You can use this monitor to check the most important values for the SAP dialog system and the number of users logged on, broken down by application server:
You can find a more detailed display of the relevant times for the dialog system including the response times at the front end in the Entire System monitor, Dialog subtree.
The monitor contains the current values for the following attributes:
This monitoring tree element (MTE) provides information about the response time of the dialog service. The dialog response time consists of the period of time from the request of the dialog to the dispatcher through the processing, to the ending of the dialog in the dispatcher and the transfer of the data to the presentation layer.
It also includes the time used during the roundtrips for data transfer from the front end to the application server and back. On the application server, this time is shown as roll wait time, while it appears as GUI time on the front end. In the case of inefficient networks, such as WAN connections, this time can contribute significantly to the response time, although it does not use any resources on the application server, since the context is paged out, and the work process is released.
The response time is usually split into wait time and execution time. The SAP response time consists of the following components:
An alert here indicates a performance problem in your system. The problem can have many different causes, and it is therefore necessary to perform a thorough analysis. You should first ensure that the application server is not overloaded. This is measured, for example, using the statistics for CPU performance, system paging, and dialog work processes. Another possible bottleneck is the database. You should ensure that the database performance is adequate.
If the dialog response time generates an alert, but the response time for the standard transaction (see Standardized Response Time) is not reporting any problems, you should first check the number of users logged on (Users Logged On). If there are only a small number of users, the user of very resource-intensive transactions can, in extreme cases, significantly increase the response time. If this is the case, there is often no serious performance problem.
The network time is the time used in the network during the first data transfer from the front end to the application server and during the last data transfer from the application server to the front end. It therefore does not include the "roundtrips" (communication steps during a dialog step).
This monitoring tree element displays the dialog response time of the standard transaction. This standard transaction simulates the normal workload of a transaction by accessing data on the database and executing a range of ABAP function modules.
You can use the standardized workload to compare the performance of the dialog system under uniform conditions. In this way, you can avoid effects that would provide distorted dialog response times on certain servers (for example, if only a small number of users is active, and at least one user is using long-running transactions).
When they log on, users are automatically distributed to the individual servers depending on the dialog quality (response time and number of users logged on) and in accordance with the logon groups that you have created.
To start the monitor, follow the procedure below:
The workload monitor is set as the analysis method for the dialog response time and the network time. You can use this to analyze performance problems in great detail, down to the level of individual dialog steps of particular transactions.
For the Number of Users Logged On, the analysis method displays a list of the users that are logged on, broken down by application servers on the Currently active users screen ‑ including information about the terminals used, transactions used, and number of open sessions.