Page History
...
The Load Balance tab includes the following:
| Anchor | ||||
|---|---|---|---|---|
|
The Load Balancing between JVMs table displays the following information for all instances:
Instance
- displays. Displays the instance name. If you select the highlighted instance, the execution tree will update itself to display the specific instance’s execution tree.
Info
...
Once a specific instance is selected, the Load Balance tab will no longer appear.
- Server - displays . Displays name of the server on which the instance is running.
- Avg. Response Time (Sec) - displays . Displays the average response time for the selected node. The average response time is displayed in seconds.
- Executions - displays . Displays the number of executions within the selected node.
- Sum. Response Times - displays . Displays the total response time for the selected node. The summed response time is displayed in hours: minutes: seconds: milliseconds.
| Anchor | ||||
|---|---|---|---|---|
|
The Response Time (Avg) overtime graph displays and compares the average response time of the selected instance in the selected time frame against the average response time frame for the AppTier. If the average response time is identical, the overtime graph will display information for the selected JVM on top of the AppTier average.
| Anchor | ||||
|---|---|---|---|---|
|
The Executions overtime graph displays and compares the average number of executions for the selected instance in the selected time frame against the average number of executions for the AppTier. If the average number of executions is identical, the overtime graph will display information for the selected JVM.
About the Impact tab
The Impact tab displays information regarding the selected method’s impact on the application. It displays general performance data in context for the selected node and specific information regarding its contribution to all entry points and methods in the application by which it is directly invoked. This provides the user with a comprehensive overview of the specific method’s performance throughout the application.
The Impact tab displays the following information:
| Anchor | ||||
|---|---|---|---|---|
|
The summary area displays the following information for the selected entity. The value displayed is the aggregated value for the selected entity’s underlying call tree in the selected time frame:
- Method (only when a method is displayed, not SQL) - displays . Displays the method type.
- Avg. Response Time (Sec) - displays . Displays the average response time for the selected node. The average response time is displayed in seconds.
- Executions - displays . Displays the number of executions within the selected node.
- Sum. Response Times - displays . Displays the total response time for the selected node. The summed response time is displayed in hours: minutes: seconds: milliseconds.
- Max. Response Time (Sec) - displays . Displays the peak response time detected in the selected time frame.
- Min. Response Time (Sec) - displays . Displays the lowest response time detected in the selected time frame.
- Avg. Work Time (Sec) (not displayed for SQLs) - displays . Displays the average actual work time of the selected node without the underlying call path.
- Max. Work Time (Sec) (not displayed for SQLs) - displays . Displays the maximum work time.
- Avg. CPU Time (Sec) - displays . Displays the average time that CPU was consumed for the selected node and its underlying call path. (Not displayed for SQLs).
- Max. CPU Time (Sec) (only when a method is displayed, not SQL) - displays . Displays the peak CPU time detected for the selected node.
| Anchor | ||||
|---|---|---|---|---|
|
The Impact on All Entry Points table displays the following information for every entry point that the selected method was invoked by in the selected time frame:
Entry Point
- displays. Displays the name of the entry point from which the entity was invoked.
Info
...
Selecting the hyperlinked entry point name selects the entry point in the execution tree and displays the highlights tab for the entry point.
- Work Impact on Entry Point - displays . Displays the percent value of the selected node’s summed response time out of its entry point’s summed response time.
- Method Avg. Response Time (Sec) - displays . Displays the method’s average response time.
- Executions by Entry Point - displays . Displays the number of times the selected method was invoked by the specific entry point.
- Sum. Method Response Times - displays . Displays the method’s total response time.
- Sum. Entry Point Response Times - displays . Displays the entry point’s total response time.
| Anchor | ||||
|---|---|---|---|---|
|
The Impact on All Direct Callers table displays the following information for every method that directly invoked the selected method:
Caller
- displays. Displays the name of the method that directly invoked the selected method.
Info
...
Selecting the hyperlinked direct caller name selects the method in the execution tree and displays the
...
Highlights tabs for the method.
- Work Impact on Caller - displays . Displays the percentage of the selected method’s total response time out of the direct caller method’s total response time.
- Method Avg. Response Time (Sec) - displays . Displays the method’s average response time.
- Executions by Caller - displays . Displays the number of times the selected method was executed by the specific direct caller method.
- Sum. Method Response Times - displays . Displays the method’s total response time.
- Sum. Caller Response Times - displays . Displays the direct caller method’s total response time.
About the Entry Points tab
The Entry Points tab displays the top level URI/iView or methods (EJB or others) that are monitored for this instance, comparing performance data of the detected entry points within the monitored application in the selected time frame. This view enables efficient identification of problematic entry points and, in the event of many entry points from the same instance showing distress, problematic instances as well.
...
