Architecture

dynaTrace Diagnostics ' open and high-performance architecture is one of its primary strengths. With software applications in general—and diagnostics software in particular—underlying system architecture determines a product's functionality, reliability, scalability—and ultimately—its impact on your company's success and operating margins.

KnowledgeSensors mark a transaction’s progress along its execution path and identify all transaction entry points (e.g., Java Servlet invocations) and method calls, as well as their sequence and nesting. For each transaction, the KnowledgeSensors record not only pure performance metrics (e.g., response times, CPU usage) of an execution path, but also contextual information (e.g., method arguments, return values, exceptions, log events, IO usage, network traffic, objects created, SQL calls, remote calls, and synchronization delays) in order to enable a precise root-cause analysis.

To support a smooth deployment and an easy administration, several KnowledgeSensors can be packaged into a single KnowledgeSensorPack™.

The Diagnostics Agents enable you to monitor and diagnose your applications using lightweight byte-code instrumentation without the need to change source code. Using a unique drop-and-go installation process, the Diagnostics Agents inject the KnowledgeSensors into the target application. The granularity of the code-level monitoring can be adapted easily to ever-changing diagnostic requirements on-the-fly without restarting the target application. Plus, all Diagnostics Agents can be deployed and managed from a central location and require only minimal system resources for sustained 7x24 operation.

The Diagnostics Servers gathers all monitoring and diagnosis data from all KnowledgeSensors, which have been injected into the target system by the Diagnostics Agents. All calculations, aggregations and synthesis operations - such as the creation of PurePaths, dashboard views or reports - are done centrally on the Diagnostics Server. This allows overhead to be sustained at a very low level of only 3-5%. This makes dynaTrace Diagnostics ideally suited for usage in load test and production environments, even in large clustered application environments.

The Diagnostics Repository stores historical performance data for forward- and backward-looking long range analyses. This includes the following types of analysis:

• High-level performance analysis
Determine the performance of your applications from a bird’s eye view. Heat field will also indicate if individual problems have occurred, while the average performance is still OK.
• Availability analysis
Determine your application’s uptime
• Service level compliance analysis
Determine your application’s service-level fulfilment
• Trend analysis
Recognize performance trends before end-users start to experience problems (e.g., for capacity planning, avoiding application downtime, etc.)
• Historical incident analysis
Review all incidents that have occurred within your system/application and review linked PurePaths that have been stored for the incident (also see Alerting)

By providing an intuitive, platform-independent user interface, the Diagnostics Client guides IT team members through the processes of monitoring and diagnosing application performance. Team members from around the world can access dashboards, detailed diagnosis data and reports in real-time via context-sensitive menus that minimize GUI navigation steps. Additionally, an interactive offline analysis of recorded Diagnosis Sessions enables developers to fully reconstruct application hotspots on their local development workstation in order to quickly understand a problem’s root cause. To simplify communication, recorded Diagnostics Sessions can be stored centrally on the Diagnostics Server or even exported to a potable file for simple transmission via e-mail.

Using its open command line and JMX APIs, as well as through its pre-defined updaters, dynaTrace Diagnostics can be integrated seamlessly into existing IT ecosystems such as enterprise management systems (e.g., IBM Tivoli), IDEs (e.g., Eclipse), load test tools (e.g., Borland SilkPerformer) and automated build-and-test systems.