Upcloud vs DigitalOcean vs Linode vs Vultr vs Hetzner Benchmark Comparison

Being a benchmark addict, I had to feed my cravings so did a 13-way KVM VPS provider benchmark comparison tests for Upcloud vs DigitalOcean vs Linode vs Vultr vs Hetzner at https://community.centminmod.com/threads/17742/ :smile:

Enjoy :slight_smile:

10 Likes

Nice! Next one please with the German Comparison Line-up:

Hetzner
Netcup
PHPFriends
Contabo
FirstRoot

2 Likes

While I find the benchmarks interesting, I wish we could reliably determine expectation of continued performance without the data being too limited to conclude. Continued performance goes all the way to company policy and how it’s enacted, how they monitor servers, what their capacities are, etc.

You now how it goes, a fresh node always performs well but you don’t know how they’ll be at their worst until the node is full and they’ve actively dealt with abusive neighbors (how long did it take, what were the thresholds, etc).

That said, I rely a lot on hetzner cloud and it hasn’t failed me yet. A few minor downtime events but always top notch performance (at least in line with expectation and consistency).

4 Likes

I see that you’ve purchased my book, “Shitposting as a statement”. Just wait until you get chapter 5, “Dick pics: Not just for Instagram and Grandma.”

3 Likes

I miss Amazon Lightsail, Netcup and Ramnode in this benchmark.

Congrats to Upcloud (again), good to see that Vultr is catching up with the HF range.

1 Like

Just a small piece of advice (and this is solely my opinion), I think you should put the main takeaways and key findings in one of the very first sections of your benchmark summaries. The average person reading this will want to simply know the leading contender and what services packs the best punch for their dollar (and maybe more importantly which services performed poorly for the price and should be avoided).

Just something to think about. I’ve always been a BLUF (Bottom Line Up Front) person. So I tend to state the conclusions and main takeaways at the onset, then the reader can decide if they want to dive into the methodology and specific details that support the conclusion.

5 Likes

Thanks for the suggestions - I guess I am not the average joe, I like to read the specific details before seeing the summaries :slight_smile:

Indeed, VPS is only as good as it’s current performance - long term consistent performance is much harder to do properly on a massive scale though as $$$ budget and financial constraints would limit how many VPSes I could test long term heh. And indeed a fresh node definitely will perform better. We will see how Vultr’s new offering fairs long term wise.

Haha need widescreen viewing for extended tabulated results - though those VPS providers I don’t normally use besides Hetzner so aren’t usually part of my focus :slight_smile:

3 Likes

@Mason can I get a TLDR?

Upcloud and Vultr beated the competition

4 Likes

Although not really related to this thread, it’s funny that I say this today and I’ve had a server down for hours now. No response, no status page (they usually post for individual node issues).

Nice and surprising benchmark. Hetzner uses NVMe SSD in both of their offerings by the way.
Gotta give Vultr a try sooner or later.

1 Like

Hi @Jarland! What a weird coincidence… I apologize for the downtime! Did you get a response from our support team in the meantime? If not, you can send me the ticket number and I’ll investigate.

Also, sorry that you did not get informed about the issue with your node. I just checked Hetzner Status…there are a couple new reports since this morning. Anyway, we’re working on a better status system at the moment so that customers will be informed as soon as the issue appears. We’re doing our best to improve.
–Julia, Marketing

4 Likes

Yup pretty much :slight_smile:

oh didn’t know about that explains there nice disk I/O performance numbers :slight_smile:

1 Like

Yep all good. About 4 hours in the status page got posted :heart:

2 Likes

we need more benchmarks on Hostballs !

1 Like

I don’t think I want any of you testing ma balls… :thinking:

1 Like

don’t worry

just a gentle squeeze…

with a pipe wrench

1 Like

Oof, that wasn’t a pleasant experience the last time. PASS.

2 Likes

It’s OK. Just the once. You’re fixed now.

1 Like

Yes, thanks for bringing the right wrench

Now lay down and close your eyes…