Bug - SampleCount Statistics invalid Lambda Error Metric values

Question

Hello Everyone, 
I am adding the total invocation count and error count of from AWS's Lambda metrics. 
I am able to see that the SampleCount metric of Error Count is showing the same as Lambda's invocation metric.
Example:
| Sl.No| Total Invocation| Total Errors| Error SampleCount Statistic (Invalid) | Error Sum Statistic (Valid)  |
| --- | --- | --- | --- |--- |
| 1| 10| 5| 10|5|

The Sum Statistic is working as expected but the SampleCount metric for Error is shows same as the invocation's SampleCount.

**Note: AWS has suggested to use only SUM Statistic **

Able to see an open question in stack overflow on the same: https://stackoverflow.com/questions/76012271/aws-lambda-error-observed-in-cloudwatch-metric-for-samplecount-statistics-for-l

Accepted Answer

Hello,

You mentioned the "Error SampleCount" statistic. This metric represents the number of data points collected for the ErrorCount metric during the specified period. However, it does not provide any additional information about the number or cause of errors.

**Some things to keep in mind:**

* ErrorCount will always be less than or equal to InvocationCount
* SampleCount just indicates how many data points were recorded, not the actual error values
* Look at logs/traces for details on specific errors
* Monitor ErrorCount/InvocationCount over time to check for changes in failure rate

Let me know if you have any other questions

Thanks

Abhinav

Bug - SampleCount Statistics invalid Lambda Error Metric values

Relevant content