App Agent Node Properties (M)

This property sets an upper limit for Analytics collectors in the agent.

Adjust this property to increase or decrease the number of asynchronous task registration requests allowed.

Adjust this property to increase or decrease the number of asynchronous task registrations allowed.

Sets a limit on the number of business transactions discovered once an agent is started. The limit helps to ensure that the Controller I/O processing capability and agent memory requirements are appropriate for a production environment. See ビジネストランザクション.

Warning: Changing this setting can affect the resource consumption of your deployment. Before you change this setting, verify that your application environment and Controller can handle any increased resource requirements.

This property represents the maximum number of elements that are collected for any call graph for a snapshot. When the limit is reached, the agent stops collecting more data for this call graph, reports what has been collected to that point, and marks the call graph with a warning that the limit was reached. AppDynamics does not recommend dramatically increasing this number as it may have overhead implications.

The maximum number of total snapshots that are allowed, including continuing transactions. When the queue goes over the value set, additional snapshots are dropped. This property is ignored in Development mode. 

 The maximum size of the correlation header.

A limit for the number of snapshots per minute due to errors. For example, if too many error snapshots are being seen, then tweak this value to reduce noise in the snapshots.

The maximum number of JDBC/ADO.NET exit-call stack samples per call graph. Only queries taking more time than the value of min-duration-for-jdbc-call-in-ms are reported. Changing the value does not require a restart.

The maximum number of JDBC/ADO.NET exit calls allowed in a snapshot. Calls after the limit are not recorded. Changing the value does not require a restart.

This property sets the upper limit to the number of metrics that can be registered by the agent.

Note: If the number of metrics—registered + unregistered—is greater than the new node property value, then the new metrics will not be created.

Maximum length of the SQL statement for SQL prepared statements. The length of the SQL statement is truncated if it exceeds the value set by this property. 

Maximum total number of service endpoints that can be registered for each asynchronous entry point type, such as worker thread. Because a single transaction may spawn many threads, you may expect more asynchronous service endpoint types than for synchronous service endpoint types.

Maximum total number of service endpoints that can be registered for each entry point type, such as servlet, struts action, web service, and so on.

Maximum total number of service endpoints that can be detected on a single node. Increasing the value of this property enables more service endpoints to be detected on a particular node.

Maximum number of service endpoints detected on a single thread of transaction execution.

If this property is set to the default value of one and two service endpoints are detected that impact one specific transaction, only one service endpoint will be evaluated at any time. If a second service endpoint is detected in the context of the first one, the second is ignored. But, if the second service endpoint starts after the first one ends, the second service endpoint will be evaluated.

Increase this property to monitor additional service endpoints on a thread. This number ensures a maximum limit on overhead and number of metrics due to service endpoints on each thread execution. 

This property increases the number of URLs the agent can track that produced a certain error. Once the maximum has been reached, all remaining errors are classified as unknown.

A JDBC/ADO.NET call taking more time than the specified time (in milliseconds) is captured in the call graph. The query continues to show up in a transaction snapshot. Setting this value too low (< 10ms) may affect application response times. Changing the value does not require a restart.

This property indicates the number of requests per Business Transaction to evaluate before triggering a diagnostic session. This property is useful to prevent diagnostic sessions when there is not enough load.

Automatic Leak Detection (ALD) tracks all frequently used Collections. For a Collection object to be identified and monitored, it must meet the conditions defined by the ALD properties. This property is the first criteria that needs to be met. The value is the minimum age of the Collection in minutes. This property takes effect after the node restarts.

From the point the collection is captured, it is monitored if it is still available for the specified period without collecting garbage. If it survives, then it is evaluated for size checks. If it meets the criteria, then it is monitored for long term growth in size.

Warning: If you reduce the default, there may be a performance hit on the CPU and memory because ALD needs to process more collections.

Automatic Leak Detection (ALD) tracks all frequently used Collections. For a Collection object to be identified and monitored for it must meet the conditions defined by the ALD properties. This property sets the number of elements threshold.

Warning: If you reduce the default there may be a performance hit on the CPU and memory because AD is processing more collections.

Automatic Leak Detection (ALD) tracks all frequently used Collections. For a Collection object to be identified and monitored it must meet the conditions defined by the ALD properties. This property sets the minimum initial size in megabytes for a collection to qualify for monitoring. The collection must also survive for the period specified in the minimum-age-for-evaluation-in-minutes property.

Warning: If you reduce the default there may be a performance hit on the CPU and memory because AD is processing more collections.

For Executor mode only, the asynchronous transactions do not get checked for stall unless they run at least the specified number of seconds.

By default, the .NET agent disables downstream correlation for MSMQ message queues. Register this node property on both the publishing and receiving tiers to enable downstream correlation for MSMQ and to specify the field where the agent writes correlation data.

The agent supports Extension or Label fields. By default, the agent writes correlation data to the Extension field, however, some frameworks built on MSMQ write data to the Extension field. Only use Label when the Extension field is not available because it is already in use by the framework. The NServiceBus implementation of MSMQ uses the Extension field, so for NServiceBus, use Label. See .NETのMSMQバックエンド.

Specify the threading architecture for the MSMQ message queue. The default value is false. For multithreaded queue implementations, change the value to false. See .NETのMSMQバックエンド.

Use this property to set the default log location of mtagent and argentoDynamicService logs.

Accepted values:

1. apm - This default value stores the mtagent and argentoDynamicService log files in the default APM logging directory along with APM logs.
2. non-apm - Use this value to store the mtagent and argentoDynamicService log files in the argentoDynamicService directory in external-services.

You can edit the property in app-agent-config.xml.