Show TOC

Background documentationAvailability Monitoring with CCMSPING Is Functioning Incorrectly Locate this document in the navigation structure

Structure

Problem Description

You want to use the availability agent CCMSPING to monitor the availability of systems and instances from the central monitoring system (CEN). You have therefore installed CCMSPING and configured availability monitoring as described in the documentation. However, one or more of the following problems occur during the monitoring:

  • The system to be monitored is not found in the availability monitoring configuration interface.

  • An ABAP system for which you have activated availability monitoring cannot be monitored. The value 0% is displayed for the system availability, and a message similar to the following error message appears in the CCMSPING log:

    ERROR: RFC error during logon: 102 RFC_ERROR_COMMUNICATION [Connect to message server failed

    Connect_PM MSHOST=bcomain, R3NAME=BCO, GROUP=PUBLIC

    LOCATION CPIC (TCP/IP) on local host

    ERROR hostname 'bcomain' unknown

    Note Note

    You can display the CCMSPING log:

    1. Start transaction RZ21, choose the CCMSPING Agents radio button in the Topology group box, and choose the Display Overview button.

    2. Select the destination of the desired CCMSPING, and choose the Agent Working Directory (Agent Working Directory) button. The files from this directory appear in a separate window. Double-click the log file ccmsping.log.

    End of the note.

  • An ABAP system for which you have activated availability monitoring cannot be monitored. The value 0% is displayed for system availability, and a message similar to the following error message appears in the CCMSPING log:

    ERROR: RFC error during logon: 102 RFC_ERROR_COMMUNICATION [Connect to message server failed

    Connect_PM MSHOST=bcomain, R3NAME=BCO, GROUP=PUBLIC

    LOCATION CPIC (TCP/IP) on local host

    ERROR partner '10.66.64.191:sapmsBCO' not reached

  • The availability tree is incomplete (for example, instance trees are missing), or the values are obsolete (grayed out).

  • You have activated the logon check (more information: Creating and Changing Monitoring Rules), but either the corresponding subtree is not created, or the reported values become obsolete. In the latter case, check the status of the data collector for the relevant nodes, by choosing the menu path   Views   Status Data Collector   in the Alert Monitor. The status Fatal Error is displayed.

  • Part of the availability monitoring with CCMSPING is the monitoring of hosts where the monitored instances are running, to help identify the cause of the error if interruptions occur in the availability. The corresponding nodes do not exist or are obsolete. In the latter case, check the status of the data collector for the corresponding node by choosing the menu path   Views   Status Data Collector   in the Alert Monitor. The status Fatal Error is displayed.

Scenario Type:

Error analysis

NetWeaver Component:

SAP Web AS ABAP

Valid for:

SAP NetWeaver 7.0 EhP 1
Decision Roadmap
Most Important Tools

  • For configuration, use the relevant interface in transaction RZ21 (menu path   Technical Infrastructure   Availability Monitoring   Configure CCMSPING Availability Mon.  ).

  • To display the availability data, use the Availability and Performance Overview Monitor in transaction RZ20 (Alert Monitor).

Analysis
System Is Not Contained in the List

If you cannot find the system to be monitored on the availability monitoring configuration interface, proceed as follows:

  • If you are monitoring double stack systems and standalone Java systems using the Java message server, you need to manually enter the connection data yourself by choosing   Monitoring   Add System   on the configuration interface. Then enter the connection data for the Java message server on the Connection Data and Monitoring Options screen.

  • If you are monitoring double stack systems and standalone ABAP systems using the ABAP message server, the connection data for the systems that can be monitored is, by default, copied from the sapmsg.ini file on the host on which CCMSPING is running. This file is used by SAP Logon. You can, therefore, manually add the system as described in the previous point, but you can also enter the desired system in SAP Logon on the host on which CCMSPING is running. Then upload the system into CEN by choosing the Load System List button.

Host unknown

In this case, CCMSPING could not create a connection to the host where the message server of the monitored system is running, and the log file ccmsping.log contains the message hostname <hostname> unknown (see above). This is a network administration problem with the following possible causes:

  • The host name is unknown in the responsible Domain Name System (DNS).

  • The host name relates to a virtual host

To solve this problem, create an alias for the relevant host in the hosts file of the host where CCMSPING is running (on Microsoft Windows XP, under [Drive]:\WINDOWS\system32\drivers\etc, on UNIX, under /etc).

