TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Cloud server showdown: Amazon AWS EC2 vs. Linode vs. DigitalOcean

212 点作者 remcobron将近 12 年前

41 条评论

oogali将近 12 年前
<i>sigh</i><p>I call these &quot;weekend benchmarks&quot; -- what you&#x27;d typically do when you have a block of free time, then spend time optimizing for said benchmark. Roll in on Monday with some staggering results, only to find one (or many) of your variables were off.<p>Did the author try multiple instances on each provider? VM tenancy is a bitch. (Think of how annoyed you get at the noise levels, when your neighbor in the next apartment throws a party)<p>Is the author&#x27;s source benchmarking machine, a physical machine, or a virtualized guest? Does it have power savings turned off so that the process is running at 100% speed, instead of a variably-clocked down core?<p>Did the author enable or disable TCP time-wait recycle? So he doesn&#x27;t bump into said ceiling when running such tests back to back?<p>Did the author run the tests back to back, or have a cool down period between tests?<p>Where was the author&#x27;s network upstream located when he tried said tests? Were there any network issues at the time of the test? Would the author even be aware of them?<p>Your page you&#x27;re testing against, does the same database call, which is presumably cached. Did he throw out the cached results? Can he identify the cached results?<p>Are we firing up ApacheBench with HTTP keepalives enabled? With parallel connections? How many parallel connections?<p>How many Apache servers (StartServers, MinSpareServers, etc)? Which httpd modules were enabled? Which httpd modules were disabled? Which PHP modules were enabled?<p>You&#x27;re trying to benchmark CPU and I&#x2F;O horsepower across three different platforms but doing it through this narrow &quot;straw&quot; which consists of your &quot;independent server&quot;, your upstream, your upstream&#x27;s upstream, your upstream&#x27;s upstream&#x27;s peering connection with Amazon&#x2F;Linode&#x2F;DigitalOcean, your web server and its PHP module, your application, and MySQL.<p>If you&#x27;re rolling your eyes at this, then you shouldn&#x27;t be doing weekend benchmarks.<p>I&#x27;ll leave you with this as well:<p><a href="http://zedshaw.com/essays/programmer_stats.html" rel="nofollow">http:&#x2F;&#x2F;zedshaw.com&#x2F;essays&#x2F;programmer_stats.html</a>
评论 #5979881 未加载
评论 #5980676 未加载
评论 #5982069 未加载
评论 #5980351 未加载
评论 #5980381 未加载
评论 #5981266 未加载
mgkimsal将近 12 年前
Yes, amazon is &#x27;overpriced&#x27; if what you&#x27;re looking at is CPU or disk speed (which is not a bad metric, really). Where amazon shines is the amount of programmatic interaction you can do with it - have build scripts kick off an EC2 instance to run tests, send reports, shut it down. Programmatically bring up more instances during peak times, spin them down at night, etc.<p>AFAICT, Linode doesn&#x27;t offer that, and they probably won&#x27;t. Amazon&#x27;s been ahead in this arena for awhile, and will probably keep that lead for the forseeable future. EDIT: Apparently they do have an API which would cover a decent variety of use cases.<p>What&#x27;s sad is the number of people that migrate over to Amazon because it&#x27;s the done thing, without realizing what they&#x27;re paying for (and that they&#x27;re not utilizing the unique features of EC2).
评论 #5978956 未加载
评论 #5978768 未加载
评论 #5979403 未加载
评论 #5982997 未加载
评论 #5978778 未加载
PaulHoule将近 12 年前
The &#x27;small&#x27; instance he&#x27;s testing is the worst deal Amazon offers. If you buy certain instances you get can a better connection to EBS, and then you can allegedly buy better I&#x2F;O for $. A fair analysis is going to involve pages and pages of graphs and charts...<p>It is hard to beat AMZN&#x27;s flexibility and easy provisioning. For instance, even my sister-in-law can walk up to AWS, spin up an instance with Windows, log in with RDP and then enjoy everything you can enjoy on Windows other than high-end gaming. (For you mac-ers in the audience, this is a clean and economical way to make sure that your landing page works for &quot;the rest of us&quot;)<p>But you don&#x27;t have to choose because you can work with Linux. On top of that, AMZN layers services such as Elastic Map-Reduce which are compatible with industry standards but eliminate so much time you could waste sysadmining.
