Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

>Whereas Cloudflare really are allowing people to deploy code at 100's of edge locations

Does cloudflare "really" deploy the code to the edge in any sense (e.g pushed to edge nodes, not just cached there)? I was under the impression this product uses some origin (likely, several) they manage, and their cdn.

Also cloudflare has a good number of POPs but I thought it was in the 100-200 range, not 200+



We have hardware in more than 200 cities in 100 countries globally: https://www.cloudflare.com/network/ It is very definitely above 200 locations. And we own the hardware there, we're not piggybacking on someone else's cloud and calling it our network.


Ah thanks, I stand corrected. That's what I get for relying on the Google featured snippet for "cloudflare number of POPs" for checking my memory.

What's the origin setup like for cloudflare pages?


They have a great breakdown on their blog[1] around the high-level request flow, but Workers run independently in each POP/colo. They aren't "faked" / running in some more centralized location.

Traffic only has to leave the edge if you want to fetch from _your origins_ behind it, but any responses the Worker generates itself are local to that edge POP. Return a static response, inspect/validate a JWT, modify something on the way out of the cache: all close to the user.

Disclaimer: used to work at Cloudflare, still think it's a great product.

[1]: https://blog.cloudflare.com/eliminating-cold-starts-with-clo...


Ah, thanks, yeah I know workers are legit at the edge, but I meant pages specifically, unless pages are somehow implemented on top of workers?


Pages is indeed implemented on top of Workers.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: