Amazon Introduces Bidding For EC2 Compute Time

ryanvm alerts us to Amazon's beta announcement this morning for what it is calling Spot Instances, which represent a name-your-own-price way of using the elastic compute service. Here is Amazon's documentation on the feature. "For customers with flexibility in when their applications can run, Spot Instances can significantly lower their Amazon EC2 costs. Additionally, Spot Instances can provide access to large amounts of additional capacity for applications with urgent needs." Customers can use the EC2 API to see recent spot prices.

Comments

Permutation city

[Brian Gordon]

Reminds me of:

and civilisation has accumulated vast amounts of ubiquitous computing power and memory which is distributed internationally and is traded in a public market called the QIPS Exchange (QIPS from MIPS, where the Q is Quadrillions)

Great book. http://en.wikipedia.org/wiki/Permutation_city.

Re:Permutation city

[MichaelSmith]

Yes. All we need now is the software which Egans characters were using to simulate their personalities.

Re:Permutation city

[afidel]

In short scale Quadrillion is only peta, so we are very much there today with the largest computers being able to do quadrillions of floating point operations per second let alone integer. Not sure if worldwide processing power is yet to the yotaflops scale though I suspect we probably aren't far off.

Re:Permutation city

[robbrit]

And YIPS sounds so much better than QIPS.

Re:Permutation city

[Anonymous Coward]

That is a good book. I had forgotten the name over the years, so thanks for the post!

Re:Permutation city

[Permutation Citizen]

I agree. Greg Egan is my favorite SF writer.

nifty

[Anonymous Coward]

if only i needed more cpu power for something..

Instance creation?

[afidel]

Last I had heard if you wanted to create more than a couple dozen new VM instances at a time you needed to get custom quotes from Amazon, with this metering in place I assume they have worked out those provisioning problems?

Re:Instance creation?

[emcron]

It's not a matter of getting custom quotes -- they just want to know what you're intended use is going to be. The price tiers (I believe) remain the same. It's a basic check to try to limit malicious users who would spawn thousands of instances for spam or other nefarious purposes.

Re:Instance creation?

[42forty-two42]

There's no custom quote for increasing the instance limit - it's just a safety to prevent someone with a stolen credit card from spawning enough instances to use up all available capacity. You just put in a request and have it lifted; pricing stays the same.

Re:Instance creation?

[Anonymous Coward]

Initially when you get your EC2 account, it is limited to 20 concurrent VMs. You then have to request to have this limit increased, but the same pricing applies (at least up to 200 concurrent VMs)

Having accounts limited to 20 VMs means that a single user can't sign up and start to flood the system. Also they make no guarantees that a request to start up a VM will succeed unless you have a reserved instance, which has a yearly cost attached.

What this spot pricing does is provide a financial incentive for customers to move load to off peak times rather than adding to the load during peak demand. This in turn should flatten out the load on EC2 (depending on how much load can and will be moved to the off peak times).

In the end this is simple supply and demand economics, and a lot of people in academia have been carrying on about this and a "cloud marketplaces" for quite some time...

Re:Instance creation?

[Slashdot Parent]

No, the limit for On Demand instance creation is still 20 instances per account (but it is allowed by amazon TOS to create multiple accounts, hint hint).

The limit for Spot instances (the kind described in the summary) defaults to 100 per account.

As you note above, you can get that limit raised by Amazon, upon request.

Indictment of cloud computing?

[teknopurge]

I wonder if amazon overestimated the uptake of EC2 and needs to have a fire-sale to get _some_ income on the investment. They need to have capacity to service new orders but by default this means that they have un-used resources that are costing them more then the unit bring in.

Re:Indictment of cloud computing?

[Anonymous Coward]

I'm sure you have no conflicts of interest, so I can take what you said without a grain of salt.

Re:Indictment of cloud computing?

[afidel]

Considering that the 3 year cost of a Double Extra Large Instance 34.2 GB of memory, 13 EC2 Compute Units (4 virtual cores with 3.25 EC2 Compute Units each), 850 GB of local instance storage, 64-bit platform is $36k and that the 3 year TCO for such a machine in my environment today would be about $8k I don't think they are losing money on EC2 unless their utilization is *really* low, but this does allow them to maximize utilization (and profit) as it allows near perfect price discrimination. It also allows them to scale *their* resources for things like cyber monday by bumping these low priority jobs off the cluster and using it to run their own dynamic site.

Re:Indictment of cloud computing?

[42forty-two42]

I don't know if they _need_ the income, but I'm sure they have plenty of excess capacity from reserved instances that aren't running for starters, not to mention the pool for on-demand instances. It's just a smart business move to try to exploit this.

Re:Indictment of cloud computing?

[fuzzyfuzzyfungus]

I suspect that we'd have to watch the prices pretty closely to find out.

Variable pricing(down to the point where electricity costs make it cheaper to just turn them off and suck up the losses, of course) is completely logical for dealing with any unused capacity, and it can't have cost Amazon all that much to hack this pricing scheme into their existing system.

If there is usually some capacity available, with a small discount in exchange for having no availability guarantee(or even implication), with occasional chunks of cheap time, then it is probably just an expected side effect of the fact that "cloud" demand(like a fair bit of classic datacenter demand) is somewhat bursty. If there is constantly capacity available at 20 cents on the dollar, well, then Amazon would seem to have a bit of a capacity problem...

Re:Indictment of cloud computing?

[enoz]

In order for EC2 capacity to be highly available (I haven't yet heard of people waiting in a queue for hours for an instance), it seems obvious that Amazon must have a large amount of computing power in standby.

This process of auctioning off the extra processing power based on fluctuating capacity seems like a win-win situation for Amazon AND users. Users who want increased processing, but are not time-bound, can get "off-peak" rates. Meanwhile Amazon can make money off the "idle" processors which are still available to be reserved as an EC2 instance.

Re:Indictment of cloud computing?

[TooMuchToDo]

In order for EC2 capacity to be highly available (I haven't yet heard of people waiting in a queue for hours for an instance), it seems obvious that Amazon must have a large amount of computing power in standby.

They do, since they have to be able to handle peak loads around Christmas and so forth. It just makes good financial sense to make some money off of all that infrastructure when it's not in use.

Re:Indictment of cloud computing?

[NeilO]

No, I don't think so. More likely they already own outrageous capacity to handle the December retail crunch but don't use anything close to that capacity in any other month of the year. So I doubt there's any additional investment in capacity for EC2. EC2 utilizes what Amazon already owns.

If you read Werner's blog entry on this new feature you'll see they reserve the right to interrupt a Spot Instance and essentially restart it later on. You need to make sure whatever you're doing with that instance you can checkpoint and resume. I think that means Amazon is not trying to "fire sale" underutilized resource. More like they're filling in the cracks between larger "full price" instances in order to maximize utilization.

Re:Indictment of cloud computing?

[Anonymous Coward]

How'd you work out your numbers?
3 year double extra is $4900 plus $0.42 per hour. Say 100 hours per week, $42 per week * 150 weeks, equals about $12k.
In your TCO above, a 32G machine may cost you $8k, but you haven't allocated rackspace, or redundant power, or bandwidth, or other failures.

Re:Indictment of cloud computing?

[afidel]

I just did hourly rate for non-dedicated * 24 * 365 * 3 and compared it to the cost of machine + power + AC fraction + warranty uplift which is my standard TCO calculation minus admin time as the admin time will be roughly the same either way.

Re:Indictment of cloud computing?

[dkf]

[Amazon] do, since they have to be able to handle peak loads around Christmas and so forth. It just makes good financial sense to make some money off of all that infrastructure when it's not in use.

There's this great myth that Amazon need all this capacity for handling Christmas. They don't. It may have started that way, but now their infrastructure for handling services is much larger than that. (IIRC, they passed that level sometime last year, or maybe even back in 2007.) They are now a cloud provider that happens to sell stuff (books, etc) on the side.

Re:Indictment of cloud computing?

[pydev]

minus admin time as the admin time will be roughly the same either way.

Admin time is not at all "the same"; do you have any idea how much time it costs and how much risk is involved in ordering, configuring, and high-end machines? Then there's the room, the power supplies, the racks, all the supporting infrastructure.

Also, when you buy them, you are stuck with those machines--you have sunk costs. The less you use them, the more you pay for the hours that you do use them.

Re:Indictment of cloud computing?

[Anonymous Coward]

>  They are now a cloud provider that happens to sell stuff (books, etc) on the side.

If I'm reading their SEC filings correctly, 54% of their sales revenue comes from books and movies, 43% from electronics and random crap, and 3% comes from cloud computing.

Looks to me like they're very much still a retailer, albeit one who has started selling off compute capacity.

Re:Indictment of cloud computing?

[afidel]

High-end? Ha, a 32GB machine is as small as I have ordered this year. The biggest was 256GB, and the average was 72GB. Of course I'm already moving towards 90% virtualization internally so the only additional cost is the physical rack and stack which fairly minimal when spread over a 3 year system life (care and feeding is MANY more hours by comparison).

Re:Indictment of cloud computing?

[pydev]

You're really confused when you compare the size of machines to host virtual instances with the size of the actual instances themselves. For instances, 32G is big, even today.

And it's nice that "you" are moving to "90% virtualization internally", but what that tells me is that you get paid to do that sort of thing, so I hope you put your salary and the salary of everybody else needed to order, deploy, and maintain your virtualization into the computation (the cost of setting up instances is the same).

Basically, the savings from cloud computing come from firing people like you and reusing your machine rooms and other space for offices. The fact that Amazon provides these CPUs without having to interact with primadonnas is an added bonus.

Re:Indictment of cloud computing?

[afidel]

OK, they fire me, who's going to setup, configure, and maintain the cloud VM's? Oh, that's right they still need someone to do that.... There is no magic pixie dust despite what IBM and PHB's would like to think, companies need digital janitor's to feed and care for the complex machines that run the company and that's exactly what I do. If you are smaller than Amazon or Google and you have someone who spends most of their time worried about the physical infrastructure then they are doing something terribly wrong.

Re:Indictment of cloud computing?

[Sulphur]

Are you predicting rain? If so how much compute power did it take?

Re:Indictment of cloud computing?

[pydev]

OK, they fire me, who's going to setup, configure, and maintain the cloud VM's?

Which part of "the cost of setting up instances is the same" did you not understand?

There is no magic pixie dust despite what IBM and PHB's would like to think, companies need digital janitor's to feed and care for the complex machines that run the company and that's exactly what I do.

So? Even if your company is so small that you're the single IT person and you spend x% of your time setting up virtualization infrastructure (not instances), x% of your salary and overhead needs to count towards that. That's likely at least $50k/year right there (I'm assuming you aren't paid that well). Add to that the time and money spent on (your) training, purchasing, janitorial staff, rent, planning, insurance, rent, and that easily doubles.

And you can only compare things if you have 100% utilization; if you have 50% utilization, what you're paying per compute hour in-house just doubled, since all those costs are mostly fixed.

Sorry, you may know how to install a VM, but your economics leave a lot to be desired.

Re:Indictment of cloud computing?

[afidel]

What you don't seem to understand is the value I bring is not primarily in the day to day operations, they can get someone far cheaper to check logs, it's in my ability to respond is a crisis and my ability to design systems to avoid crises. Even if my utilization were 50% (it's not) I would still be at less than half the 3 year cost of the Amazon solution. I guess if I needed a massively scalable web presence it would be an option, but our primary systems are remote access, content managment, OLTP, reporting and business intelligence with only remote access really fitting into a bursty demand model and even that needs to be internal to access all the data systems that the company is not likely to ever outsource. We do use remote hosted solutions when it makes sense, but like most businesses it's not likely to ever be the majority of our IT systems.

