X hits on this document

Word document

SAP with Microsoft SQL Server 2005: Best Practices for High Availability, Maximum Performance, and ... - page 57 / 76





57 / 76

SAP with Microsoft SQL Server 2005: Best Practices for High Availability, Maximum Performance, and Scalability53

System performance is poor

When SAP system performance is poor, determine which component is having problems before checking the database server. In a three-tier SAP landscape, the database server might not be the immediate cause of slow system performance.

The first step is to analyze which of the SAP layers introduces the general slowdown. Use the SAP ST03 transaction to:

Check the performance of the SAP application servers.

List the different categories of the SAP workload (dialog, update, batch, and so on). Check the response times and the time spent in different stages of the overall system.

Check the response time of the database.

Check the time it takes the SAP application server to deliver the data to the SAP user interface at the end user’s computer (also listed in the ST03 transaction).

Check the aggregated data over several days and weeks.

ST03 Monitoring Transaction

Use the ST03 monitoring transaction to determine which SAP component is having a problem. ST03 differentiates between problems such as CPU consumed on the SAP application server, user requests waiting for the application server, database response time performance, and time to deliver data to the user interface.

Overall Performance (One Day)

The ST03 sample in Figure 18 shows the performance of an entire system of nine application servers for a one-day period. ST03 row entries show the different SAP process types in the grid. The columns show the measurements for different components.

The ST03 sample shows that the dialog process (in row six) has an overall response of 775.9 milliseconds (ms) for each SAP Dialogstep (SAP unit of work).

Figure 18

Microsoft Corporation © 2005

Document info
Document views158
Page views158
Page last viewedWed Oct 26 20:03:27 UTC 2016