Show TOC

Procedure documentationConfiguring the CCMS Alert Monitor Locate this document in the navigation structure

Procedure

For PI, the Computing Center Management System (CCMS) provides a special alert monitor template called Process Integration.

You use this alert monitor to monitor the ABAP and Java components (including the Business Process Engine) of your process integration landscape centrally, and to identify different categories of system errors and application errors in the various interfaces and interface namespaces of the components involved.

Besides the information on the monitored components, the alert monitor also provides information on the qRFC queues and alert categories relevant for PI, as well as on adapter-specific errors.

Creating a System Group

You have to define the system group ExchangeInfrastructure_Systems to obtain monitoring information when you call the CCMS alert monitor Process Integration from the SAP CCMS Monitor Templates within transaction RZ20.

The definition of this monitor includes a rule for the uppermost node of the monitor tree. This rule determines the systems of system group ExchangeInfrastructure_Systems that are to be displayed.

This system group must be created and contained in the system from which all systems of your PI landscape are to be centrally monitored, that is, in the central monitoring server.

The system group is maintained with transaction RZ21. Two steps are required:

  1. Choose Create remote monitoring entry from the Technical infrastructure menu and maintain all systems of your PI landscape.

  2. Choose Maintain system groups from the Technical infrastructure menu and define the system group ExchangeInfrastructure_Systems consisting of the remote systems created before.

If the local system in which you configure your monitor is part of this group, you can include it in the group by specifying its system ID. You do not need any remote monitoring for this system.

Enabling Performance Alerts for Specific Sender-Receiver Combinations

To make performance-related alerts possible for specific sender-receiver combinations and display these alerts in the CCMS alert monitor, you have to do the following:

  • Activate the writing of performance data in the respective Integration Engine.

  • Maintain logical sender-receiver assignments in table ALXMBPFALERT.

Activating the Writing of Performance Data

Perform the following steps to activate the writing of performance data in your Integration Engine:

  1. Call transaction SXMB_ADM and choose   Integration Engine Configuration   Change Configuration  .

  2. Select category PERF, parameter MEASUREMENT_LEVEL, and set the Current Valueto 1 or higher.

  3. Select category PERF, parameter MEASUREMENT_PERSIST, and set the Current Valueto 1.

    For more information, choose the button in the Documentation column.

Maintaining Logical Sender-Receiver Assignments

Perform the following steps to maintain appropriate logical sender-receiver assignments in table ALXMBPFALERT:

  1. Use the table maintenance transaction SM30 to maintain table ALXMBPFALERT.

    Note Note

    You need the authorizations of the SAP_XI_ADMINISTRATOR role to perform this step.

    End of the note.
  2. Add a new entry to the ALXMBPFALERT table to activate the monitoring of messages.

  3. In the CCMS Alert Monitor Nodefield, enter the name for the CCMS node to be used to monitor or display the processing times of the messages (for example, <SenderService_ReceiverService_Application>).

  4. In the Sender and Receiver frames, specify the values required to determine the messages to be monitored under the specified monitor node or displayed in the PI Alert Monitor.

    Specify fully qualified names, since generic entries with an asterisk (*) are not supported.

  5. In the Alert Thresholds frame, specify the values that lead to an alert if they are exceeded by the average processing time of messages.

  6. Save your new entry.

    When you save a new entry, the corresponding node is created and regularly updated every 15 minutes. By default, the average processing time of messages also refers to the last 15 minutes.

    If required, you can use the CCMS node in transaction RZ20 to change this time period as well as the threshold values. In the PI Alert Monitor, these values are contained under the node   IEngine   Message Performance by Sender/Receiver   of the Integration Server system.

    Note Note

    Changes to threshold values in the ALXMBPFALERT table only take effect on the corresponding CCMS node if this node is newly created. For this purpose, you must use transaction RZ20 to delete any existing node beforehand.

    End of the note.
Configuring the CCMS Heartbeat Monitor for Java Components

Perform the following steps if you have installed a PI Java component, for example an Adapter Engine, and you want to monitor its heartbeat in the PI monitor of CCMS. The heartbeat will be implemented based on the GRMG framework of CCMS.

  1. Define a CRMG scenario for the Java component to be monitored.

    Any Java component to be monitored has to be defined as a GRMG scenario. You do this by writing a corresponding XML specification and importing it with transaction GRMG.

    A template for such a CRMG customizing file is provided with SAP Note 634771. Use this template and refer to this note for more information.

    When writing your XML:

    • Note that it is always the text between the desc tags that appears in the CCMS user interface.

    • Ensure that you fill in the other tags as well (just like in the example), since the GRMG framework needs all these entries.

  2. Call transaction GRMG to upload your XML document once you have finished and saved your XML specification.

    For a better display in the alert monitor, we recommend that you do this in the system that belongs to the Integration Server. However, you can also do it in the central monitoring system. In this case, you have to add this system to the system group ExchangeInfrastructure_Systems.

  3. Log on to the central server of the system to which you have uploaded your XML document (the server that hosts the enqueue service).

  4. Call transaction GRMG to start the scenarios.

Also use transaction GRMG if you want to stop the CCMS heartbeat monitor configured for a particular Java component while keeping other heartbeat monitors.

More information about GRMG: Activating and Customizing Monitoring with the GRMG.

Configuring CCMS Alerts for the Adapter Engine

If processing of a message on the Adapter Engine fails, the message state is changed to faulty. To notify administrators immediately of such processing failures, they can be propagated as alerts to CCMS. As CCMS is the SAP standard interface to third-party monitoring tools, these failure events may also become visible in non-SAP monitoring tools.

Configuration Steps

To make use of Adapter Engine alerts, you must configure the connection to CCMS as follows:

  1. Make sure that the AS Java on which the Adapter Engine is running is configured to be monitored by the central CCMS system.

    To do so, execute transaction RZ20 on your central monitor server. There, it should be visible in the CCMS monitor under   SAP J2EE Monitor Templates   All Systems  .

  2. Check whether the Adapter Engine is configured to report errors to CCMS. To do so:

    1. Start the SAP NetWeaver Administrator.

    2. Choose   Configuration Management   Infrastructure Management   Java System Properties   Details   Services   (alias: /nwa/sys-config).

    3. Select the service XPI Service: AF Core.

    4. Check whether the value of property CCMSEnabledis set to true (in lower case letters).

    5. If it is not, change the property value to true and restart the service (  Start & Stop   Java EE Services  ).

  3. Verify that the setting is effective. To do so:

    1. Start the SAP NetWeaver Administrator.

    2. Choose   Availability and Performance Management   Resource Monitoring   Java System Reports   (alias: /nwa/java-sys-reports).

    3. Select the report Monitor Browser.

    4. Show Predefined Local J2EE Views and All.

    5. Go to   <SID>   Server   Services  . You should find a node for the Adapter Framework.

Alerting

Once you have performed the configuration steps in the previous section, errors that occur during message processing in the Adapter Engine should show up in both the SAP NetWeaver Administrator of the AS Java on which the Adapter Engine runs, and transaction RZ20 on the central monitor server:

  • Start the SAP NetWeaver Administrator, choose   Availability and Performance Management   Resource Monitoring   Java System Reports  (alias: /nwa/java-sys-reports), select the report Monitor Browser, and navigate to the Adapter Engine node.

    Here you should find that the number of errors of the appropriate error category and error code increases when messages fail on the Adapter Framework.

  • Execute transaction RZ20 on the central monitor server and select the Process Integrationmonitor.

    You should find a node for the Adapter Engine here with the same error information as in the SAP NetWeaver Administrator.

Accuracy

The default polling period for the monitor is 30 minutes. If you want to have shorter polling periods, you can change them in the SAP NetWeaver Administrator as follows:

  1. Choose   Availability and Performance Management   Resource Monitoring   Java System Reports  (alias: /nwa/java-sys-reports).

  2. Select the report Monitor Browser.

  3. Select an arbitrary entry in the monitoring tree of the Adapter Framework node.

  4. Choose Configuration in the Details view.

  5. Choose Edit Configuration Group.

  6. Change the polling period to 5 minutes, for instance.

  7. Choose Save and wait until saving is finished (be patient, since this may take a while).

  8. Choose   Configuration Management   Infrastructure   Java Configuration Browser  .

  9. Navigate to   apps   sap.com   com.sap.aii.af.app   Monitoring Configurator   monitor-configuration.xml  .

  10. In the xml file, change the polled-by-template period to the same value as above, for example, 5 minutes.

  11. Save the modified xml file.

  12. Choose   Operations Management   Infrastructure Management   Start & Stop   Java EE Services   to restart the Monitoring service to apply the changes.

Displaying PI Alerts in CCMS

To display alerts from alert categories you defined for PI, you have to perform the following configuration steps:

  1. Call transaction RZ20.

  2. Choose   Extras   Activate maintenance function   to switch to change mode.

  3. Copy the CCMS alert monitor Process Integrationunder a new name.

  4. Select the top node of your new monitor and add a new rule node

  5. Select the rule CCMS_GET_MTE_BY_CLASS.

  6. As R3System specify the system ID (SID) of your central monitoring server.

  7. As MTEClass specify SXI_RWBAlerts.

  8. Select the display option (preferably Short name).

  9. Choose Continue to confirm your entries.

  10. Choose Save to save the definition of your new monitor node.