Monthly Archives: March 2011
Amazon’s dedicated instances
Back in December, I blogged about the notion of Just Enough Privacy — the idea that cloud IaaS customers could share a common pool of physical servers, yet have the security concerns of shared infrastructure addressed through provisioning rules that would ensure that once a “private” customer got a virtual machine provisioned on a physical server, no other customers would then be provisioned onto that server for the duration of that VM’s life. Customers are far more willing to share network and storage than they are compute, because they’re worried about hypervisor security, so this approach addresses a significant amount of customer paranoia with no real negative impact to the provider.
Amazon has just added EC2 Dedicated Instances, which are pretty much exactly what I wrote about previously. For $10 an hour per region with single-tenancy, plus a roughly 20% uplift to the normal Amazon instance costs, you can have single-tenant servers. There are some minor configuration complications, and dedicated reserved instances have their own pricing (and are therefore separate from regular reserved instances), but all in all, these combine with the recently-released VPC features for a reasonably elegant set of functionality.
The per-region charge carries a significant premium over any wasted capacity. An extra-large instance is a full physical server; it’s 8x larger than a small instance, and its normal pricing is exactly 8x, $0.68/hour vs. a small’s $0.085/hour (Linux pricing). Nothing costs more than a quadruple extra large high-memory instance ($2.48/hour), also a full physical server. Dedicated tenancy should never waste more than a full physical server’s worth of capacity, so the “wasted” capacity carries around a 15x premium on normal instances and a 4x premium on the expensive high-memory instances, compared to if that capacity had simply been sold as a multi-tenant server. It’s basically a nuisance charge for really small customers, and not even worth thinking about by larger customers (it’s a lot less than the cost of a cocktail at a nice bar in San Francisco). All in all, it’s pretty attractive financially for Amazon, since they’re getting a 20%-ish premium on the instance charges themselves, too. (And if retail is the business of pennies, those pennies still add up when you have enough customers.)
Amazon has been on a real roll since the start of the year — the extensive VPC enhancements, the expansion of the Identity and Access Management features, and the CloudFormation templates are among the key enhancements. And the significance of the Citrix/Amazon partnership announcement shouldn’t be overlooked, either.
Cloud IaaS special report
I’ve just finished curating a collection of Gartner research on cloud infrastructure as a service. The Cloud IaaS Special Report covers private and public cloud IaaS, including both compute and storage, from multiple perspectives — procurement (including contracting), governance (including chargeback, capacity, and a look at the DevOps movement), and adoption (lots of statistics and survey data of interest to vendors). Most of this research is client-only, although some of it may be available to prospects as well.
There’s a bit of free video there from my colleague David Smith. There are also links to free webinars, including one that I’m giving next week on Tuesday, March 29th: Evolve IT Strategies to Take Advantage of Cloud Infrastructure. I’ll be giving an overview of cloud IaaS going forward and how organizations should be changing their approach to IT. (If you attended my data center conference presentation, you might see that the description looks somewhat familiar, but it’s actually a totally different presentation.)
As part of the special report, you’ll also find my seven-part note, called Evaluating Cloud Infrastructure as a Service. It’s an in-depth look at the current state of cloud IaaS as you can obtain it from service providers (whether private or public) — compute, storage, networking, security, service and support, and SLAs.
Cloud IaaS market segmentation
Over the past couple of months, I’ve been mulling over a way to structure and segment the cloud infrastructure as a service market. Some of those ideas have appeared on my blog, and have since been refined, heavily peer reviewed, and then trial-ballooned at clients. The result is a new research note, called The Structure of the Cloud Compute IaaS Market. (Sorry, Gartner clients only.)
In brief, I’ve used a two-axis strategy to break the market into eight segments.
The first axis is your general use case. Are you sourcing infrastructure that is focused on a single application (or a group of tightly-related applications, like your e-commerce application)? Or are you sourcing infrastructure for a range of diverse applications, essentially replacing a part or all of your data center? For the former, you are essentially doing a form of hosting. For the latter, you have a whole host of significantly more complex requirements.
The second axis is the level of management services. The first possibility is unmanaged — you’re doing pretty minimal operations, probably because this is a test/dev environment. The second possibility is self-managed — the provider offers the IaaS platform (data center, hardware, and virtualization), but you do the OS layer on up yourself. The third possibility is that the core foundation is service-provider managed — they also handle the OS management, usually with a security emphasis (patch management et.al.). The fourth possibility is that some or all of the rest of the application stack, minus the app itself, is service-provider managed (which usually means DBA support, maintenance of a Java EE or .Net stack of middleware, etc.).
That gets you eight market segments, as follows:
| SCENARIO | Single Application | Multiple Applications |
|---|---|---|
| Unmanaged | Developer-centric cloud hosting | Virtual lab enviroment |
| Self-Managed | Scale-out cloud hosting | Self-managed virtual data center |
| Core Foundation Managed | Simple managed cloud hosting | Turnkey virtual data center |
| Application Stack Managed | Complex managed cloud hosting | Cloud-enabled data center outsourcing |
Each of these segments has very different buyer profiles and requirements. No single service provider serves all of these segments. At best, a service provider might serve a few of these segments well, at the current state of the market. These are all cloud IaaS, but each segment serves a different kind of customer need.
Want more details? Read the research note.
Contracting in the cloud
There are plenty of cloud (or cloud-ish) companies that will sell you services on a credit card and a click-through agreement. But even when you can buy that way, it is unlikely to be maximally to your advantage to do so, if you have any volume to speak of. And if you do decide to take a contract (which might sometimes be for a zero-dollar-commit), it’s rarely to your advantage to simply agree to the vendor’s standard terms and conditions. This is just as true with the cloud as it is with any other procurement. Vendor T&Cs, whether click-through or contractual, are generally not optimal for the customer; they protect the vendor’s interests, not yours.
Do I believe that deviations from the norm hamper a cloud provider’s profitability, ability to scale, ability to innovate, and so forth? It’s potentially possible, if whatever contractual changes you’re asking for require custom engineering. But many contractual changes are simply things that protect a customer’s rights and shift risk back towards the vendor and away from the customer. And even in cases where custom engineering is necessary, there will be cloud providers who thrive on it, i.e., who find a way to allow customers to get what they need without destroying their own efficiencies. (Arguably, for instance, Salesforce.com has managed to do this with Force.com.)
But the brutal truth is also that as a customer, you don’t care about the vendor’s ability to squeeze out a bit more profit. You don’t want to negotiate a contract that’s so predatory that your success seriously hurts your vendor financially (as I’ve sometimes seen people do when negotiating with startups that badly need revenue or a big brand name to serve as a reference). But you’re not carrying out your fiduciary responsibilities unless you do try to ensure that you get the best deal that you can — which often means negotiating, and negotiating a lot.
Typical issues that customers negotiate include term of delivery of service (i.e., can this provide give you 30 days notice they’ve decided to stop offering the service and poof you’re done?), what happens in a change of control, what happens at the end of the contract (data retrieval and so on), data integrity and confidentiality, data retention, SLAs, pricing, and the conditions under which the T&Cs can change. This is by no means a comprehensive list — that’s just a start.
Yes, you can negotiate with Amazon, Google, Microsoft, etc. And even when vendors publish public pricing with specific volume discounts, customers can negotiate steeper discounts when they sign contracts.
My colleagues Alexa Bona and Frank Ridder, who are Gartner analysts who cover sourcing, have recently written a series of notes on contracting for cloud services, that I’d encourage you to check out:
- Four Risky Issues When Contracting for Cloud Services
- How to Avoid the Pitfalls of Cloud Pricing Variations
- Seven Ways to Reduce Hidden Upfront Costs of Cloud Contracts
- Six Ways to Avoid Escalating Costs During the Life of a Cloud Contract
(Sorry, above notes are clients only.)
CloudPundit: Massive-Scale Computing 