评论 #5979461 未加载
评论 #5986627 未加载
qeorge将近 12 年前
I&#x27;ve been able to make good utilization of Linode&#x27;s multiple cores for our Sphinx server (<a href="http://sphinxsearch.com/" rel="nofollow">http:&#x2F;&#x2F;sphinxsearch.com&#x2F;</a>)<p>In sphinx&#x27;s config files, its super easy to split an index up into 4 parts, and then assign each partial index to one core.<p>Doing this, we find that all 4 cores are utilized nicely (&gt;50%) almost all of the time. I feel like we&#x27;re getting a really good value out of that machine.<p>AirBnB has a good writeup on this config: <a href="http://nerds.airbnb.com/how-we-improved-search-performance-by-2x/" rel="nofollow">http:&#x2F;&#x2F;nerds.airbnb.com&#x2F;how-we-improved-search-performance-b...</a>
评论 #5979069 未加载
bdcravens将近 12 年前
EC2 always comes out behind in these comparisons. However, EC2&#x27;s benefit is the synergy of the platform: painless images, flexible volume management, VPC, programmatic DNS, queue services, Heroku style service, baked-in Chef support, easy RDBMS cloning and replication, etc.
uts_将近 12 年前
ServerBear does everything the OP is already trying to do:<p><a href="http://serverbear.com/237-small-amazon-web-services#benchmarks" rel="nofollow">http:&#x2F;&#x2F;serverbear.com&#x2F;237-small-amazon-web-services#benchmar...</a><p><a href="http://serverbear.com/10-linode-1gb-linode" rel="nofollow">http:&#x2F;&#x2F;serverbear.com&#x2F;10-linode-1gb-linode</a><p><a href="http://serverbear.com/1990-2gb-ssd--2-cpu-digitalocean" rel="nofollow">http:&#x2F;&#x2F;serverbear.com&#x2F;1990-2gb-ssd--2-cpu-digitalocean</a>
imperialWicket将近 12 年前
These are fine benchmarks for consideration, IFF you are interested in using AWS EC2 instances as a vps. This might not be a terrible endeavor, but it&#x27;s really not the intended use.<p>EBS performance is an issue - Yes it is, don&#x27;t use EBS. Use instance store and push to S3 for persistence. If you need performance I&#x2F;O for something, there is likely a separate service available that pushes your I&#x2F;O bottleneck further away (RDS, ElastiCache, SQS, etc.).<p>AWS costs more for weaker CPU - Indeed, this can be the case. But, it&#x27;s often cheaper (but not much) to put up an Elastic Load Balancer with an Auto Scaling Group and dynamically support your peak traffic than it is to pay for an enormous VPS that sits idle 60% of the time.<p>As these benchmarks suggest, I agree that if you&#x27;re only using one EC2 instance (and you can&#x27;t get away with a micro), you should probably be investigating other solutions. If you want to architect your app&#x2F;project&#x2F;service&#x2F;whatever to be more distributed and fault-tolerant, AWS can probably make that easier (not necessarily cheaper).
评论 #5980485 未加载
leetrout将近 12 年前
I am always surprised Rackspace Cloud Servers are left out. We&#x27;ve been using them for over a year with only a handful of outages (VM host reboots- 3~5min each).<p>For us it is worth the price (&gt;$40&#x2F;mo for 1 GB ram) for great stability and support. The Openstack API is really nice, too.
评论 #5980562 未加载
throwit1979将近 12 年前
...and the author is using Laravel, which delivers stunningly terrible latency and throughput performance in the techempower benchmarks:<p><a href="http://www.techempower.com/benchmarks/" rel="nofollow">http:&#x2F;&#x2F;www.techempower.com&#x2F;benchmarks&#x2F;</a><p>Would be nice to see similar benchmarks using a framework that doesn&#x27;t resemble a honeypot&#x27;s tarpit.
评论 #5978762 未加载
评论 #5978720 未加载
harper将近 12 年前
Here is the original post: <a href="http://www.cosninix.com/wp/2013/06/amazon-aws-ec2-linode-digitalocean-cloudserver-showdown/" rel="nofollow">http:&#x2F;&#x2F;www.cosninix.com&#x2F;wp&#x2F;2013&#x2F;06&#x2F;amazon-aws-ec2-linode-dig...</a>
trimbo将近 12 年前
&quot;Is Amazon overpriced&quot;<p>Well, yeah. If you use it as just commodity servers. At Radius, we migrated our index build process over to using Elastic Map Reduce on spot-priced servers and it&#x27;s been a huge cost savings.<p>Long story short, move to Amazon if you want elasticity and can design around saving from their services. Otherwise, look elsewhere.
mark_l_watson将近 12 年前
Well, AWS does have some very useful value add services: Elastic MapReduce, S3, and DynamoDB being my favorites.<p>Way back when, Amazon gave me about $1200 of free use credits over a two year period, and I played, experimented, used it for most customer projects, etc. I also used AWS for almost all of my own projects.<p>In the last year or two however, I have started going back to renting large VPS by the month (I use RimuHosting, but there are a lot of good providers) because you get so much more capabilities for the same amount of money.<p>A little off topic, but another way I have found to save money is to wean myself off of Heroku by taking the little bit of time to set up a git commit&#x2F;push hook to automatically deploy my web apps. I was using a manual deployment scheme before that took me a minute for each deployment - not so good.<p>All that said, AWS is really awesome for some jobs like periodically crunching data with MapReduce, etc. I bought a very useful little book &quot;Programming Amazon EC2&quot; a few years ago, and I recommend that as a good reference for using the AWS APIs.
