I have a nginx server running as a container in a pod behind an ELB and in front of a uwsgi container (so client request -> AWS ELB -> nginx -> uwsgi).
My ELB times out at 60 seconds. Recently I noticed some requests log a 400 at the nginx layer after 59 seconds ... I'm not sure what could be the cause for the 400. The closeness of the timeout value seems interesting.
Has anyone encountered something similar before? Or have pointers on how to investigate?
It seems my error_logs log to stderr, so I'm not able to retrieve past error logs.