Monitoring the Server Performance
You can monitor the server performance in two ways: using Logi Report Server Monitor or using an external method: JMX MBean API.
Select the following links to view the topics:
- Monitoring the Server Performance on Logi Report Server Monitor
- Monitoring the Server Pperformance Using JMX MBean API
Logi Report Server Monitor can show performance counters in graph (Line chart and Bar chart) and text mode.
To monitor the performance of Logi Report Server on Logi Report Server Monitor:
- In the left panel of the Logi Report Server Monitor home page, select to select any server node.
- Select the Performance tab and the performance chart of the specified server will be displayed.
- If you want to configure the performance chart, select the Graph Options button on the toolbar. In the Graph Options dialog box, set how many tick marks you need to display on the X axis in the Keep Last N Records text box. In the Y Axis Limit and Y2 Axis Limit text boxes, set respectively the maximum value on the left Y axis and the right Y axis. In the Interval text box, set the time interval (in seconds) the performance chart will use to get data and refresh itself.
Select the Clear Display button if you want to clear the current display of the performance counters in the performance chart.
To stop the current display of the performance counters, select the Stop button on the toolbar.
The following are the available counters:
|The number of the currently waiting reports.
|The number of the currently running reports.
|The number of the finished reports.
|Finished Report Pages
|The number of pages of the finished reports.
|Report Average Processing Time
|The average processing time of each report.
|Report Average Waiting Time
|The average waiting time of each report.
|Valid User Sessions
|The number of valid user sessions.
|Average Submitted Tasks per User
|The average number of tasks that each user has submitted since Logi Report Server started.
|The number of database connections.
- Make your application environment consistent with the JMX specification.
- In server.properties, set the properties server.rmiserver.enable and server.profiling.enable to true.
- Run your code. Here is an example:
ObjectName profilingName = new ObjectName("jinfonet:name=profileCounter");
ProfilingFactory ft = new ProfilingFactory("127.0.0.1",1129, "d:\\monitor");
ProfilingCounter profilingCounter = new ProfilingCounter(ft);
MBeanServer server = MBeanServerFactory.createMBeanServer();
ProfileNotifyListener notifyListener = new ProfileNotifyListener();
// Create an RMI connector and start it
JMXServiceURL url = new JMXServiceURL("service:jmx:rmi:///jndi/rmi://localhost:9999/server");
JMXConnectorServer cs = JMXConnectorServerFactory.newJMXConnectorServer(url, null, server);
In code part
ProfilingFactory ft = new ProfilingFactory("127.0.0.1",1129, "d:\\monitor"), the parameters serverHost and serverRMIPort are used to get server's ProfilingService remote object, parameter homepath is the path of the rmi.auth file which is used for RMI authority checking public ProfilingFactory(String serverHost, int serverRMIPort, String homepath).
- Check the monitoring result.
The available counters used in JMX MBean API are:
|Total Completed Tasks
|The tasks that are totally completed.
|Number of Reports per Minute
|The number of the reports running per minute.
|Total Number of Pages Exported
|The total number of the exported report pages.
|Exported Pages per Minute
|The number of exported report pages per minute.
|The tasks that have been run successfully.
|The tasks that have been run on-demand.
|Average Number of Pages
|The average number of report pages for each task.
|Maximum Number of Pages
|The maximum number of report pages for all tasks.
|Average Running Time per Task
|The average running time of each task.
|Maximum Running Time
|The maximum running time of the tasks.
|Average Waiting Time per Task
|The average waiting time of each task.
|Maximum Waiting Time
|The maximum waiting time of the tasks.
|Average Engine Time per Task
|The average running time of each task on engine.
|Maximum Engine Time
|The maximum running time of all tasks on engine.
|Average Concurrent Engines
|The average number of concurrent engines for each task.
|The maximum number of concurrent engines for all tasks.
In JMX MBean API, the data type of the profiling counters has to be Decimal. See a figure below of built-in data type hierarchy that you could refer for the return value: