AWS ApiGateway metrics report 4xxErrors and 5xxError when there are none

I have setup an integration with AWS to monitor API Gateway metrics via CloudWatch and Kinesis Firehose, as described in the NR documentation. I noticed that the 4xxError and the 5xxError metric counts are following the Aws.apigateway.count metric almost exactly. To be precise, when calling the API gateway 6 times (200 OK response), the count metric reports 18 calls (why?) and 4xxError and 5xxError both report 12. Cloudwatch reports the correct metrics 6x OK and 0x 400 and 0x 500 errors. I have no idea what could be the issue. Any suggestion or help is appreciated.

Hello there and welcome to the community!

I am happy to create a ticket on your behalf with our infra team for better assistance with this error.

While waiting on a response it may be great to produce a screen shot of or permalink to the report you are seeing to offer our team more insight into what you are seeing.

Someone will be in touch soon!


As @jshearwood mentioned, if you still have your AWS integration connected to your NR One account, it would be helpful to get some additional details / screenshots and permalinks to what you are seeing in your account and we’d be happy to help further. :slight_smile: