Earlier this week, we host a live virtual event on our organization's website that had higher-than-expected attendance. We had several complaints from users saying they either couldn't get into the event or couldn't click on links we were pushing in the chat because of 502 Gateway errors.
I am trying to determine what exactly went wrong to see if we need to move our AWS EC2 instance up a tier or just need to make config changes.
When I look at the PHP-FPM error log for that event, I am seeing that that child processes are maxing out.
However, when I look at the NGINX error logs, I see this error:
[crit] 4505#4505: connect() failed (101: Network is unreachable) while resolving
I assumed this was related to upstream server being busy and NGINX not connecting. But then I noticed that this error is showing up frequently in our logs -- even in today's log. However, during the previously stated live event, the error increased in frequency.
I see that it has something to do with changing IP addresses, but I am still not sure what it means or how to address it.
I do not see any errors related to PHP FPM maxing out. Just this error.
My question is can "101: Network is unreachable" be related to the upstream capacity issue? or does it only pertain to IP addresses?
