Amazon’s Scorpion Problem
feld.com
feld.com
#1 - "AWS is not the low price provider", "Being low priced is in Amazon’s nature so this will be intensely challenging to them". This is true, but as the market-leader, do they need to be?
#2 - "most of their features are mediocre knock offs of other products". As @ToddMathews said here (https://news.ycombinator.com/item?id=8104364), he doesn't list any examples of these superior cloud services. I can't think of any. Sure, ELB is a knock off of HAProxy, but in that sense, Gmail is a knockoff of Postfix.
#3 - "AWS is unbelievably lousy at support", "Support is just not in Amazon’s nature". Really? Compared to whom? In my experience, they have always been one of the best providers of support of any online company.
#4 - "Once you are at $200k / month of spend, it’s cheaper and much more effective to build your own infrastructure". And yet many businesses spend more than $200k PCM. Are they stupid? Or is the author wrong?
edit: Sorry, didn't see "vendor" before. True, when you are using marketplace you seem to be pretty much on your own.
ec2 and associated services' customer service and support has been pretty shitty, and these are from companies paying anywhere from $8k to $96k/mo
Does not list a single example.
I'm pretty sure Netflix uses it due to the highly variable usage situation. It never makes sense to have a server sitting idle for 23 hours a day.
EDIT: found someone who's made a feature grid here: http://www.planforcloud.com/pages/resources/cloud_services.h...
This doesn't even include more esoteric stuff like Redshift and Elastic Beanstalk.
In addition most sysadmins and developers will be more likely familiar with AWS's tools, API and dashboard. They'll be more likely to already have their development environment set up with AWS command lines tools. You ever heard of someone with HP cloud tools installed [1]? I frequently browse dotfiles on GitHub and often see non-sensitive AWS credentials and aliases. I've never seen a Rackspace or Google Cloud thing in them.
AWS wins on mindshare. It's also a proven system. Infrasture decisions are really important, and downtime is very bad and when you choose AWS you know at the very least that if your shit goes down because of provider issues everyone else will be down too and you can kind of get a pass (although outsourcing downtime is kind of shitty), but if you're not on AWS and your shit goes down because of the provider people will ask you why aren't you using AWS. Your customers and investors might actually be unhappy with you because of this.
As for support Google also has terrible support, Rackspace's managed support is very expensive and actually probably not what a competent team would want, otherwise it is easy to get a person on the phone at Rackspace, so that is commendable. Support doesn't scale, but if you are paying AWS a lot I bet you can get support.
I don't really like the mono-culture of AWS everywhere, but I feel like the points above need to be raised. A standard on APIs like OpenStack would be nice, but implementations will differ and AWS will fight this kind of change and because of their dominance getting traction will be hard.
[1] this is actually a thing: https://docs.hpcloud.com/cli
Compared to whom? Google? How many times have you seen tickets like [1] for Android, Gapps, etc? Additionally, the google cloud platform onboarding was so bad, I almost gave up a $500 credit in frustration.
[1] https://productforums.google.com/forum/#!topic/gmail/HTX9cLb...
On the other hand, when Amazon first released EC2, they gave me a $1000 credit to use anyway I wanted, and that I did use.
I really like Google products like GMail, Blogger, and G+ and working at Google the infrastructure was more than wonderful, but for external developers the Google Cloud Platform products really could be a lot better.
From the title, I thought it would be about Amazon competing with its customers, like Netflix using AWS to compete with Amazon Prime, and how its inevitable that Amazon will eventually have to sting Netflix.
The server room we have at our office have had max 5 hours of total down time since 2011 And we have moved our entire office under that period. I will try to estimate our general costs in EUR.
Inital costs for servers, switches, cooling, Racks UPS 150K
Replacement of servers every 3-4 years 60K/3 = 20K
Internet line = 10K a year
Electricity = 2K a year
Consulting cost for hardware maintenance = 10K
Old servers we put those in an a room a bit away and use them for backups/failsafe.
That is about 50K a year.
Counting in initial costs and just cause is is till almost nothing compared to 200K a month, I would say that switching to self hosting is financially beneficial when getting to about 10K a month with 3party hosting.
But to be honest there is not like one has to do much width the hardware.
When one hardware server fails all the virtual machines will still just continue on another hardware server. Vmware takes care of that stuff and it just works, at least in our testing of pulling the cords to the server (stuff has not failed yeat).
