Nginx subdomain down issue

Question

I have two rails app which are running on linode. OS is ubuntu, nginx server. The subdomain instance giving problem. It is getting down just after 1 day. On restarting the server, it is working fine.

The error log says- "*1 upstream timed out (110: Connection timed out) while reading response header from upstream,".

I googled for the problem and found that increasing proxy_read_timeout value will solve the problem. But I am unable to find the reason.

Is there an issue of over utilization of resources? I have 24 GB of storage and 512 MB of RAM as shown in linode manager. I have 10 cron jobs in total (5 in each app). They all start at the same time. Can that be the issue?

Please tell me the reason and solution for it.

Answer 1

It definitely sounds like a resource issue... Or perhaps something else is killing / hogging your app. Generally the upstream request is a request from the web server to the app server, so if your app is doing something wonky, this would cause the timeout to occur. I'm not sure what the default timeout is, but I'm guessing it's rather short. Increasing the timeout would at least buy you time to look at the system resources the process stack to try to figure out what's going on.