Re:Indictment of cloud computing?

[Danny Rathjens]

It also allows them to scale *their* resources for things like cyber monday by bumping these low priority jobs off the cluster and using it to run their own dynamic site.

The EC2 systems are completely separate from the systems that run amazon.com. (According to a seminar on EC2 by Amazon folks I attended at an NYLUG meeting in October.)

Re:Indictment of cloud computing?

[xenocide2]

And you don't think their clients suffer from Christmas peak loads? Even if you don't sell things, it's a time of year with a lot of vacation time.

Re:Indictment of cloud computing?

[Anonymous Coward]

What you don't seem to understand is the value I bring is not primarily in the day to day operations,

I have no idea what "value you bring" in general, but I do know that it isn't in setting up and maintaining a VM infrastructure, at least not based on th analysis you provided.

that the company is not likely to ever outsource.

As I was saying: there are many reasons not to outsource. However, the cost analysis you presented is not one of them.

Re:Indictment of cloud computing?

[lonecrow]

I guess you were using the "on-demand" pricing. If you use a reserved instance on a 3-year term you start at $4900 + $0.42/hour. I think this works out to only $15,937 over the three years which is less then half the $36k you quoted.

I am looking at a single smaller machine. My current dedicated server is ~$200/month. Thats about $7.5k over 3 years. The equivalent EC2 reserved instance is only $350+$0.03/hour. That's $788 total for 3 years or a tenth of the price! Sounds like a hot deal to me :)

Of course at this time you can only reserve Linux instances and I happen to need a windows one so I have to go with the on-demand prices :(

Re:It's too proprietary

[dikdik]

If you're using Amazon for hosting, you can't switch hosting services; their system is too nonstandard. Do you want to be in a position where they can raise prices or cut off your air supply?

EC2 allows you to setup your own servers on their infrastructure. Ultimately, this is as standard as getting a virtual or dedicated server at any one of thousands of other hosting providers. Switching is as easy as replicating the environment you've created for yourself (which is likely a standard LAMP stack anyways) and then doing a DNS change.

Re:It's too proprietary

[fuzzyfuzzyfungus]

The actual ec2 instances are pretty much bog-standard linux VMs. For everything else, there's eucalyptus.

This doesn't mean that EC2 is necessarily a good idea, or that Amazon couldn't slip you a nasty dose of strategic downtime if they, for some reason, felt like it; but their ability to exert lock-in in the medium term is pretty weak.

Brilliant market for surplus

[seifried]

Since an EC2 server not running an EC2 instance is literally burning cash for Amazon (depreciation value, electricity, maintenance, building cost, etc.) being able to sell surplus time like this through a secondary market that allows Amazon to recoup value for the EC2 instances at a minimum price is brilliant, of course the infrastructure needed to support EC2 let alone is pretty staggering, no wonder EC2 is stomping everyone with their insane flexibility.

Re:Brilliant market for surplus

[Chuck Chunder]

Since an EC2 server not running an EC2 instance is literally burning cash for Amazon

I know the USD has been slipping but surely grid electricity is still a cheaper source of energy then using your own greenback powered generator?

Re:It's too proprietary

[Alpha830RulZ]

but their ability to exert lock-in in the medium term is pretty weak

Which is related to their strength, for me. I use them precisely because of the short term availability. We ran a test job just today on their service to test the scalability of a routine. Bought 20 2 core machines for 4 hours for I think $64 or so. I'd have to have at least 4 big multi-core servers lying around idle to have that available for a test bed, best case about $12,000 for the bunch (I've priced them in the last month, this is the best choice I found. That's 4 separate servers in a single 2U chassis. I have several boxes from these guys, they are very solid, are 2/3 the price of Dell, and don't need a blade cabinet.)

