Hey there Chris, Thanks for mentioning my blogs; happy that they're helpful.
My 2 cents on this topic: 1. Just because a CDN supports HTTP/3 does not mean it enables it (by default) for all customers. For example, while Akamai (my current employer) provides support, it is explicitly opt-in instead of opt-out for existing customers, meaning we're probably at a much lower level of adoption than what's conceptually possible. I can't speak for other CDNs/large providers, but I imagine similar things happen there (e.g., only enable it by default for free/small customers, but require manual action from larger entities, which can be slow in the uptake). 2. Even if it's enabled, the alt-svc-based bootstrapping/discovery means you'll still have plenty of HTTP/2 by design. And while Chrome and Firefox pretty aggressively switch from H2 to H3 when discovered (even during a page load), Safari/Apple (at least for a long time, not 100% sure what they do not) kept using H2 for as long as possible, only switching to H3 when a new connection was really needed. This will (probably) be improved by HTTPS DNS records and alt-svc-bis, but those also have (current) deployment issues IIUC. 3. Even if the client tries HTTP/3, it will sometimes fail due to UDP rate limiting/blackholing/firewalling/flukes or things like MTU nasties that make HTTP/2 win the happy eyeballs race. I doubt this would contribute more than single-digit percentages to the overall number, but still a factor (I haven't seen recent numbers on this. Maybe Matt or Ian can comment, since they track more detailed client-side info?) 4. Outside of the browsers, there aren't many clients that actively support HTTP/3 (at least not by default, or in stable versions, see e.g., curl). Assuming again that say a single digit percentage of traffic (or should I say connections) is from non-browser clients, if that's counted in the stats, it'll also skew H3 potential. All that said, I must admit that I too have been surprised by Cloudflare Radar's numbers and their relative stableness over time :) Maybe Lucas can share some wisdom ;) With best regards, Robin On Mon, Jun 24, 2024 at 4:46 PM Chris Box <[email protected]> wrote: > On Mon, 24 Jun 2024 at 13:51, Aaron Ding <[email protected]> wrote: > >> we had a recent IEEE ICDCS work related to this subject: >> https://sk-alg.tbm.tudelft.nl/aaron/files/pre-icdcs2024.pdf >> >> > Thanks Aaron! > > Figure 2 is most illuminating. It shows that Google mostly serves HTTP/3, > but Cloudflare less than half H3, with all other major CDNs being mostly > H2. You also refer to https://w3techs.com/technologies/details/ce-http3 > which shows a slow upward trend from 26% to 30% in a year. > > So perhaps I should conclude that HTTP/3 adoption will significantly step > up when non-Google CDNs decide to deploy it more widely? > > In this space, Robin's APNIC posts are also useful: > https://blog.apnic.net/2023/09/25/why-http-3-is-eating-the-world/ and > https://blog.apnic.net/2023/10/23/the-challenges-ahead-for-http-3/. > > Chris > > -- Marx Robin +32 (0)497 72 86 94
