Hi,
I noticed the following strange thing. My Server is located in Singapore
If I do GTMetrix test for my site with caching ON using GTmetrix Server in Canada - I get the B score.
But if I use GTmetrix Server in Hong Kong (that is much closer to Singapore) I get F with caching ON but still B with caching OFF (Developer mode).
So the result without cache is much better. How is it possible? Any problems in CloudFlare network?
Hello,
I think this has to do with countries’ Internet connections.
Cloudflare selects the nearest server based on the number of routers and pings.
Try this site:
Don’t look at the single score but the individual metrics. I wrote a guide for pagespeed in era of Google Core Web Vital metrics at https://community.centminmod.com/threads/google-page-speed-insights-and-google-core-web-vital-metrics.20735/ that you might want to read.
Check here for more information provided by @eva2000:
Depending on your cache settings at Cloudflare dashboard, moreover there could be some re-routing in the time of testing or blockage due to execution of some script at your website.
Just make sure to have the best possible TTFB, while using Cloudflare you can really achieve it.
Keep in mind, on the Cloudflare dashboard you can set up optimization stuff like Minify JS, CSS, HTML, enabling the Rocket Loader, setting the cache, etc. as from the Free plan 
Other stuff, as already stated via the provided links from user replies to Google Page Speed, to consider the ones from it are the LCP and CLS.
But, keep in mind also the fact that you cannot achieve 100% if your Website has got images and some scripts, or even some advertisements.
Thank you for the answers guys.
DotMrCode Thank you for your link. Pingdom shows a similar result for all testing locations.
eva2000 thanks. I will read. I think anyway I need to improve the performance but first I want to understand this Hongkong issue. The Asian region has priority for me so if Cloudflare uses the wrong routing for Hongkong I want them to resolve this issue.
If I just had a bad result, I would agree that I need to look towards website optimization. But the problem is that the result in the case of testing from Hong Kong with caching is much worse (B-> F) than without caching from Hongkong or even with caching from Canada.
And this result is reproducible. I get it every time for testing from Hong Kong if caching is enabled and magically I get a good result if I disable caching. Since this happens every time, it’s not about the scripts that run on my server. It is not clear to me why I get different results when testing from different servers. If the problem is on my site the results should be similar, but they are not.
I still believe that the problem is on Cloudflarer’s side, but they stubbornly refuse to notice it.
Cloudflare’s headers will report which colo is serving the site for a given request.
Not all sites will be advertised from all datacenters on all plan types.
You could check how we manage our bandwidth here: The Relative Cost of Bandwidth Around the World.
This topic was automatically closed 5 days after the last reply. New replies are no longer allowed.