Page History
...
The JVM identifier that was chosen or given (depending on the installation type) during installation is referred to as JVMID. The JVMID directory is used to store JVM-specific settings. It is referred to within the file system paths, such as \i3\products\j2ee\config\JVMID. For example, if you choose PetStore as a JVMID, the JVMID directory would be \i3\products\j2ee\config\PetStore.
...
products\i3fp\registry\clusters\j2ee\<cluster-name>\<settings>.xml
| Anchor | ||||
|---|---|---|---|---|
|
Configuring the instance registry only affects the specific instance. Every element in the instance registry can also be specified in the server registry file, in which case it affects all the instances on that server.How to
To change the instance registry1. On
- On the relevant Precise node (proxy), manually change the products\i3fp\registry\instances\j2ee\<server-name>\<instance_name>\<settings>.xml file.
...
- Save the file.
...
- Run the
update-main-registryCLI command on the same Precise node.
...
- UNIX. ./infra/bin/psin_cli.sh -action update-main-registry -i3-user admin -i3-clear-password admin -registry-path /registry/clusters/j2ee/<server-name>/<instance-name>
...
- Windows. infra\bin\psin_cli.bat -action update-main-registry -i3-user admin -i3-clear-password admin -registry-path /registry/clusters/j2ee/<server name>/<instance name>
This will force a registry sync from the node's FocalPoint to the main FocalPoint.
| Anchor | ||||
|---|---|---|---|---|
|
Configuring the cluster registry affects all instances of the specific cluster.
| Info |
|---|
...
All instances in the cluster need to be restarted via AdminPoint after changing the cluster registry. |
To change the cluster registry:1. On
- On the relevant Precise node (proxy), manually change the products\i3fp\registry\clusters\j2ee\<cluster-name>\<settings>.xml file.
...
- Save the file.
...
- Run the
update-main-registryCLI command on the same Precise node.
...
- UNIX. ./infra/bin/psin_cli.sh -action update-main-registry -i3-user admin -i3-clear-password admin -registry-path /registry/clusters/j2ee/<cluster-name>
...
- Windows. infra\bin\psin_cli.bat -action update-main-registry -i3-user admin -i3-clear-password admin -registry-path /registry/clusters/j2ee/<cluster_name>
This will force a registry sync from the node's FocalPoint to the main FocalPoint.
| Info |
|---|
...
Cluster registry changes override the FocalPoint registry changes. |
| Anchor | ||||
|---|---|---|---|---|
|
This section includes configuration details for exception seeking and data filtering.
About configuring exception seeker options
Precise for J2EE catches exceptions during a time slice, based on its instrumentation configuration and filtering criteria. Filtering can be performed at the JVM level (that is: instrumentation and collector.xml level) and the GUI level to configure portlets.
Exceptions filtered at the JVM level will not be sent to the Precise for J2EE FocalPoint. View the exceptions on the Precise for J2EE Dashboard tab.
Configuring instrumentation parameters (JVM)
Perform the following steps to configure instrumentation parameters.
To configure instrumentation parameters (JVM)1. In
- In any Precise for J2EE tab, go to Settings>Monitor Settings.
...
- In the Monitor Settings dialog box, mark the “Monitor Exceptions” check box.
...
Restart the JVM.
Info
...
The changes take place at the instrumentation level. Therefore, the changes will only occur after restarting the JVM.
If you have a large amount of exceptions, focus on a specific exception by defining a filter on an exception class name (JVM).
To define a filter on an exception class name1. Open
- Open the collector.xml file in the registry folder. Use this file to define a filter on an exception class name, using regular expressions.
...
- Edit the following tag:
<exceptionFilterByName></exceptionFilterByName>
The default filter tag is:<exceptionFilterByName>.*</exceptionFilterByName>
...
- Run the relevant CLI command as it appears in step 3 of
...
About configuring data filtering
To reduce the quantity of collected data and propagate mainly meaningful data, a filtering mechanism is used. This mechanism is used to filter nodes and invocations before loading the data into the PMDB.
About the data filtering stages
The Precise for J2EE data filtering mechanism contains the following three stages:
1. Top N branches filtering - sorts each invocation tree by service time, only keeping the top n entry points (with their underlying call tree), as defined in the filtering configuration.
2. Filtering by service time - for each entry point, all underlying branches (including locked methods) with a service time lower than the defined percent of the entry point’s service time are filtered.
NOTE No information regarding the filtered branch’s SQLs or exit points will be displayed.
3. Filtering by work time - filters branches (only branches, and keeps their underlying call tree) with a work time lower than the defined threshold.
NOTE Locked methods are always displayed.
It is highly recommended not to alter the default data filtering values. Changes to the default values should only be made in extreme cases that either require maximum data, such as POC environments, or in cases of data overflow where stricter data filtering must be applied.
Configuring data filtering
To configure data filtering:
1. Open the collector.xml file in the registry folder.
2. Add the following element:
<agent-config>
<vendor>undefined (version undefined)</vendor>
<memloginterval>100</memloginterval>
<agginterval>30</agginterval>
<topnsql>5</topnsql>
<minSQLThreshold>100</minSQLThreshold>
<loggerEnabled>true</loggerEnabled>
<dirconnenabled>false</dirconnenabled>
<dirconnport>20764</dirconnport>
<exceptionFilterByName>.*</exceptionFilterByName>
<filteringEnabled>true</filteringEnabled>
<filteringMaxInvocationTrees>100</filteringMaxInvocationTrees>
<filteringPercent>1</filteringPercent>
<filteringThreshold>5</filteringThreshold>
<useStartupClass>false</useStartupClass>
</agent-config>
3. Run the relevant CLI command as it appears in step 3 of “Instance registry” on page 103 or “Cluster registry” on page 103.
The following table lists the available parameters:
Table 8-1 Data filtering configuration parameters
Parameter Description
filteringEnabled Activates and deactivates data filtering. Type: Boolean
Default value: true
filteringMaxInvocationTrees The maximum number of entry points to keep, based on service time. Type: Integer
Default value: 100
Note: For more information, see “About the data filtering stages” on page 105.
filteringPercent The minimum percent value of the entry point’s service time necessary to keep a node.
Type: Integer
Default value: 1
Note: For more information, see “About the data filtering stages” on page 105.
filteringThreshold The minimum work time value (milliseconds) threshold to keep a node. Type: Integer
Default value: 5
Note: For more information, see “About the data filtering stages” on page 105.
Keeping exceptions of filtered invocations
After selecting the data filtering method, the user can choose to mark the option: “Keep exceptions for filtered methods”. The exceptions will be included in the total per JVM, even though their source is not displayed.
...