Eivind Uggedal, autor do serviço gratuito de monitorização do website wasitup.com, publicou recentemente uma comparação exaustiva do desempenho dos fornecedores de cloud/VPS e dos seus pontos de referência ao longo do tempo. O post compara Linode, Rackspace, Slicehost, Amazon's EC2, e Prgmr, e testa o desempenho de Unixbench, SQL-bench, e tempos de renderização de páginas utilizando Django e PostgreSQL e SQLite.
Esta citação do artigo fala por si só: "Resumir os valores de referência dá-nos um vencedor claro: Linode". Pode ver o artigo completo aqui.
I got to know your services thanks to that article. Congratulations!
Damn, am I using an i686 or x86_64 system?
Um, anyway, you guys are the best, keep up the good work.
Switched to Linode from Slicehost thanks to that article =)
Love your servers!
Well deserved review, as you guys offer an excellent service!
I use Linode for personal web/email and we have a few for the company to.. and let me assure you all… linode rock the VPS world!!!
[…] employer) has been the subject of an extensive VPS performance review, and has come out as the clear winner. For those who don’t already know, a virtual private server (VPS) gives you the power of a […]
Waitaminut. Howcome x86_64 performance is so bad relative to i686 (which itself is relative)?!!! I recently provisioned a new VM with 64-bit b/c I thought the server would be better.
PS: Congrats, I knew I picked a winner.
x86 performs better on smaller nodes (360) our experience.
Anybody can explain Unixbench advantage on x86 and PostgreSQL on x86_64?
I wonder if Mysql has a similar performance advantage on x86_64 platforms as we use it for many database centric applications and extra performance is always welcome.
Also, in case of a Rails application, would rendering be faster on x86 like Uniqbench or not?
Just to say congratulations .. This is the most fantastic service I ever used 🙂
Keep up the goo work for many, many years!
Under x86_64, every pointer in a program is 4 bytes wider. This memory overhead can impede performance for a few workloads; however, 64-bit integer arithmetic is potentially quickened by an order of magnitude.
For crunching apps and fast string compares (database) and such, 64-bit can be better, particularly if you have a lot of memory to work in. If you’re running Apache in a load balancer, you’re not going to see a big improvement from a 64-bit wide architecture.
We always recommend 32-bit for most scenarios.
The x86_64 performance should in theory be better if the host is 64-bit itself, which I’m sure it will be for the demand on RAM required on the hosts!
I’m shocked at that result.
If you look at the actual article, you’ll notice that Linode performance numbers where spiking all over the place.
The only truly stable environment is Amazon EC2. You’ll notice Amazon EC2 graphs are all flat (which is a good thing).
Wavy lines, like Linode (and all other VPS) indicate who your “neighbors” adversely effects YOUR performance
Thanks for the plug of http://wasitup.com. The Djano benchmarks does not exactly test rendering times. It’s a run of the large Django unit test suite, testing amongst other things various rendering related functions.
Please, please open a Europe-based datacenter, or find a way to bring the latency between the UK and the NJ DC below 50ms.
I know performance with you guys is great (having been a customer for a year) but I’m finding it hard to justify staying with you when the latency makes it less responsive for me to work on sites.
I have been to Amazon EC2, Slicehost, and Linode. At the end I went back to Linode (and stayed for good). Now all of my 50+ web sites are hosted only between Linode and Google App Engine.
@Tim, more stable but less powerful and much more expensive.. I noticed that with all the “instability” Linode still performs better than Amazon 95% of the time. Also imo the support service here at linode is much better (even though I rarely need it).
On a different note, I wish the test had included FiveBean.. their prices are very tempting even though them using OpenVZ is a serious turnoff.
[…] on various startup/programming community sites (Hacker News, Reddit, etc) and got a mention on Linode’s blog. A small thread has been started on Linode Forums on migrating from SliceHost to Linode, when […]
I was a Linode customer in the past but went through some changes and ended up canceling my account. I will soon be in the market for another VPS and after catching that post on reddit, I will be re-opening a Linode account.
I am a current linode customer (2 nodes) since 2003. I feel the love with linode all the time. Service, reliability, and performance are great. I’m sticking with this winner!
Perhaps linode can come out with an “announcement” in February, as slice host plans on?
caker points to https://www.linode.com/2009/12/07/linode-expands-into-europe/
[…] Linode is getting a lot of buzz based on their performance/cost ratio. I recommend you take a look at them and see if they will work for […]
I actually feel lucky to be a customer… lol pretty rare! Keep up the great work!!!
Keep on rocking guys, I love to be a happy linoder 🙂
I found Amazon EC2 virtual servers to be very sluggish. They might be guaranteed speed, but from the looks of this, even a bad moment on Linode is better.
Yes, Linode rock!
I am pretty comfortable with Linode and this benchmark is one more reason for others to switch to Linode. I am now able to focus on the development the core area of mine without any server hassles
I am shocked at how bad the English is in comments posted here. You may be great with computers, but if you can’t write a proper sentence in English, you’re not going to get paid as well for your skills or work as effectively with others. In addition, bad English puts a strain on the reader, reduces the effectiveness of your communications, and reduces the likelihood that what you have written will be read or understood.
Take some time out from obsessing over *nix and the ‘net (sure, it’s fun) and learn to communicate. Linode will keep your site up in the meantime! Make more money and be more successful!
@Gregg, English is not the only language in the World (sure, is the most spoken but not all of us know enough).
ps. sorry for my bad English 😀
In the comparison page http://journal.uggedal.com/vps-performance-comparison the “Memory/$” don’t have sense… must be $/Memory 🙂
You guys do rock, service is excellent, fast, your servers are fast, the choices of distributions (Ubuntu for me), VPS choices, everything rocks.
If you get bought out, I hope the current owners get rich and the new owners have half as much common sense and business sense.
I bought a slice at linode instead of slicehost based on this benchmark. And so far it sticks.
Used the referral code from his site, might aswell support the work he has done 🙂
Great result, especially comparing to slicehost where i used to host my site!
Been using for more than an year – with out any glitches – the support is exceptional and the tools are the best!
Moved some of our web apps from Mosso and SliceHost to Linode.
I started with Slicehost but then I did some benchmarks on Linode and they crushed them, so I switched!