To control data flow to the application triage map for the purpose of assuring good performance, Introscope uses a threshold, set using the property introscope.apm.query.max.results (found in APMEnterpriseManager.properties), to clamp the amount of data the Workstation will attempt to display in the map. When you select a tree node or attempt to unroll a map element which would exceed this threshold, you see a message stating "The map is too large to display." You can adjust this threshold by adjusting the value of the property.
Note: Adjusting the value for the introscope.apm.query.max.results property may affect Workstation performance.
For more information about this property, see introscope.apm.query.max.results.
To edit the application triage map display threshold:
The application triage map data in your APM database can grow over time. Excessive data in the APM database can affect Enterprise Manager retrieval of application triage map data as well as Workstation and Enterprise Manager performance. You can prune application triage map data to avoid collecting excessive map data in the APM database.
You can configure how long map data remains in the database until it is pruned and how often you want to scheduling the pruning to occur. Map data pruning is controlled by a scheduler, which can be configured to prune data, on a daily, weekly, or monthly basis.
If application triage map data is pruned while a user is viewing the map, the pruned data remains on the map until the user changes the map time range, which triggers the map to be redrawn.
If you allow more than one Enterprise Manager to prune the database, multiple pruning processes compete for database access and create a database deadlock.
Follow these steps:
(Optional) Configure the following properties:
Note: Set the value of the introscope.apm.data.preserving.time property low to manage the volume of application triage map data in your environment, or exceptions can display. To prevent exceptions, verify that your environment is tuned in accordance with the application usage before you install the APM SOI connector. If you still see exceptions, increase the value for the introscope.apm.query.max.results property in the APMEnterpriseManager.properties file.
Restart the Enterprise Manager.
When the Enterprise Manager starts, you can go <EM_Home>/logs to review the Enterprise Manager log file messages associated with the map data pruning properties. For example:
11/23/09 12:02:13 PM PST [INFO] [Manager] [ The APM Data property : introscope.apm.pruning.enabled set to value : true] 11/23/09 12:02:13 PM PST [INFO] [Manager] [ The APM Data property : introscope.apm.pruning.cron.trigger.expression set to value : 0 0 6 * * ?]
CA APM refers to Application Triage Map objects and the relationships between them with the terms edge and vertex. The APM database stores data about these objects.
Over time, when connections between the Enterprise Manager and the monitored application are lost, edge objects “age out” according to two factors:
When edge objects age out, all associated objects are classified as obsolete if they are not used in any other active edge object. The Application Triage Map displays the elements associated with these obsolete objects as "greyed out" to make it clear they represent old, not current, data.
These "aged" objects remain in the APM database and are not automatically deleted. Objects are deleted only when the Application Triage Map Pruning process occurs, based on the properties that are set for pruning. The Pruning stored procedure physically deletes the objects based on pruning rules, which differ from Obsolete rules. For more information about pruning, see Configure Application Triage Map Data Pruning.
Rules for Application Triage Map object aging
(database table apm_edge)
|An Edge is considered to be obsolete when its update date is older than the value in the introscope.apm.data.obsolete.time property.|
|Vertex (database table apm_vertex)||A Vertex is considered to be obsolete if it is only referenced by obsolete Edges and no other Edges in the Application Triage Map.|
|Owner (database table apm_owner)||An Owner is considered to be obsolete if it is only referenced by obsolete Edges and no other Edges in the Application Triage Map.|
|Agent (database table apm_agent)||An Agent is considered to be obsolete if it is only referenced by obsolete Vertices in the Application Triage Map.|
You can return requests for obsolete objects from the database by configuring the Time Aging properties for the application triage map. If a time is not configured, then any request pointing to a time point within the period (introscope.apm.data.preserving.time + introscope.apm.data.agingTime - introscope.apm.data.obsolete.time) can specify the object as obsolete.
The Default values are:
To retrieve information on the obsolete objects, run the query Web Services APIs on November 1at 0800h and 1500h, and on November 2 at 1000h. This query provides the following results for each request:
The application triage map objects are marked obsolete based on the introscope.apm.data.obsolete.time property. If the introscope.apm.data.obsolete.time uses the default value 300 DAYS, then objects older than 300 days become obsolete.
Follow these steps:
View the message logs in the <EM_Home>/logs directory. For example:
10/19/12 12:02:13 PM EST [INFO] [Manager] [ The APM Data property: introscope.apm.data.obsolete.timeset to value : 2 DAYS]