Over the past year, the number of objects stored on S3 has grown from 54 billion to 100 billion, according to Amazon CTO Werner Vogels, who mentioned this startling growth curve in his recent presentation at the Cebit computer trade show in Germany.

It’s a fuzzy milestone, to be sure, as we don’t know how much infrastructure is required to store those 100 billion objects, or how much revenue Amazon is generating from them. But in an industry where we’re used to big numbers, 100 billion is an eye-popping total. By any measure, that’s a huge storage cloud, and likely a sign of things to come.

Gartner defines three opportunities for cloud brokers:

* Cloud Service Intermediation: Building services atop an existing cloud platform, such as additional security or management capabilities.

* Aggregation: Deploying customer services over multiple cloud platforms.

* Cloud Service Arbitrage: Brokers supply flexibility and “opportunistic choices” – and foster competition between clouds.

To help you determine which cloud broker fits your needs, we’ve pulled together a list of cloud brokers and open source cloud management projects (in alphabetical order) with a brief description of their offerings. We will continue to update this list as new suppliers come on board. Have we missed someone? Send us your feedback.

CloudKick
Launched in March 2009 as part of the winter class at start-up incubator Y Combinator, Cloudkick provides management tools for Amazon and Rackspace. Users monitor their clouds through a dashboard, which also allows for tagging and color coding of nodes for easier identification. Users can see how their clouds are doing through visualized graphs. The service is hosted on the SliceHost service owned by Rackspace. Cloudkick also developed the open source project libcloud, a pure python client library for interacting with Amazon EC2, Slicehost and Rackspace Cloud Servers. In September Cloudkick 2009 raised $750,000 in funding. Its services are currently free and it is working on a suite of commercial offerings.

CloudSwitch
CloudSwitch claims to move data center applications to clouds without modification, allowing customer to manage their apps from within the datacenter using existing tools and processes. Established in 2008, its beta customers have been using Amazon EC2, but CloudSwitch plans to support other platforms from Rackspace, Microsoft, VMware and Terremark. CloudSwitch is delivered as a software appliance and contains the management components for discovering applications, orchestrating cloud deployments, and managing cloud usage, the company explains. Its CloudFit function automatically selects the appropriate combination of processor, memory, and storage. Secure communication and storage services are automatically provisioned via a local control point within each cloud deployment. These are virtual instances that run on behalf of the enterprise user and manage the cloud infrastructure, including data management, synchronization, and long-duration data transfers.

CohesiveFT
CohesiveFT says it has been providing enterprise-grade virtualization and cloud products since 2006. It’s flagship products are Elastic Server, Context-Cubed, and VPN-Cube. Elastic Server is a multi-cloud virtual server image design and management system, Context Cubed is a cloud topology management software, and Elastic Server is a multicloud server image design and management system.

DeltaCloud
An open source project aimed to develop an ecosystem of tools, scripts and applications for the cloud. The project also aims to write a common, REST-based API to enable developers to write once and manage across multiple clouds. Cloud supported include Amazon EC2, RHEV-M RackSpace and RimHosting, and private clouds based on VMware and Red Hat Enterprise Virtualization Manager.

Elastra
With Amazon.com among its investors, Elastra says it enables enterprises to use their legacy applications in public and private clouds. In November 2009, it announced Elastra Cloud Server (ECS) 2.0, Enterprise Edition, which enables applications based on Oracle databases and WebLogic Application Servers and integrated with management tools such as Tivoli and OpenView to be used on private VMware or Citrix clouds, or Amazon’s public cloud. There is a free version of ECS running on Amazon Web Services, while the commercial Enterprise Edition is aimed at private data centers. For more, seeElastra Cloud Server Extends SaaS.

enStratus
EnStratus offers a cloud management platform for enterprise applications running on Amazon and Rackspace clouds. In addition to offering the usual configuration and monitoring of cloud resources, enStratus claims its security architecture offers up to six nines of availability. All customer data is stored on enStratus’ servers running on redundant virtualized environments on dedicated physical servers. The data centers are located in downtown Minneapolis. EnStratus operates three VLANs; two house the Web console, Web services, and provisioning systems, and are accessible from the public Internet. A third houses customer credentials and is not addressable from the public Internet. The file systems are encrypted using SHA256 encryption.

Eucalyptus Systems
Eucalyptus is an open source private cloud platform that enables customers to build private clouds that are compatible with public clouds such as Amazon EC2. Customers will test out their applications on their private clouds with a view to moving them to public clouds. Eucalyptus software has been adopted by Canonical as the engine behind the Ubuntu Enterprise Cloud (UEC), which is bundled with Ubuntu 9.10 Server Edition. Customers include Eli Lilly, NASA and Argonne Leadership Computing Facility.

Kaavo
Kaavo offers enterprise management of cloud services from Amazon, Flexiscale and GoGrid. Kaavo’s IMOD provides security atop of those public clouds by configuring firewalls and VPN connectivity with both the cloud server and an internal data center or network. Kaava uses AES-256 bit encryption. IMOD monitors the application service levels and takes appropriate actions based on predefined workflows in the System Definition for the application. The System Definition also enables customers to configure multiserver configurations with different operating systems and middleware.

