New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Tiny Benchmarks to Compare Linode, Digital Ocean, Vultr and Wable
PHP 5.5.11 compiled:
3m16.052s on Digital Ocean 2GB
3m9.606s on Linode 2GB
1m50.630s on Vultr 2GB
2m15.880s on Wable 2GB
PHP source directory compressed (tar zcf):
0m15.037s on Digital Ocean 2GB
0m10.395s on Linode 2GB
0m7.649s on Vultr 2GB
0m8.826s on Wable 2GB
PHP source tar.gz decompressed (tar zxf):
0m7.987s on Digital Ocean 2GB
0m2.107s on Linode 2GB
0m1.735s on Vultr 2GB
0m1.831s on Wable 2GB
VM Generation:
45 seconds on Digital Ocean
7 seconds on Linode
64 seconds on Vultr
37 seconds on Wable
Price for 2GB 2Core:
$20/month on Digital Ocean
$20/month on Linode
$15/month on Vultr
$8/month on Wable (1 more extra vCPU included)
All VPS's have 2GB memory and 2 vCPUs.
Comments
Vultr looks pretty good.
That's pretty fast!
Vultr = best CPU performance
Vultr = best disk performance
Linode = fastest VM spin-up
Everyone else loses
Nice.
Vultr 2GB also has 2 cores or not?
All has 2GB and has 2 cores.
Hm.. cause you mentioned it for wable and not for the others.
Edit: Nvm.
20 customers in a openvz node and 200 in a ovz node will be a big difference. Wait wable fill up their nodes with customers and try again
I think based on using openvz at most low end budget hosts that that's a valid concern. However we've spent over a year working on this project (we BETA'ed it at SXSW in 2013, pics on our facebook) and then had a public BETA since then. We've got a ton of time and code invested into smart allocations of new vps and controlling resources to prevent noisy neighbor issues. We track CPU by vm, hostnode, time, client, and by type (disk related etc) http://d.pr/i/kMa7/16dmGje2.
We use second by second cpu usage and history to keep a model of each physical machines current and predicted usages and use this data to determine which is healthy, e.g. a server which is healthy now but usually has peaks at 3pm up to 80% total cpu wouldn't be considered healthy to deploy to etc. I can't share too much of our secret sauce, but I'm confident that we won't have noisy neighbor issues. We also have approx ~150 spare machines to use for Wable as needed without having to spend a dime on new hardware. So there would be no point or incentive for us to overload our nodes.
Give us the benefit of the doubt for now ;-)
It also depends on which linode he was put on. There is still some very old hardware on the linode platform.
Mun
What's "wable"? Google gives nothing
@Incero's new brand based off the EduVPS project.
Just a domain that gordon got as he was a domain squatter:
http://web.archive.org/web/20120303135009/http://www.wable.com/
For the company: https://angel.co/wable
Even the logos look almost the same!!
Incero's bundle based VPS hosting
I bought the domain back in 2009. Thanks for the love though.
wable 2G is the second.
But there's a big jump from Bundle #3 to Bundle #4, $8 to $24.
Anyone want to benchmark Ubiquity? https://www.ubiquityhosting.com/cloud
What about iwStack?
iwStack is Italy only right? the first three in the title are available in different locations, as is ubiquity so they have broader appeal maybe? not sure where wable is though... i guess just TX
It would be nice to see similar results from well known providers like BuyVM, INIZ, Prometeus & Ramnode. I'm aware of the benchmarks on various sites like ServerBear, but having something similar to what @serverian just posted would be great.
I'm curious, just how much is wable oversold?
good to know, thank you
iwStack is in America aswell.
Yes, it is in US too, but on SSD, not SAN so no failover, there are also other limitations, such as no virtual router, so, no private networking, NAT, etc.
This is first time I hear about Wable. Any reviews?
Wable doesn't have a need to oversell, quoting a previous comment from @incero.
" Incero said: We also have approx ~150 spare machines to use for Wable as needed without having to spend a dime on new hardware. So there would be no point or incentive for us to overload our nodes."
New product, so reviews will come. However here are some comments from the Behold the Wable thread. http://lowendtalk.com/discussion/24839/behold-the-wable/
http://lowendtalk.com/discussion/comment/545519/#Comment_545519
http://lowendtalk.com/discussion/comment/543535/#Comment_543535
http://lowendtalk.com/discussion/comment/536463/#Comment_536463
http://lowendtalk.com/discussion/comment/536088/#Comment_536088
Is that domain really worth >= 10k?
One important test missed - how weird does the name sound
So, Wable a clear winner in name weirdness. Hopefully this info will make you all rethink your purchases.
One other point. Will they all do the job the majority want them to do regardless of where they sit on benchmarks? Yes. So just pick whatever one has the prettiest colors on their website and go with that :P
So I've got a project and I've decided to give Incero/Wable a shot. Looks like an interesting concept, plus allows me to distribute resources differently. Well see how this goes.
https://www.petabyet.com/result/2014-04-21-607795e26a7dc88038fd9915dcd0561e/
Here's a petabyet of one of the VM's I started. 9.58/10.