评论 #5978931 未加载
stevedomin将近 12 年前
I found your conclusion a bit misleading and unfortunately a lot of people often comes to the same conclusion.<p>Indeed the CPU is limiting factor in your bench, and while you say you&#x27;re comparing apples to apples, it&#x27;s not true in my opinion.<p>Maybe I&#x27;m wrong but I never see these actors as really competing in the same market : of course if you&#x27;re looking for the best CPU perfomance&#x2F;price ratio AWS is one the worst choice. But that&#x27;s not what you&#x27;re paying for imo, as someone already mention you&#x27;re paying for the programmatic access, the ecosystem, the auto-scaling ability, monitoring, etc. Move to AWS if you need the &quot;Elastic&quot; part in EC2 (Elastic Cloud Compute) or if you know what you&#x27;re doing and what you&#x27;re paying for.<p>Same for DO, what you&#x27;re paying for is the SSD, they&#x27;re advertising about it everywhere, so it&#x27;s kind of obvious that a big part of what you&#x27;ll pay for is the SSD.
mratzloff将近 12 年前
If security is a consideration, I personally wouldn&#x27;t use Linode at this point. AWS has a dedicated security team. I don&#x27;t know about DigitalOcean.
dschiptsov将近 12 年前
Oh, I&#x27;m too late to the party.) Nevertheless, virtualization is not for production.) Even para- or hvm. I&#x2F;O is a bottleneck.<p>Linode is just plain xen, and domUs packed not so tightly, so it performs more or less quickly if other instances are 100% idle.)<p>The main assumption about virtualization is that there is no two high load domU instances running in parallel, same scenario as it was with primitive apache virtual host based hostings - we could pack them tightly, each one have 100 requests per week on average.<p>So, virtualization works fine for almost always idle development servers, but everything would fall apart in a I&#x2F;O intense production environment.<p>The mantra is &quot;I&#x2F;O request should be separated and data partitioned&quot;. In case of cheap virtualized &quot;servers&quot; storage is the first bottleneck, because you share it with other domUs. Once they&#x27;re idle, you&#x27;re OK. Should one next to you running, say, a torrent tracker - you&#x27;re screwed.<p>It doesn&#x27;t matter what you&#x27;re running under - Xen or just FreeBSD&#x27;s jails (still love them). The problem is that a HDD could perform only one operation (read or write) at a time.) So, your dom0 is deeply in IOwait, and your, say, mysql on domU locked all your tables, waiting for insert&#x2F;update completion.<p>I could write a brochure, but in short - virtualization on productions is the same unnecessary complications as a Java Virtual Machine. They are nice toys, but in production everything is better without them.)
bryanlarsen将近 12 年前
Did you try multiple instances on each provider? Performance between VPS&#x27;s from the <i>same</i> provider varies wildly. If there are 64 VPS&#x27;s on a server, you will probably get a lot more than 1&#x2F;64th of the processing power of that server because most of the other VPS&#x27;s are idle. But how much more than 1&#x2F;64th you get will vary wildly from machine to machine.
评论 #5979111 未加载
franklaemmer将近 12 年前
My two cents here: Benchmarking is all fine, but from my point of view, the performance-price-ratio is not sooooo important in hosting.<p>This discussion here reminds me of PC customers buying behavior in the 90ies. What&#x27;s better AMD or Intel? ... Nowadays other features are key: What&#x27;s the weight of this device? How thick is it even? Apple has changed the way we look at these things today.<p>Convenience also matters in hosting a lot. How much time do i have to spend to have my app up and running? Do i really want to set up and maintain everything myself? How good is the support? Do i want just bare metal computing resources or a solution provider with an eco system?<p>What matters the fastest server ever, when the queries are slow? The performance of any app&#x2F;website relies heavily on the engineering skills of the developers. See caching, see i&#x2F;o load, see frontend technolgies, see #perfmatters.<p>disclaimer: i am co-founder of a PHP PaaS.
calinet6将近 12 年前
And compare it to a dedicated box, and they <i>all</i> suck. If you&#x27;re looking for a stationary server and just need raw power, get a dedicated server plan and good backups.<p>If you need what Amazon offers, all of the AWS services or most of them, then go for it. But you don&#x27;t use AWS for raw power.
okrasz将近 12 年前
There are at least two tools which are very useful in such comparisons - <a href="http://www.cloudorado.com" rel="nofollow">http:&#x2F;&#x2F;www.cloudorado.com</a> for pricing, and <a href="http://cloudharmony.com/" rel="nofollow">http:&#x2F;&#x2F;cloudharmony.com&#x2F;</a> for benchmarks.
