Node/Apollo - A connection attempt failed because the connected party did not properly respond after a period of time

Question

We have a Node + Apollo Server GraphQL API that calls multiple microservices. Under performance testing, we are suddenly hit with a lot of 500 responses with the message

A connection attempt failed because the connected party did not properly respond after a period of time, or established connection failed because connected host has failed to respond

Everything works fine under 'normal' loads, but when we increase request activity, these become more and more prevalent.

It seems the request does not even reach the resolver chain. APM Traces from DataDog show this:

and the requests terminate within 1.5 ms. This leads me to believe that the issue lies with the Node server and not the services downstream.

Any ideas for debugging or experiences with dealing with the error would be much appreciated.