You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We are using the java agent 8.14.0 in our application. We observed that the start up of the application takes 2 minutes longer than before.
As a comparison, we tested 8.12.0, 8.13.0 and the newest 8.18.0, and found that starting from 8.13.0, the DockerData.java was modified and start to use the java.net.URL class to retrieve the docker id.
Although the log detail is hidden since 8.17.0, I still located it according to 8.13.0's log that when attempting retrieve the docker id, exception occurs at InstrumentationImpl::noticeTracer. The call of URL.openStream and the underlying call of HttpURLConnection.getInputStream cannot complete.
We also observed that the agent attempted twice to obtain the docker id by different uri, and each costed around 1 minute.
java.lang.NullPointerException: Cannot invoke "com.newrelic.agent.profile.ProfilerService.getTransactionProfileService()" because the return value of "com.newrelic.agent.service.ServiceFactory.getProfilerService()" is null
at com.newrelic.agent.instrumentation.InstrumentationImpl.noticeTracer(InstrumentationImpl.java:381) [newrelic.jar:8.13.0]
newrelic_agent.log(8.18.0)
XXXX-XX-XXT21:31:57 ... INFO: Attempting to fetch ECS Fargate container id from URL (V4): http://xxxxx
...
XXXX-XX-XXT21:32:58 ... FINEST: Exception in noticeTracer. This may affect thread profile v2. Tracer:null. Error message: Cannot invoke "com.newrelic.agent.profile.ProfilerService.getTransactionProfileService()" because the return value of "com.newrelic.agent.service.ServiceFactory.getProfilerService()" is null
XXXX-XX-XXT21:32:58 ... WARN: Error opening input stream retrieving AWS Fargate metadata
XXXX-XX-XXT21:32:58 ... INFO: Attempting to fetch ECS Fargate container id from URL (unversioned): http://xxxxx
...
XXXX-XX-XXT21:33:58 ... FINEST: Exception in noticeTracer. This may affect thread profile v2. Tracer:null. Error message: Cannot invoke "com.newrelic.agent.profile.ProfilerService.getTransactionProfileService()" because the return value of "com.newrelic.agent.service.ServiceFactory.getProfilerService()" is null
XXXX-XX-XXT21:33:58 ... WARN: Error opening input stream retrieving AWS Fargate metadata
Steps to Reproduce
It is not easy to share, but I will provide the detailed environment description.
Your Environment
New Relic Java Agent 8.13.0/8.18.0
AWS ECS (ECS_CONTAINER_METADATA_URI_V4 avaliable)
Docker image from amazoncorretto:21.0.2-alpine3.19
springboot(3.3.0) application
I tried to connet to the ECS task right after it starts up, and found that the accessibility to metadata uri is no problem.
Additional context
The text was updated successfully, but these errors were encountered:
Description
We are using the java agent 8.14.0 in our application. We observed that the start up of the application takes 2 minutes longer than before.
As a comparison, we tested 8.12.0, 8.13.0 and the newest 8.18.0, and found that starting from 8.13.0, the DockerData.java was modified and start to use the java.net.URL class to retrieve the docker id.
Although the log detail is hidden since 8.17.0, I still located it according to 8.13.0's log that when attempting retrieve the docker id, exception occurs at InstrumentationImpl::noticeTracer. The call of URL.openStream and the underlying call of HttpURLConnection.getInputStream cannot complete.
We also observed that the agent attempted twice to obtain the docker id by different uri, and each costed around 1 minute.
newrelic-java-agent/newrelic-agent/src/main/java/com/newrelic/agent/instrumentation/InstrumentationImpl.java
Line 381 in d492642
Expected Behavior
The exception happend during tracing should not interfere with the woven class's(java.net.HttpURLConnection) behavior.
Troubleshooting or NR Diag results
newrelic_agent.log(8.13.0)
newrelic_agent.log(8.18.0)
Steps to Reproduce
It is not easy to share, but I will provide the detailed environment description.
Your Environment
New Relic Java Agent 8.13.0/8.18.0
AWS ECS (ECS_CONTAINER_METADATA_URI_V4 avaliable)
Docker image from amazoncorretto:21.0.2-alpine3.19
springboot(3.3.0) application
I tried to connet to the ECS task right after it starts up, and found that the accessibility to metadata uri is no problem.
Additional context
The text was updated successfully, but these errors were encountered: