Tyk_http_status_per_api prometheus metric

Hello Tyk fellows!

Do you know if a 500 for the tyk_http_status_per_api Prometheus metric also includes failures of the tyk grpc server and any tyk failures coming before the call to the backend service behind the API call?

Thank you!

I will try to make my question easier to understand with different scenarios. Let’s say we have a API request going through Tyk whose API definition specifies a few custom middlewares and one proxy (aka upstream service). Let’s consider these 3 scenarios:

  1. Tyk returns a 500 because one of the customer middleware failed
  2. Tyk returns a 500 because Tyk has an internal failure before the call is proxied to the upstream service
  3. Tyk returns a 500 because the upstream service returned 500

Does tyk_http_status_per_api has label code=500 in all the above cases?

I don’t see tyk_http_status_per_api as part of the base metrics. The status code should be there if specified in the custom metric

Hey @Olu I simply meant tyk_http_status.
All of my concerns above stay the same.
Thank you.

It will as long as analytics are generated for it and the correct pump details are configured.

I am sure about 1 and 3. For 2 I may have to confirm