applications running DB2 and Optim pureQuery Runtime (or pureQuery Runtime for z/OS) can reduce CPU usage per transaction for access to DB2 up to 42 percent when compared to dynamic JDBC, as shown in Figure 2.

A key way to achieve double-digit performance enhancements is to leverage the full power of DB2’s static execution mode. Static execution mode means most of the internal overhead of determining how a particular query will access the database is handled prior to execution time (that is, the queries are prebound). This saves both CPU and network overhead (see Figure 3). For database-intensive applications that access for DB2 for z/OS data, this can mean significant savings in software-licensing costs, which can help any organization grow application throughput without purchasing new hardware. Using pureQuery Runtime, you can enable static execution seamlessly (with no change to the application) for either Java or .NET applications. This is the least-disruptive means to achieve this goal.

In addition, if you’re leveraging IBM System z servers using pureQuery Runtime, it’s possible to migrate more function to System z Integrated Information Processor (zIIP) and System z Application Assist Processor (zAAP) to lower costs of applications that access DB2 for z/OS data.

Tweet