All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
PeeringBench - First ever peering bench!
ManofServer
Member
Hi guys,
Long time lurker here and user of MANY low end boxes - usually dedicated servers in my case. Was never a fan of VPS...
Anyways, for a long time I had a suspicion that the things we know about the "network quality" of certain providers shouldn't be taken at face value. You know, sometimes providers that cost more than the others, don't always offer better peering - especially over longer routes.
So, since I had 4 of the big ones - Online, SyS (OVH), Hetzner and LeaseWeb at the same time, I decided to run a heck of a lot of different peering, latency and bandwidth benchmarks in order to more offer a better view of which providers is better at specific areas.
My website is PeeringBench.com (it's a bit long but I believe if you're shopping for a new box, it's definitely worth a read).
Would love to hear if my research helps any of you, and feel free to offer any suggestions!
If everything goes well, I would love to test more boxes, like from RedStation, FDC, Server.Lu, etc., etc.
Thanks!
Comments
Cool site but the red hurts my eyes >_>
So, who "won"?
The red is indeed hard to read...
Aight, will work on the colors and update the design today. Sorry about that, not very good at web design
Anyways, the page is like a long spreadsheet with more detailed images inside. It doesn't really offer a winner as all providers have a different "edge", although I guess I would say the most value was offered by the Online box, while I was the most amazed at the Hetzner's speed, although I didn't like the bandwidth limit of their boxes.
Fixed the red, now it's dark brown and much easier to read IMO
One thing cracked me up on the website...
You are aware that there are 51247 currently active ASN's and so your test unfortunately has zero value, as it measures a single or a handful ISPs only ?
That doesn't really make much sense. This is not a benchmark of the Internet, it's about comparing 4 of the most popular budget dedicated server choices, so it's a comparison between 4 of probably the most sold ASN's.
And you are comparing what @ManofServer ? Random destinations? How does it offer any value ?
If you need help with that. PM me.
Random destinations with a sample size of 1 using a benchmark system primarily aimed at the home ISP market, great idea not convinced on the execution.
@Clouvider, it still offers more value than simply looking at a provider's sales page. Plus, for example if you can see the bandwidth and latency results for Comcast in Boston, you can with some certainty still deduce what the bandwidth would be to neighboring states & other Comcast users (of which there are over 28 million in the US).
Plus, there is a torrent speed test, which probably measures the results of peering with hundreds if not thousands of different ASN's. The results from the torrent test are similar to what the random destination tests showed, so I think that still shows that good peering with major ASN's does translate into good peering with the smaller networks as well.
@sdglhm Is the offer free?
J/k I'll see what I'll do in the future and PM if needed.
@AnthonySmith - thanks for the reply! I wasn't very sure of the idea and how many users would find it useful. Will work on improving today, with a more concise ending conclusion (like in a real review), and yeah, when I come in the possession of the 4 some boxes soon, will do multiple runs of the same destinations (I guess 3 should be enough?).
I did intentionally targeted home ISP's, since I figure most people that run their own websites are going to target the B2C market, so that would be the most useful for them.
Looking Glass... Let the client run their own tests and determine what's best for them.
Yes. If you're willing to provide this for free. I'll help you for free.
I sincerely don't think this has a much effect but end customers may find it useful.
So... the idea is nice, but I see a few major issues with the execution:
If you mean 3 per day then yes.
How about writing a small software for client PC's (or MAC's, don't start the flame war) to install. They're contributing to this website by hosting the software and sending up to xx (example) tests per day from their location.
Let's assume that I'm visiting your website. You can use GeoIP to filter the unwanted result (accessible from menu) and show me the results which are from Sri Lanka users. Then end users will get more accurate view on their overall ISP performance with that server.
Also you can use customer base PING tests to see if the server is up or down for that location.
But... It makes sense.
One thing that was questionable to me... Why so much personal stuff? How is this relevant to the site?
And along that, what are we talking on the low 3 digits? McDonalds at $10/hr would still get you ~$1200 a month...
^ To all the above, guys this are seriously good suggestions and I'm already working on implementing some of them (the ones that I'm competent to do, which I'm afraid to admit is not as much as you'd think).
I sincerely appreciate the input, I wouldn't even think about half of the things mentioned - the desktop app, the colors, converting some of the text-based images into crawable text. This kind of makes sense why my adult business never took of.
@Pyr02k1 I didn't want to specifically write my niche, but I try to build, ummm, I guess they could be called adult oriented websites? That's why the design of PeeringBench is a bit like those TGP porn websites you visit, you know the ones with lots of scroll down thumbnail galleries. I guess it didn't quite fit as nicely as a spreadsheet as I wanted, so that's why some people find the page painful to consume.
Btw, I was talking about my online income, this is just a hobby for me, as I have the wrong college degree and work a job I hate. And I was kidding about the last part, I have a dog actually.
I was just making a joke to go with jars. Good to know that your total income isn't low 3 digits... That'd of been concerning
Very useful, add more please
How often do you test/update this? Seems unfair to designate a hosting provider as being inferior when network connectivity can change minute by minute, hour by hour.