If I have to select between Vultr high
frequency and high performance for a very dynamic website(buddyboss),
which one would you choose? And why? I asked support the difference and
here’s their replyhttps://d.pr/i/THULON.

Vultr’s just added High-Performance VMs (3rd Generation AMD EPYC CPUs). I’d love you to do the comparison with these.

“Has anyone done their own performance comparisons between similarly provisioned UpCloud servers vs. Vultr HF vs. Digital Ocean’s Premium AMD Droplets?

I know Vultr HF boxes are all the rage as of late, but I’m super intrigued by UpCloud’s offerings. And DO’s AMD Droplets also look promising. I’m planning on running my own batch of tests.

But if anyone else has done a shootout comparing something like a clone of a brochure/eCommerce/LMS website across those different hosting environments, I’m interested in your findings if you care to share.

I have an LMS site on a Premium DO droplet right now. And it slaps. But I’m going to clone it over to those other setups with similar resource provisions to see what performs best in a few different server load tests using Loader.io
https://loader.io/

Gijo at WP Speed Matters has a great tutorial on how to use Loader to simulate all kinds of traffic surges for testing purposes if anyone is interested in testing their own stuff.

Just don’t get yourself in trouble with your hosting provider’s billing if you cause yourself some overages when testing a bunch of stuff with this method. Hopefully hosts can filter out the bit traffic and note it’s not legitimate, human visits to your site. Fair warning though.”

With all ^that said, have you done any testing on similarly provisioned servers from UpCloud vs. Vultr HF vs. DO Premium AMD droplet?

I’ve recently started using GridPane and I’m planning on spinning up a few servers on the providers mentioned above. I’ll run the same kind of test like you described in your post here. Are there any other kinds of tests you recommend when using Loader.io?

Just curious if you’ve done any tests comparing these 3 specific providers anytime recently. Obviously, you’ve already compared Vultr HF to DO Premium AMD droplet. But I’m curious if there’d be any difference if the same batch of tests was run again with UpCloud in the mix for comparison.

When I run my tests, I’ll link my results here for anyone interested.

Yes, will do more tests

NGINX is the best option at this time?

Thanks!

I’d have to find the numbers somewhere :), but we ended up going with Vultr as well: performances were very somolar to DO’s, but some values ended up being a little better on Vultr.

(From memory, the only test DO was substantially better was one of the SSD read tests, cannot remember if sequential or random)

(Linode didn’t even made the podium: performances were substantially worse)