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

Difference in RPM of thorughput and error rate


#1

On the APM page, I am seeing different value of rpm in throughput and its almost double value of rpm in error rate for same time frame. I brought this up in future stack conference and was suggested to update the Java agent as it was ~2 years old. Now I have updated the java agent and yet seeing the same problem. Please suggest.


#2

Hi ,

I can take a look. Could you send us a permalink to the affected application? To create a permalink to any page within the New Relic APM UI, scroll to the bottom and click ‘Permalink’ - all the way on the right next to ‘Kiosk Mode. Once I have the permalink I should be able to see what’s going on.

Thanks!


#4

There you go.
https://rpm.newrelic.com/accounts/1986032/applications/162033025?tw[end]=1557961140&tw[start]=1557957540

Also attaching the screenshot here.


#5

Hi,

Are you taking the total of Web versus Non-Web transactions into consideration? Toggling between those two indicates a roughly equal amount of throughput. Could you add a screenshot here of the throughput value in the errors page that you’re using for reference?


#6

Hi Joey,
No I am taking only Web transactions. Attaching a new screenshot and highlighting what I am referring to.


#7

Hey @surabhi.jain the error rate will always show the combined throughput for web and non-web as the metric is actually /Errors/all, which does not distinguish between the two. So it doesn’t matter if you’re targeting web or non web transactions on the Overview page; that chart remains the same.

You can see that metric in the Insights data explorer, like:

https://insights.newrelic.com/accounts/1986032/explorer/metrics#entityId=162033025&entityName=MA1%20APB%20APP&searchTerms=%2FErrors%2Fall&entityType=application&entityTypeName=Applications&view=METRICS_OF_ENTITY&timeWindow=30&metricAgentIds[]=162033025&metricNames[]=Errors%2Fall&metricFacets[]=false&metricSegments[]=Errors%2Fall&metricFuncs[]=errors_per_minute

Hope that helps!


#8

Thanks for the explanation @netheroth . However this does not explain why we are seeing different value for request per minute in threshold and Error rate. In my other accounts where NR was installed directly with java Agent 4.5 , I am not seeing this problem. In the account where I am currently seeing problem was first installed on agent 3.43 and I have updated now to 4.11, but the problem is still there. I dont think uninstalling NR and installing again with latest agent would be the solution. Is there anything I need to check to validate my agent upgradation was all good.


#9

@surabhi.jain,

The RPM in the error rate chart includes both Web and non-Web transactions since the error rate is calculated based on both types transactions. The RPM in the throughput chart depends on which transaction type you choose in the overview page (Web or None-Web).