Target response time metric

I cannot copy\find the right query from this metric in AWS.
Is there a way to get it\ create it as a query.
Even only the maximum then i will be able to work from there.

Hi @yanir.cohen,

It looks like there is a dropdown under the statisics label at the top. Is there a max value there?

Hi ,
I do not understand the question, but yes there is a max value meaning the highest of each time.
I couldn’t find those metrics in new relic, they do not seem to exist
Please let me know how can i use those metrics,
The picture was taken from AWS

ELB metrics through the AWS integration are stored in Insights, So you can query and use operators on that data such average, min, max, etc if you are pulling in valid data from AWS through the integration

Ok, I tried this SELECT max(provider.targetResponseTime.Average) FROM LoadBalancerSample WHERE provider.targetGroupName = ‘4da14ddd-5667d43bb3361ea4517’ TIMESERIES since 3 hour ago

and this
SELECT max(aws.alb.targetGroup.targetResponseTime) FROM Metric WHERE aws.alb.targetGroupName =‘4da14ddd-5667d43bb3361ea4517’ TIMESERIES since 3 hour ago this is what i got

Now this is what i see in AWS which is clearly different

Is there a way to correlate this metric as well?
Is there a way to correlate alerts from cloudwatch?
First option is what i prefer.

From the screenshots, it looks like the graph is different. Try your query with a timeseries max if you don’t mind.

SELECT max(provider.targetResponseTime.Average) FROM LoadBalancerSample WHERE provider.targetGroupName = ‘4da14ddd-5667d43bb3361ea4517’ TIMESERIES max SINCE 3 hours ago

Sorry for the delay, While AWS may show/calculate the data differently, it does looks like Insights is following generally the graph in AWS.

Yes probably, but I can’t reach the other metric by performing calculations. it sometimes can be *5 and sometimes * 200.

Hi @yanir.cohen

Do you know what length period is being used for the datapoints in the CloudWatch graph? The information presented in New Relic will be scaled according to the number of “buckets” based on the size of the interval and downsampling retention. Some great articles that covers this in more depth can be found here:

Yes, 5 minutes is the number.

Hi @yanir.cohen

How does the graph compaer when specify TIMESERIES 5 minutes in your NRQL query?

Still not the same one of them is the average and one of them is maximum of the same time.
Is there a way to pull the one with the maximum of the same time?

Hi @yanir.cohen,

If that metric has max aggregations released from CloudWatch, it should be captured in the provider.targetResponseTime.Max attribute, rather than provider.targetResponseTime.Average.

Ok, but there isn’t one.
I will be fine thanks for your help btw.

No problem! Be sure to mark this as solved unless you have some other questions :slight_smile: