New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Random CloudFlare Timeout 522 errors
Does anyone else have random CloudFlare Timeout Errors? I have site which receives literally almost no traffic, it's CPU usage is at 0.1%. Yet, from time to time I end up in CloudFlare Timeout errors.
The server is working perfectly fine, I can reach it with direct IP. It's only CloudFlare what makes this random errors.
Why is this happening and how do you solve it?
What are the recommended settings for nginx's keepalive_timeout when behind CloudFlare?
Also, is it possible that sendfile, tcp_nopush, client_max_body_size or accept_filter=httpready have any effect on this?
Again, without CF, the website has a 100% uptime, it's only when behind CF that the problems start.
Comments
Cloudflare was crap and continues to be crap.
Always.....they are stupid errors....I did not know how to resolve the error...So I just left it with the error/bug
Read this.
https://support.cloudflare.com/hc/en-us/articles/200171906-Error-522-Connection-timed-out
@mrtz I read that one a dozen times, but I couldn't do anything. Obviously there is no firewall, etc. issues, as it only happens randomly, usually works 99% of the time.
@noaman they are not just stupid errors, they are a full website down experience for users who run into them.
What can I do to monitor this on my side? Should I put an
mtrcommand in crontab and run it every minute and log it?Also, if anyone understands CF's infrastructure, do they connect to my site from different nodes, or only from the same ones? This anycast IP network is making me crazy, the same IP always returns a <20 ping all around the world, so it must be a different node from each location. But then how can I run an mtr, if the node is not even the same?
I've had (and have) a lot of sites with nginx behind Cloudflare with no issue whatsoever (other than my own ineptitude of course).
Try to use an uptime monitor (newrelic synthetics or one of the other countless services) directly to your server / site (not through cloudflare) and see if you can spot a pattern.
@zsero
I dont have any production websites so It doesnot matter to me
@mrtz good idea, I'll set up a direct one.
I suspect it might be network related, as I'm on iwstack's new NL zone, and it's new so they might have some network issues. But how can I find out if it's CF or iwstack NL? Should I mtr in crontab?