Copyright © TIBCO Software Inc. All Rights Reserved
Copyright © TIBCO Software Inc. All Rights Reserved


Chapter 6 Monitoring and Managing a TIBCO BusinessEvents Cluster with MM : Working with Cluster Explorer

Working with Cluster Explorer
Using Cluster Explorer, you can use functionality available at various nodes on the left, and you can view information about that node level on the right.
Navigating Cluster Explorer
To navigate the cluster explorer, do any of the following:
Starting, Stopping, Pausing, and Resuming Cluster Engines
You can also use the MM-tools utility to start and stop engines at the command line. See Deploying, Starting, Stopping PUs with the MM-tools Utility.
To Start, Stop, Pause, or Resume an Engine
1.
2.
Click the appropriate icon: Start, Stop, Pause, or Resume
3.
Purging Inactive Unpredefined Processes
To purge, click the Purge icon. The Cluster Explorer view is cleaned. For configuration related to the purge feature see Chapter 4, MM Metrics and Features Configuration.
Viewing Monitored Objects
To view the Monitored Objects details, select the Monitored Objects node from the Cluster Explorer. The Cache Overview page is displayed in the right pane.
The details of various processes are displayed. For example, name of the entity, count, gets, puts, and so on.
Executing Methods
For a table listing all the available methods, see MM Method Reference.
To Execute a Method
1.
From the Cluster Explorer, expand the methods node to the level at which you want to execute the method, cluster, process or agent. You see method group nodes.
2.
3.
Specify the argument or arguments according to the dialog that appears, and click Execute. Tooltips explain the values required for the arguments.
Generating Thread Analyzer Reports
Thread analyzer connects to hosts through JMX ports and collects the thread dump for each host. Thread Analyzer analyzes the thread dump to provide information such as the following:
Thread dump compression  Threads with same stack trace are compressed into one to provide a compressed thread dump report.
Deadlock analysis  A resource allocation graph identifies the deadlocks in each of the thread dumps. (Thread analyzer creates a wait-for graph for a given set of stack traces and analyzes the graph to identify “Circular Wait Conditions” (CWC)).
Thread Dump Summarization  Provides a detailed call flow summarization of the thread dump.
The name of the thread analyzer report uses the format: IPAddress_Port_x.y.log. The y element is the number of the file, 0-9. You can generate up to ten log files for each set of reports for a machine, and the default size of each file is 10MB.
The x element is used if you stop and start again. It is a number used to distinguish each set of reports, when multiple sets of reports are generated for the same host.
To Generate Thread Analyzer Reports
1.
2.
Host name and IP  Displays the name and address of the machine whose thread dump you want to get and analyze. (Thread Analyzer is always started on the server. Only a remote JMX connection is established with this host to obtain the thread dump). When not provided, it is assumed the host is localhost. MM connects to the JMX port configured in the site topology file.
Username and password  Enter the credentials (if any are required) used to connect to the JMX server running on the target machine. Note that these are not the JMX credentials and not those used to log on to the machine.
Report Directory  Specify where on the MM server to save the Thread Analyzer report. If blank or incorrect, the thread report is saved here by default: BE_HOME/mm/logs/thread-analyzer/.
Time Interval: The interval in seconds between thread dumps. For example, if you enter 10, a thread dump is obtained every 10 seconds.
3.

Copyright © TIBCO Software Inc. All Rights Reserved
Copyright © TIBCO Software Inc. All Rights Reserved