🔴 - To support my channel, I’d like to offer Mentorship/On-the-Job Support/Consulting - me@antonputra.com
@primux2007
Жыл бұрын
good job bro with this videos keep working i know you will success
@AntonPutra
Жыл бұрын
I appreciate it
@kamurashev
Жыл бұрын
Interesting, I was really surprised by the results. I would expect at least equal performance.
@AntonPutra
Жыл бұрын
Well it’s too early
@lucaskunze
Жыл бұрын
I tested last weekend, and I got the same result: HTTP/2 outperformed HTTP/3 by at least 100ms. I also expected better or at least equal results, but unfortunatelly that was not the result 😐. Let's wait a little more.
@kamurashev
Жыл бұрын
@@lucaskunze thanks for the update!
@TheSergWolf
6 күн бұрын
КЛАД!!!!! Спасибо огромное
@AntonPutra
6 күн бұрын
:)
@hashcode01
4 ай бұрын
Thanks for the benchmark Note that In scenarios where a website has numerous assets to load, like images, scripts, and stylesheets, HTTP/2 is better But for real-time applications such as streaming video or online gaming, HTTP/3's can provide lower latency and faster recovery from network disruptions
@AntonPutra
4 ай бұрын
Okay, thanks. I'll try to include more scenarios next time.
@bakhtiyargarashov805
Жыл бұрын
Gives very nice idea about performance. Thank you
@AntonPutra
Жыл бұрын
welcome!
@SergeiKliuikov
4 ай бұрын
Kudos for doing such practical research to spot the differences, but this test needs some tweaking. The biggest issue is that it assumes a stable connection. If we test it with a network with high packet loss (like a 3G signal or shaky WiFi), we’d see a whole different story. Another factor is dealing with high-res images or videos, like in an online store; the results would be totally different. Another snag is relying on the full page load times. This metric is heavily dependent on JS speed and the rendering engine, which doesn't really reflect the capabilities of the protocols. Lastly, the tests should be run in a highly competitive environment, with lots of users loading pages simultaneously. We should also check the traffic speed from the proxy because for HTTP/3, it’ll show lower values.
@AntonPutra
4 ай бұрын
Thanks for the valuable feedback. I'll take a look at how I can reproduce the scenarios that you mentioned!
@creo_one
Жыл бұрын
Infrastructure not diversified enough, how do you ensure that you're not measuring your noisy neighbors?
@AntonPutra
Жыл бұрын
I ran this test at least 3 times, by recreating VMs.. By the way you can find the ansible to reproduce in couple of steps.
@Babbili
Жыл бұрын
your videos are awesome
@AntonPutra
Жыл бұрын
Thanks Babbili!
@tinpa1756
Жыл бұрын
Hi, great video first of all. Secondly, I'm trying to recreate the test, but it seems like the playwright client doesn't use the HTTP/3 protocol. Whenever I try the endpoint in my Firefox browser I can see that it uses HTTP/3. Do you have any ideas why that might be?
@AntonPutra
Жыл бұрын
Thank you! No, unfortunately, it's been a while since I tested/recorded this video.
@KomentarSaya
7 ай бұрын
please do benchmark again in 2024, add openlitespeed in the list
@AntonPutra
7 ай бұрын
ok :)
@home-uf3tg
Жыл бұрын
I have no experince with grafana. anyone cloud you please tell me how can I import test dashboard after I login localhost:3000 ?
Пікірлер: 28