评论 #5980500 未加载
评论 #5978858 未加载
ausjke将近 12 年前
Linode for hosting, DO for experimenting and Amazon for Glacier, the combination worked well for me.
评论 #5980728 未加载
yellow将近 12 年前
I&#x27;m interested in seeing Google&#x27;s Compute Engine show up in some of these benchmarks. Any reason why someone would not want to use them? Some benchmark&#x27;s I&#x27;ve seen show them with awesome stats.
henryw将近 12 年前
Something to consider is that DO doesn&#x27;t offer internal network IP&#x27;s. So if you have something like a cache farm, you have to access them via external IP&#x27;s. Linode is supposed to internal IP&#x27;s, although some manual tickering is required. I&#x27;m not sure about EC2, but I&#x27;ll assume they do.<p>On a cost per hour basis, DO is the cheapest. Linode next. And Amazon most expensive.<p>Also, DO is only located in NY, while the others offer central and west cost locations.
评论 #5981405 未加载
akurilin将近 12 年前
What&#x27;s the recommendation for hosting Postgres on a VPS these days? DigitalOcean doesn&#x27;t seem to offer NAS drives, and I believe Linode is in the same boat.
评论 #5979183 未加载
hcarvalhoalves将近 12 年前
This is only true with the small instance type though, which is severely CPU handicapped. The larger instances (&gt; medium) get the full CPU throughput. Their &quot;EC2 Compute Unit&quot; rating is mostly marketing, I could never verify it by practice. Some instances with twice the ECU actually give more than twofold performance, while instances with 1 ECU should be renamed to 0,25 ECU.<p>I would expect anyone who works with AWS to know this by now.
评论 #5979390 未加载
wh-uws将近 12 年前
Has anyone used Cloud66 or Cloudfoundry to deploy to either of these (Linode, AWS, Digital Ocean)?<p>I&#x27;m about to deploy an app and I would like to get other impressions.
评论 #5979289 未加载
ww520将近 12 年前
One of downsides of AWS is the bandwidth cost. It&#x27;s 5x to 10x more expensive. Running heavy media based websites is just not feasible.
评论 #5983384 未加载
评论 #5980417 未加载
davidedicillo将近 12 年前
I think the reason many choose Amazon is also all the other services they offer, keeping everything under the same umbrella.
samcrawford将近 12 年前
It would be helpful if the author included the round-trip latency (even some simple pings) alongside the benchmark so we could judge how far the servers are from the client.<p>This will have a significant impact on the ab numbers (particularly as he&#x27;s not using the -k option and therefore establishing a new TCP session for every request).
akurilin将近 12 年前
Those of you who made the switch from EC2 to any of the other providers, what are you missing the most? I&#x27;m right now pretty integrated into the whole VPC&#x2F;Route53&#x2F;IAM&#x2F;S3&#x2F;SES web and get the feeling that it&#x27;d be pretty rough to untangle myself from that.
mzarate06将近 12 年前
The benchmark is on web requests per second, so it really needs to cite each server&#x27;s location or ping latency relative to the machine doing the benchmarks. It&#x27;s unfair if one of the servers is on the opposite coast of the other 2.
pindi将近 12 年前
I don&#x27;t know that comparing EC2 reserved instances to DigitalOcean is really &quot;apples-to-apples&quot;, since DigitalOcean has hourly billing and no upfront cost just like EC2 on-demand, and unlike EC2 reserved.
cpolis将近 12 年前
For perspective on pricing and different plan information, I built the following tool last week:<p><a href="http://cloudpricegrid.com/" rel="nofollow">http:&#x2F;&#x2F;cloudpricegrid.com&#x2F;</a>
mp99e99将近 12 年前
Hi, rudimentary but interesting results. Please consider including atlantic.net in your testing in the future.
tszming将近 12 年前
I am quite surprised no one has mentioned SoftLayer. Isn&#x27;t IBM buy SoftLayer for their cloud offerings?
praguebakerr将近 12 年前
Well depends on application you need, spot on instances are cheapest for batch processing.
HyprMusic将近 12 年前
Has anyone had any good experiences with any European VPS providers?
sunsu将近 12 年前
If linode offered an RDS replacement, I would switch now.
sathishmanohar将近 12 年前
Thanks for the comparison.<p>My first impression? graphs are better :)
daakus将近 12 年前
fwiw, i had a similar experience as well. i tried prgmr too, and linode was just much better and more consistent than all the others.
roma1n将近 12 年前
I never see Gandi offerings mentioned. Anyone has a review &#x2F; advice?
评论 #5980007 未加载
lifeguard将近 12 年前
Conclusion<p>Well, Amazon just sucks.