Compuware Strobe

Pinpoint Inefficiencies Causing Excessive CPU Consumption and Elapsed Runtimes

Mainframe applications provide essential back-end support for customer-facing technology. Mainframe application performance monitoring and analysis is key to reducing the time it takes to identify and fix performance problems and prevent elongated elapsed times for customers.

Strobe^®, Compuware’s modern performance management and analysis solution for mainframe applications, lets performance analysts pinpoint application inefficiencies causing excessive CPU consumption and elongated elapsed times to reduce hardware and software costs.

With Strobe, mainframe teams can:

• Locate and improve resource-consumptive database calls
• Identify tuning opportunities to reduce monthly software licensing costs (MLC)
• Resolve performance issues before they impact application service levels
• Apply analytics to captured data to produce reports showing high-resource consumption and applications that need tuning
• Measure activity of both online and batch-processing applications in single and multi-system environments
• Identify several I/O problems, including inefficient VSAM file buffering
• Create and schedule automatic, reusable measurements
• Automatically detect performance problems by using a rolling average of CPU consumption and elapsed-time metrics

The Value of Strobe

• Improves user code efficiency, transaction response time and throughput
• Enables new users to be more productive using a graphical interface, intuitive charts and graphs, as well as built-in intelligence
• Reduces complexity of mainframe application performance management
• Minimizes time and resource consumption
• Enables developers to focus on application performance management earlier in the application development lifecycle
• Integrates with industry-leading tools from BMC, AppDynamics and Dynatrace

Key Features of Strobe

Expose Hidden Performance Issues

Using iStrobe, a browser-based Strobe reporting and analytics interface, performance analysts can identify and view:

• Program statements consuming excessive CPU resources
• Percentage of CPU time used within modules and control sections
• Procedure names and statement numbers
• Complete lines of source code

• Treemaps of CPU usage and wait report details, enabling easy navigation to performance issues along with drill-down capabilities

An Easy and Repeatable Measuring Process

• Intuitive, easy-to-use measurement request form provides helpful target, sampling and reporting options and suggestions to help users of all experience levels easily create and submit repeatable Strobe measurements
• Proactively initiate a measurement when abnormal behavior occurs in a program

Global Monitoring SMF

Using Strobe’s Global Monitoring SMF collection, you can identify and initiate measurements on the jobs using the most CPU, running the longest and performing the highest I/O. Strobe will then provide “insight” reports with data collected from these profiles and present the most important tuning targets available in categories like COBOL, Db2, Language Environment and file access. These reports also let you see which CICS transactions, SQL statements and modules are using the most CPU.

AutoStrobe with Global Batch Monitoring

AutoStrobe boosts productivity and increases the flexibility of Strobe and iStrobe by making the measurement process predictable and repeatable. AutoStrobe includes Global Batch Monitoring, a feature that automates the detection of performance problems by using a rolling average of CPU consumption and elapsed-time metrics. When jobs run over, code quality isn’t at the level it should be. In that case, Strobe’s Batch Global Monitoring kicks off a Strobe measurement of the job so you can see what’s going on and correct it.

Using Strobe Insight’s built-in analytics, you can:

• Execute reports to find problems across profiles
• Query SMF data to determine what jobs should be measured
• View analysis for all major subsystem components

Db2 SQL Strobe Advisor

Strobe Advisor, accessible via iStrobe, enables performance analysts to:

• Identify access paths used by SQL statements and relevant catalog statistics
• Perform “What If” analysis, making changes to SQL statements to see if they affect Db2 access paths
• Determine what causes excessive CPU and/or wait time
• Receive catalog statistics and the number of GET PAGE requests for each SQL statement

IMS Transaction Profiling

With IMS Transaction Profiling via Strobe Advisor in iStrobe, you can:

• Capture data on all transactions regardless of the region they execute in
• Receive performance data on all IMS transactions whether or not they are executed in the measured MPR
• Identify the modules in an IMS preload list and determine if there was activity with them, allowing you to optimize the IMS Preload list and avoid a short-on-storage situation

Rolling 4-hour Average (R4HA) Reporting

Well-tuned applications allow more workload to be processed at less cost. Glean insight into your R4HA MSU usage with Strobe Insight soft capacity peak time reporting. Using SMF data, Strobe identifies the peak period and exactly what was executed during that time regardless of LPAR. Data is collected in Coordinated Universal Time (UTC), enabling analysts to collect more accurate measurements and focus on tuning jobs that can result in lower licensing charges.