Upstream timed out 110: Connection timed out on elastic beanstalk

Question

I have a node.js server hosted on Amazon Linux 2/5.4.9 elasticbeanstalk. This API server is running fine for the past year. But lately I found that the server was in severe condition and had to be restarted first so that things got better. In the logs file I get lots of error codes like the following:
`2022/12/27 16:42:07 [error] 23659#23659: *7403 upstream timed out (110: Connection timed out) while reading response header from upstream, client: 172.31.24.78, server: , request: "GET /url-my-api HTTP/1.1", upstream: "http://127.0.0.1:8080/url-my-api", host: "api.mydomain.com", referrer: "https://mydomain.com/"`

I've never done any updates. Is there a solution that I should do?

Additional information:
1. Instance types : t2.small
2. Scaling cooldown : 180s
3. Servers live 24 hours

Cloud_G · Answer

Are you using cloudwatch for any metric monitoring of your elastic beanstalk environment or servers? Have you also tried to integrate x-ray to determine further latency problems?

Ismahel · Answer

Hello,

Are you using nginx and configured proxy_pass with upstream server. If so you have some issue on resolving DNS.
After a time, the upstream is not reachable and you have to restart to have access again. Is that the issue you have ?

https://distinctplace.com/2017/04/22/nginx-upstream-timed-out/

Upstream timed out 110: Connection timed out on elastic beanstalk

Relevant content