500 errors from api gateway for large Ion responses

Question

I'm having trouble with 500 errors that I'm pretty sure are coming from the Jetty container.

My internal logging is showing that I am responding with a status 200 and a valid response.

Smaller responses work fine but larger responses are failing with a 500 at the client and I can see that 500 in the API Gateway logs.

We are using HTTP direct in this case so it's an HTTP API gateway.

All these attempts have failed and don't have any telemetry or logs indicating the true source of the failure.

To learn more I suspect we need to enable jetty logs to stop being swallowed and start appearing in the CloudWatch logs but I haven't have any luck in finding how to do this.

Are there instructions somewhere on how to enable or increase the level of logging from the Jetty container?

Answer 1

For anyone getting opaque 500 errors for large responses through IONs in the future, in other words the problem that I was reporting here, in my case the problem was a response header was > 10k which exceeds this documented limit here...

https://docs.aws.amazon.com/general/latest/gr/apigateway.html#limits_apigateway

I hope this saves somebody some time.