Well if money is no object, you are definitely doing the right thing.
How much would a comparable setup cost on AWS per year? Once we know that we can draw conclusions.
But that left out some cost like Shared network folders that our users use, so we would have needed to get some kind of replacement for our netapp(a form as NAS).
He doesn't give a single example of why "AWS is not the best product at anything". This is obviously true, but the same could be said about Google, Microsoft, or Rackspace. So what's the point? Who has a better offering?
The discussion of AWS being more expensive than bare metal is valid, but that's an obvious statement. Amazon uses bare metal (albeit at scale), and adds their application stack + support. You're paying for the convenience of purchasing all services integrated from a single vendor. If your business is simple enough and you can just throw hardware and bandwidth to solve your needs, then bare metal will always be cheaper. But for more complex environments (e.g., multi-regions, global distribution, auto-scaling), or when you need to quickly scale up and down (to accomodate surge in demand, for example) bare metal is not an option.
Net net, this sounded like a rant against Amazon's world domination plans, and the threat to the very existence of some of the companies on his fund. He said it best:
"We used to think of Amazon as a potential acquirer for these companies, or at least a powerful strategic partner. Now we know they are just using the bait of “we want to work more closely with you” as market and product intelligence."
If your business model depends on a strategic partnerships with or being acquired by Amazon, you're doing it wrong.
Isn't it a self obvious truth that AWS is most appropriate when you need lots of computational resources in bursts?
If you often need many extra servers for a short period of time, then AWS is really good for that. Another thing that AWS provides is some very handy infrastructure APIs like S3, DynamoDB, Eleastic Mapreduce, etc.
For flat usage scenarious, then renting beefy VPSs or physical servers is more cost effective.
To be fair, that is part of the cost. However, unless you are really big, you will just pay remote hands which is like $100-200/hr. The number of incidents are probably way too few, even at 200k a month, for 24/7 on-site sysadmins/techs.
[1]http://www.webhostingtalk.com/showthread.php?t=920078 [2]http://redd.it/16m9pq
A decent internet connection from level3 or the colo house internet for another few hundred.
So even at $2k/month we've still got $180k/month to buy servers and storage (or $2M for the year).
$200k/month is a lot of money.
So you build your colo'd private cloud on openstack, and then use their's to handle surges. As your traffic grows, you add your own gear to handle the base line.
But ~200K per month, you're operating at a very different scale from 2k or even 20k. Having a couple of idle servers 23 hours a day is not wasteful at that scale, unless your CIO is Ebeneezer Scrooge.
The failure rates vs physical hardware are incredibly high (just personal experience but I doubt you'll find many who disagree).
Once your systems are up and running you don't need to do much. Remote hands will replace a hard drive for $100, for example.
1) you aren't including operations staff.
2) what are you doing for DR/HA?
3) how about CDN etc?
4) how about load balancing between datacenters? DNS failover?
5) how many units of computer would $200k be at amazon? I'd imagine a lot. I recently (nearly a year ago) specced out 16 servers on AWS with HA/DR on opsworks for $15k over 3 years ($5k/year). Back of the envelope calculations make that 640 cores for your $200k. With 400GB transfer a month.
6) what are you doing for backups?
7) what if you don't need to utilize a whole rack? Most of the time systems don't fully utilize racks. But you're paying for them anyway.
1- you still need staff to deal with aws. I would argue that you need more.
2- HA is the same; DR is different. Use AWS for your DR ;)
3- What about them? there is literally no difference
4- see 3
5- a lot, I'm sure. Enough to justify owned physical equipment that you aren't sharing and doesn't get "retired" every other week
6- whatever I want, including putting them on glacier or sending them to my data center's cheap tape storage
7- it's a trivial part of the cost, why worry about it? or rent half a rack if you don't mind taking on a little extra cord-pulling risk
I think EC2 makes total sense for variable load (I paid like $0.25 to run some Coursera homework on a huge computer for 15 minutes and that was amazing) but if you're running something that needs to be up constantly and reliably (e.g., a database) EC2 gets expensive pretty fast.
A secular shift? As opposed to a religious one?
http://www.merriam-webster.com/dictionary/secular
…Feld is using the sense (3), and especially (3)(c), a quite common usage in economics (as a contrast with transient/cyclical changes).
You might benefit from a more catholic appreciation of vocabulary. (And no, that has nothing to do with the Catholic church.)