5 ms·
Hundreds of milliseconds per TCP (or UDP) session. The CGNAT boxes are handling hundreds of thousands of flows. And they're part of failover clusters so each fl
by zlynx 8y ago
Hundreds of milliseconds per TCP (or UDP) session. The CGNAT boxes are handling hundreds of thousands of flows. And they're part of failover clusters so each flow has to be shipped to the other cluster members and acknowledged before starting the flow. And these things are often overloaded and adding tens of milliseconds delay to each packet passing through. Because whoever heard of an ISP spending more than "just enough" on hardware?
- dvfjsdhgfv 8y agoYou see, my point is that these milliseconds are negligible in the case of a website like Facebook that takes 20 second to fully load on a reasonably modern machine.
- zlynx 8y agoYou may already be behind an overloaded CGNAT then. Over IPv6 on Comcast cable, with the Edge browser, Facebook is fully loaded for me in 3 seconds. Surface tablet in power-save mode so this isn't some overpowered desktop monster.
- dvfjsdhgfv 8y agoThat's very interesting. I've just measured it in private mode, all plugins off, from the moment of clicking login till the time the page load indicator stopped running. 13 full seconds.
- zlynx 8y agoFor fairness I just did that myself. 4 seconds.
- ben0x539 8y agoI mean, we're probably not talking about a dozen milliseconds out of 20 seconds total, we're talking about extra milliseconds for every request or roundtrip or w/e constituting those 20 seconds. It's an extra cost that scales with the bloat you're decrying, it doesn't become insignificant next to it!
- zlynx 8y agoIndeed. When I use Firefox dev tools to open Facebook I see over 300 requests. 300! Many of those are loaded over the same HTTP/2 connection. And a lot of those are cached. But that's still a lot of TCP connections to multiple servers opening up and every bit of added latency slows it all down.