ALERT: Some images may not load properly within the Knowledge Base Article. If you see a broken image, please right-click and select 'Open image in a new tab'. We apologize for this inconvenience.

APM Websphere Agent high CPU load

book

Article ID: 194869

calendar_today

Updated On:

Products

CA Application Performance Management Agent (APM / Wily / Introscope) CA Application Performance Management (APM / Wily / Introscope) INTROSCOPE DX Application Performance Management

Issue/Introduction

 

High CPU load after agent upgrade on Websphere Agent has now consumption of 2-3 CPUs.
Java Version = 1.8.0_211 ( 64 bit )
Agent Stats:Build and Release = Release 10.7.0.197 (Build 992906) 
In attachment is IntroscopeAgentNoRedef.profile and used websphere-typical-NoRedef.pbl

 

Cause

 

Although smart instrumentation features had been turned off as part of the diagnostics, an AutoPersist.pbd for temporary instrumentation was still in the hotdeploy folder of the agent and was still instrumenting classes.

Even though entry point detection had been disabled in the agent profile (introscope.agent.deep.entrypoint.enabled=false), the AutoProbe log still showed entry point tracers inserted into classes

 

Processing class org/apache/soap/rpc/Call
        WASinvoke:0                        inserted method tracer object allocation: com/wily/introscope/agent/trace/hc2/EntryPointTracer
        WASinvoke:0                        inserted method tracer object allocation: com/wily/introscope/agent/trace/hc2/EntryPointBoundaryMarkerTracer

Environment

Release : 10.7.0

Component : APM Agents

Resolution

The AutoPersist.pbd was removed from the hotdeploy folder and CPU returned to expected levels.

The contents of the pbd can be reviewed manually to see if there is any instrumentation there that could be added into a custom pbd

Powered by Wolken Software