The Avg. Response Time (Sec) vs. Executions overtime bar graph displays the relationship over time between the average response time of the entry point selected in the table above and the number of times the entry point was executed in the same time frame. Hovering over any point in the bar graph will display a ToolTip with the date and time of the time slice, the average response time, and the number of executions.
About the SQL & Exit Points tab
The SQL & Exit Points tab displays the following information for all SQLs and exit points in the call tree beneath the selected entity.
...
- Name - displays the destination name that can be either the Web service request name, 'DB Name' on 'Server Name' for SQLs, and so on.
- Type - displays the exit point destination type such as, Web service, DB type, and so on.
- Work Impact - displays the percentage of work time of the selected exit point in relation to the total work time of the invoking method
- Executions - displays the number of executions within the selected node.
About the Methods tab
The All Methods tab contains both URI/iView and methods from the entire call tree beneath the current context - instance. All data is aggregated from the underlying call tree.
...
The Avg. Response Time (Sec) vs. Executions overtime bar graph displays the relationship over time between the average response time of the exit point selected in the table above and the number of times the exit point was executed in the selected time frame. Hovering at any point in the bar graph will display a ToolTip with the date and time of the time slice, the average response time, and the number of executions.
About the Locks tab
The Locks tab displays locked methods from the entire call tree beneath the selected node. All data is aggregated from the underlying call tree.
...
The Avg. Lock Time (Sec) of the Methods Selected in the Tables Above overtime bar graph shows an overtime view of the periods where the methods selected in the tables above acquired locks. This view enables you to understand the mutual locking relationship of the two methods - are they always locked at the same times (meaning, they lock each other only), is one locking (indicated by a short lock time) and other locked (indicated by a long lock time), and so on. Hovering at any point in the bar graph will display a ToolTip with the date and time of the time slice, the average lock time, and the number of locks.
About the Exceptions tab
The Exceptions tab displays data for exceptions thrown from the entire call tree beneath the current context - instance. All data is aggregated from the underlying call tree.
...
See Precise Administration Guide, Appendix A, regarding configuring exception seeker options.
About the Leaks tab
The Leaks tab displays object allocation data from the entire call tree beneath the current context - instance. All data is aggregated from the underlying call tree.
...
You can identify a performance problem in Tree View by doing one or more of the following:
- Identifying slow Service Requests
- Identifying slow Methods
- Identifying the most invoked Service Request or Method
- Searching for a specific Service Request or Method
...
| Info |
|---|
The performance attribute by which the Tree View is sorted is shown on the workspace heading as well as on the ToolTip displayed when you place the cursor over a tree node. The number in brackets is the value of the entity according to its sort attribute. |
| Anchor | ||||
|---|---|---|---|---|
|
To identify slow Service Requests
- In the Time Frame list, select the period of time you want to analyze.
- In the Instance list, select the J2EE instance you want to analyze.
- Sort the tree by the performance attribute you want to explore (for example, Response Time (Avg) or Response Time (Sum). The default is Response Time (Avg), but you can change it using the Tree View Settings.
- In the Tree pane, select the Service Request you want to analyze. Focus on the Service Request that has a performance counter (for example, Response Time) that is higher than it should be.
- In the Details pane, look at both the Response Time graph and the Executions graph. Select a problematic time period to investigate and hover over it with the mouse. Additionally, analyze the details that are displayed in the Overview.
- On the displayed ToolTip, look for the time period.
- To view the problem in more detail, in the Time Frame list, select a narrower time frame which includes the time period that is displayed on the ToolTip.
- To investigate the root cause of the slow performance, expand the Service Request tree node to see its invoked methods. Continue with the Identifying slow Methods procedure.
| Anchor | ||||
|---|---|---|---|---|
|
To identify slow Methods
- In the Time Frame list, select the period of time you want to analyze.
- In the Instance list, select the instance you want to analyze.
- Sort the tree according to the performance parameter you want to investigate.
- Select the Method you want to analyze.
- Continue expanding additional tree levels until you can identify the slow Methods.
| Anchor | ||||
|---|---|---|---|---|
|
To identify the most invoked Service Request or Method
- In the Time Frame list, select the period of time you want to analyze.
- In the Instance list, select the instance you want to analyze.
- On the Precise bar, go to Settings>Tree View Settings.
- In the Tree View Settings dialog box, in the Sort by drop-down list, select Executions. Click A to display in Ascending order or D to display in Descending order.
- Select the Method you want to analyze. Methods with the highest number of executions are most frequently invoked.
| Anchor | ||||
|---|---|---|---|---|
|
To search for a specific Service Request or Method
- In the Time Frame list, select the period of time you want to analyze.
- In the Instance list, select the instance you want to analyze.
- In the Tree pane, right-click the Service Request, Method, or instance node to search under.
- From the popup menu, select the Search option.
- In the Search dialog box, specify the search parameters by inserting the string to search for and the number of results to display.
- Click OK. The results are highlighted in bold font and you can decide to focus on one of them by selecting it.
...