Your data. Anywhere you go.

New Relic for iOS or Android


Download on the App Store    Android App on Google play


New Relic Insights App for iOS


Download on the App Store


Learn more

Close icon

How to recover monitor to JVM "10.147.33.139"


#1

###Java Agent Question Template

Is this a configuration issue?

  • Are you not seeing data? What frameworks are you using?

  • Describe what you are seeing:

  • How does what you are seeing differ from what you expect to see?
    Our APP " lrn-144722-feature @ bbpd.io" had only a host instance “10.147.33.139” before 5th Feb.
    Due to huge request pressure about 13:00 on 5th Feb, there are too many errors on the instance host “10.147.33.139”.
    newrelic_agent.log
    Feb 5, 2019 13:48:15 -0500 [3855 525] com.newrelic WARN: Dropped 179 error events out of 279.
    Feb 5, 2019 13:49:16 -0500 [3855 525] com.newrelic WARN: Dropped 291 error events out of 391.
    Feb 5, 2019 13:50:12 -0500 [3855 525] com.newrelic WARN: Dropped 230 error events out of 330.
    Feb 5, 2019 13:51:11 -0500 [3855 525] com.newrelic WARN: Dropped 244 error events out of 344.
    Feb 5, 2019 13:52:11 -0500 [3855 525] com.newrelic WARN: Dropped 244 error events out of 344.
    Feb 5, 2019 13:53:33 -0500 [3855 525] com.newrelic WARN: Dropped 414 error events out of 514.
    Feb 5, 2019 13:54:16 -0500 [3855 525] com.newrelic WARN: Dropped 164 error events out of 264.
    Feb 5, 2019 13:55:21 -0500 [3855 525] com.newrelic WARN: Dropped 280 error events out of 380.
    Feb 5, 2019 13:56:28 -0500 [3855 525] com.newrelic WARN: Dropped 233 error events out of 333.
    Feb 5, 2019 14:02:21 -0500 [3855 525] com.newrelic WARN: Dropped 138 error events out of 238.

From APM UI monitoring ->overview page we can not see data yet. So we do scaling up to our app on 12th Feb. The host instance was increased to 2.
After scaling up, now there are two hosts “ip-10-147-32-103.ec2.internal” and " ip-10-147-33-139.ec2.internal" on my apps “ lrn-144722-feature @ bbpd.io”.

We can see monitor data from the host “10.147.32.103”. but the monitor to host “10.147.33.139” can not be back normal. I think, like this sceanario,after scaling up , the monitor should be back normal automatically. Could you please tell me how to recover monitor automatically, not need to restart tomcat process?
And another question, i want to know whether there is a default value what the error rate is up to , the newrelic java agent will not report monitor data to the newrelic server and on the monitoring ->overview page we will can not see data. Thanks.


#2

Hi @elina.ma, thanks for posting your question here in the New Relic Community!

I think there are two different issues here, which may be unrelated. I’d like to first address the error event entries in the log files that you shared. These are not tied to the ability of the agent to report, but rather our cap on error event data which gets sent to New Relic in each harvest cycle. We limit the number of error events to 100 events/instance to keep payload size and network traffic down, but this isn’t tied to any errors that the agent is experiencing.

Regarding the JVM’s drop in reporting, it looks like the application was reporting, then stopped reporting, and then resumed reporting again. I assume this drop correlates to when you scaled your application instances up or down? Prior to the 5th, I actually see 3 instances reporting, then the period you mentioned with no instances reporting, and then one instance reporting. If the agent is not reporting when the Tomcat instance is spun up, it won’t start reporting later. If it is reporting, it should continue to report during the duration of that JVM. There isn’t a way to cause the agent to start reporting partially through the life cycle of a JVM without a restart, unfortunately. This is mostly due to the agent needing to perform it’s instrumentation of your application during the startup. Does that shed any light on what you’re seeing? Please feel free to follow-up with any questions you may have!

Best,
@jlangdon