System Cannot Be Monitored

If it is not possible to monitor an ABAP system that is to be monitored, this is usually because the services file on the CCMSPING host does not contain the relevant system. This file contains the port numbers of the individual ABAP message servers. Therefore, proceed as follows:

  1. Open the services file (on Microsoft Windows XP systems, under [Drive]:\WINDOWS\system32\drivers\etc, on UNIX systems, under /etc). The entry has the following format: sapms<SysID> <Port No.>/tcp.

  2. The required port number usually has the format 36<system number>. Enter the corresponding line in the services file, and save the file.

The availability tree is incomplete or obsolete

There are a number of possible causes for this problem. Firstly, the storage area of the monitoring segment, which contains the system-wide context Availability, may no longer be sufficient. To check this, proceed as follows:

  1. Start transaction RZ21, and in the Topology group box, select the Context Overview radio button. Choose the Display Overview button.

  2. Find the system-wide context Availability. You can determine which instance this context is on from the associated segment name.

  3. Start transaction RZ20, and, from the SAP CCMS Technical Expert Monitors monitor set, open the CCMS Selfmonitoring monitor.

  4. For the instance determined above, expand the subtree   <SysID>   MoniInfra_<instance name>   Space   and double-click the FreeAlertSlots attribute. This starts an analysis method that checks whether there is still enough space for the various node types and for alerts in the monitoring segment for the selected instance.

  5. If the values displayed indicate a space problem (there should still be sufficient space for every slot), increase the size of the monitoring segment (more information: Extending the Shared Memory Area), or move the Availability context to a segment with more space (more information: Monitoring Contexts in the Topology Display, command Activate Fixing).

If the incorrect display is not due to a space problem, check the connection data for the affected system in the configuration of the CCMSPING availability monitoring. To do this, proceed as follows:

  1. Start transaction RZ21, and choose the menu path   Technical Infrastructure   Availability Monitoring   Configure CCMSPING Availability Mon.  .

  2. Double-click the affected system and, depending on the system type, check the connection data for the ABAP or the Java message server.

  3. Check the destination of the responsible CCMSPING. A CCMSPING that was registered with the –push option, has a destination with the naming convention CCMSPING.<hostname>.<instance number>.

    Note Note

    We recommend that you only operate the availability agent CCMSPING with the option –push. If the destination CCMS_SYSTEM_PING_DEST is specified for your monitored system, you may have specified the wrong destination when adding the monitored system; in this case, correct the destination. The other possibility is that you are operating an obsolete availability agent or are operating the availability agent in an obsolete mode; in this case, reregister the agent (more information: Installing Availability Agent CCMSPING).

    End of the note.
The logon check nodes do not exist or are obsolete

If the logon check and the availability monitoring of logon groups are not being provided with data for any monitored ABAP systems, this is usually because the corresponding data collection method is not executed.

The logon check is performed by the data collection method CCMS_AVAIL_LOGON_CHECK, which, by default, runs in a dialog work process. This means that upper limits for the runtime apply, which can be exceeded if RFC problems occur. If the runtime is exceeded in this way, the status of the data collection method changes to Fatal Error, that is, the method is no longer executed, meaning that the values for the corresponding nodes become obsolete.

In this case, execute the method CCMS_AVAIL_LOGON_CHECK in the background. To do this, proceed as follows:

  1. Start the configuration of the CCMS availability monitoring by calling transaction RZ21 and choosing the menu path   Technical Infrastructure   Availability Monitoring   Configure CCMSPING Availability Mon.  .

  2. In the Monitoring dropdown list box, choose the Move logon check to batch option.

The host ping nodes do not exist or are obsolete

As with the problem The logon check nodes do not exist or are obsolete above, the cause of this problem is the exceeding of the maximum runtime for the method execution. You should therefore also run the responsible method CCMS_HOSTPING run in the background. Proceed as follows:

  1. Start the configuration of the CCMS availability monitoring by calling transaction RZ21 and choosing the menu path   Technical Infrastructure   Availability Monitoring   Configure CCMSPING Availability Mon.  

  2. In the Monitoring dropdown list box, choose the Move host ping to batch entry.

More Information

Availability Monitoring with CCMSPING