Reducing the number of similar requests

You can configure the App Visibility agent for Java to reduce the collection of similar requests, and in that way, lower the impact of monitoring on the application servers.

Each server request instance is counted and measured for latency violations and errors, and the requests are represented in the general metrics. However, the system monitors only certain server requests for extended viewing in the call tree. To monitor a server request, the request must have the following conditions:

Be included in the sampling mechanism
Have performance violations (latency), availability errors, or exceptions
Have a detectable reference in the HTTP header (meaning that another BMC product has requested monitoring)

You need to watch the performance and availability rates to ensure they represent a service level percentage for the application.

The following table lists the parameters in the Agent for Java Policy file that control the number of similar requests that the agent collects.

Tip

Press F to view the table in full-screen mode.

Parameters that control the number of similar requests

Parameter	Default value	Description
`persisting.single.worst.sla.breach. server.request.timeframe`	300000 ms (5 minutes)	Time, in milliseconds, in which only the worst SLA breaching requests are persisted
`persisting.error.server.requests. timeframe`	300000 ms (5 minutes)	Time, in milliseconds, in which error requests are buffered
`persisting.max.num.unique.errors.per. timeframe`	3	Number of different error types that are sent in the time that is specified in the `persisting.error.server.requests.timeframe` parameter

Example

persisting.single.worst.sla.breach.server.request.timeframe=300000
persisting.error.server.requests.timeframe=300000
persisting.max.num.unique.errors.per.timeframe=3

Page tree

Related topics