Root Cause Analysis with Solution Manager Diagnostics

The cross component diagnostics tools in SAP Solution Manager are based on a central diagnostics database that is filled with data by agents on the component systems.

E2E root cause analysis in SAP Solution Manager supports components implemented in ABAP, Java, or C(++), or that run on the Microsoft .NET framework. SAP Solution Manager standardizes, aggregates, and correlates the following functions:

Performance and resource metrics
Access to technical configuration
Exceptions (critical logs and dumps)
Traces
Transparency on changes to software (code), configuration
       or content

Exeption Analysis

The End-to-End Exception Analysis provides unified access to exceptions reflecting in high severity log entries and dumps. Component –specific log and dump viewers can be accessed directly from the End-to-End Exception Analysis.

The End-to-End Exception Analysis acts as a starting point for analyzing functional problems within a solution landscape. After identifying the exception that caused an incident, you can jump to the expert analysis tool for this kind of exception.

The End-to-End Exception Analysis also enables the support employee to correlate the different exception types like ABAP dumps, J2EE dumps, and the different application log files, and use this correlated information to find the exception that was the root cause for the incident.

After identifying the sought-after component and exception, component-specific analysis tools are used for in-depth component analysis, for example, the Log Viewer is used for the JAVA environment, while Transaction ST22 is used for analyzing ABAP runtime errors.

End-to-End Exception Analysis provides an overview of all exceptions in a Solution Landscape and helps to identify trends in the occurrence of exceptions. For each support component, there are special selected key performance indicators available that are most important for analysis.

The End-to-End Exception Analysis tool can be used for two different cases. The first use case is an exception trend analysis, which monitors the system behavior over a longer time period. This can clarify whether, for example, a single component in the Solution creates a particular high number of errors.

Alternatively, the exception trend analysis can also determine whether the solution has a higher rate of error occurrence after a change has been made to one of the solution´s components.

The second use case comes into play when the exact point in time of the incident is known, and the analysis is required to identify the specific component on which the error occurred. Here it is recommended looking at not only the exact point in time when the incident occurred, but also a few minutes before and after it occurred.

Workload Analysis

To identify the performance bottleneck, the End-to-End Workload Analysis is used. As soon as you can identify the part of the system that is causing the problem, there are additional tools for either Java- or ABAP-based analysis available.

In the area of Java the memory analysis gives you details about the Java memory management process and helps you to identify out of memory situations like application leaks. In the event of a long running request, the Java Thread Dump Analysis is used to find out the root cause of these issues, which may be a deadlock situation or blocked threads.

The integrated Wily Introscope dashboards are preconfigured by SAP in order to give you an overview of the current situation regarding different problems.

End-to-End Workload Analysis provides key performance indicators such as response times for different technical components in a single view.

Wily Introscope is designed to successfully manage a service-oriented architecture by providing real-time visibility into the performance and availability of critical web applications and infrastructure systems.

Java Memory Analysis shows statistical data regarding the garbage collection behavior and memory consumption of Java-based systems.

Thread Dump Analysis can be used to trigger a snapshot of the currently running threads in a Java Virtual Machine. The generated thread dump file is parsed in order to group and identify threads that are deadlocked or in other critical conditions.

The End-to-End Workload Analysis helps you to get workload information of your complete solution in order to analyze overall performance bottlenecks in your solution. There are different monitors and analysis tools that provide you with key performance indicators for different components. Most commonly, an initial check of the overall response time is done to check the overall workload.

For a more detailed analysis, the portfolio view helps to identify bottlenecks by correlating the average response time of the system with a high-accumulated response time. This indicates a performance bottleneck (high workload with high average response time).

In addition, for each separate component, component-specific KPIs, which can be checked for deviation from average values, are available.