I wouldn't use EC2 for long term use, but for short term, or speculative use, it's a pretty good deal. I've found them to be pretty stable and easy to use.

Spare Capacity

[Anonymous Coward]

I think this is a great way for them to sell spare capacity. I would imagine that they have alot of it during off-peak times.

Overbuilt & Aging Rapidly

[ratm999]

I thought the main benefit of "elasticity" was remove "peak usage" as a primary buying factor. If you have a routine that can run, well whenever, then is EC2 any cheaper than using whatever is sitting around unused in your shop already? The cloud may be a useful resource, but the size & speed of the cloudrush has been overestimated by the entire freakin' industry. By the time it gets sorted and a mature market develops, Microsoft, Amazon, and whomever else will be taking write-downs on all the obsolete kit they built-up in 2008-2009.

Re:Overbuilt & Aging Rapidly

[Anonymous Coward]

Except Amazon didn't go out and buy servers for EC2. They essentially use the servers that are only needed for the big holiday rushes.

Re:Overbuilt & Aging Rapidly

[DragonWriter]

I thought the main benefit of "elasticity" was remove "peak usage" as a primary buying factor.

That's certainly a benefit.

If you have a routine that can run, well whenever, then is EC2 any cheaper than using whatever is sitting around unused in your shop already?

Possibly not, but if you've already built an infrastructure around EC2, you may not have stuff sitting around your shop intended for that use, and it may make more sense to leverage EC2 for the occasional task of this type, if Amazon provides pricing for it better than the "I need this to run right now" price.

The Next thing that will happen...

[ub3r n3u7r4l1st]

hackers will going to drive up the price of using the EC2 cloud.

Should VMware vCloud Go This Route? Feedback?

[Anonymous Coward]

If anyone has any feedback about whether or not VMware vCloud partners should take this route, post some questions or thoughts on the Twitter feed using @reply http://twitter.com/vcloud or the vCloud Facebook Page http://www.facebook.com/#/pages/VMware-vCloud/ We'll try to react to feedback and respond to questions as much as we can.

good. Now can I buy mobile bandwidth that way?

[MancunianMaskMan]

seems it would be be a brilliant way to marked mobile data bandwith as well: when the cell is not too busy and neither it the upstream link, they can sell it off cheap and people can facebook all they like (while i read /.), and when the traffic heats up i still have my work email going but the web2.0rhea flow is temporarily halted.

This is good

[fulldecent]

This is the same way that electricity is sold between providers -- on 6-minute interval bidding arrangements.

Of course, today only Amazon is providing the service.

ELASTIC Compute Cloud

[Slashdot Parent]

3 year TCO for such a machine in my environment today would be about $8k

Ahh, but what would the 1 hour TCO for that machine? And how quickly could you get me 100 of them? And what if after a week, I don't need them anymore. Do I have to continue paying for them?

What if I need 1000? But only for a few days? Can you even fit 1000 in your data center?

EC2 is for elastic computing needs. The price will never compare favorably to static computing on a 3-year basis.

By the way, your 8k is very low compared to what you'd have to pay to get the same featureset of Amazon Web Services. What happens if your data center catches fire? How quickly could you get that machine up and running in a new data center, and at what cost? How quickly could you upgrade the storage? Backup online to fault-tolerant storage? Clone that machine?

What if I want to to load testing of my application? Can you get me a full copy of my production environment and let me quit paying for it once my load testing is done? How much would that cost?

What about staging my application before production deployment? Do I have to pay a full year for a server I plan on using for like 100 hours of that year, tops?

Bottom line: There are a many use cases that call for elastic resources. Comparing EC2 with an ordinary server makes no sense, because they are different tools for different jobs.

Re:ELASTIC Compute Cloud

[maxume]

I think his point was more that they are setting prices in a way where they aren't losing money (they do eventually have to have actual hardware somewhere, which is where the cost comparison comes in).