Layer7 Techologies
Layer 7 Technologies is approaching cloud computing with a background in service-oriented architecture and Web services security. Its SecureSpan XML Virtual Appliance supports VMware/ESX and supports both private and public clouds. It enables customers to implement policies that manage requests to cloud apps, and controls, monitors and adapts public, private and hybrid clouds. A cloud vendor SLA enforcement feature measures and tracks cloud service provider performance; and supports strict failover between public and private cloud providers, round robin, best effort and latency-based routing.

LTech
LTech was originally set up as a Google Enteprise Partner, working with customers to deploy Google Apps and the Google Search Appliance. It later added cloud services to its portfolio and is a partner of Amazon Web Services and RightScale. Its services for enterprise cloud customers include cloud assessment (recommending either a public, private or hybrid cloud), proof of concept, production migration, and management/training. In January 2009, LTech merged with IT consultancy The Matlen Silver Group, which provides project management and IT staffing to the financial and pharmaceuticals industries.

RightScale
RightScale offers a cloud management platform that enables organizations to deploy and manage applications across multiple clouds. At the core is RightScale’s Server Templates, which the company says differs from emerging common cloud APIs because the templates enable users to “take advantage of the unique capabilities of different clouds.” ISVs build one server template that automatically installs their software on other supported cloud infrastructures. RightScale is based on Opscode Chef, an open source server configuration management framework. Customers can select, migrate and monitor clouds of their choosing from a single management environment. RightScale supports clouds from Amazon Web Services, Eucalyptus Systems, Flexiscale, GoGrid, and VMware. For more see RightScale and Managed Clouds.

Vordel
Headquartered in Dublin, Ireland with offices in London, Paris and Washington D.C., Vordel is an XML and service-oriented architecture appliance specialist. The beta of its Vordel Cloud Service Broker was launched in November 2009, and the service is scheduled for release in the first quarter of 2010. The service manages multidomain cloud services by registering them in a single repository to enable monitoring and policy enforcement.

Zimory
Zimory is a spinoff of Deutsche Telekom, and claims to be the first online marketplace for cloud computing, bringing buyers and sellers of computing resources. Zimory Public Cloud for sellers aggregates available server capacity from around the world and makes it available through an Internet trading platform, according to the company. Buyers purchase capacity using Zimory Public Cloud, with Zimory handling pricing, contracts, security, virtual machine migration, and billing. Customers can choose the geographical location of the physical data center, and Zimory provides a choice of three levels of quality of service. Customers choose from a range of appliances, from a simple Linux or Windows image to fully-configured application stack.

In December, Amazon Web Services introduced Spot Instances, a pricing approach which allows customers to bid on unused Amazon EC2 capacity and run those instances for as long as their maximum bid exceeds the current spot price, which changes periodically based on supply and demand.

A day before Amazon’s announcement, Joe Weinman, strategy and business development VP for AT&T Business Solutions wrote an opinion piece for GigaOm in which he describes similar pricing possibilities for the cloud computing market. Weinman notes that cloud computing shares similarities with other on-demand, pay-per-use offerings such as airlines or car rental.

“Microseconds Could Mean Millions”
Among Weinman’s suggestions was the notion of “time-based pricing” in which computing cycles at 2 a.m. should be priced lower than prices at 2 p.m.; and even pricing based on location: just as hotels would put a premium on ocean-view rooms, cloud suppliers offering capacity located near stock exchanges could charge more because “microseconds might mean millions.”

Other suggestions include dynamic pricing, similar to the way airlines and hotels use “sophisticated yield management algorithms to maximize revenue, reducing prices to increase demand when utilization is low, and raising prices when utilization is high,” Weinman writes.

Reuven Cohen, founder and chief technologist for Toronto-based enterprise cloud platform provider Enomaly, writes in his Elastic Vapor blog that cloud computing could be traded as a commodity. He writes: “[T]he concept of distributed batch processing and compute elasticity have become critical parts of modern business IT strategies. These kind of flexible and elastic compute usage models are ideally suited to that of a spot market for commodity compute capacity (provided via a method that is quoted for immediate (spot) settlement for both payment and delivery.” He adds that Amazon’s Spot Instances has helped to legitimize this concept.

Useful for Demand Spikes
Cohen compares the opportunity for trading of excess cloud resources to the energy market because they are commodities. “The concept of selling excess capacity in cloud centric data centers may also make sense in that cloud providers must have significant additional capacity on hand just in case of demand spikes,” he writes.

Such an exchange would have to be managed by a trusted independent agency, believes Tom Lounibos, CEO of Soasta, which provides cloud-based load and performance testing of Web applications, and who has been in informal discussions with Weinman and Cohen about this subject.

Sabre As Potential Model?
“The best example that I see of a viable model would be the Sabre System that emerged in 1960s for the airline industry which allowed passengers to search for seat availability and pricing from several airlines, Lounibos writes in an e-mail interview. “This system was originally developed by American Airlines and IBM and was developed as a separate division until it was sold by AA in 2000. So the possibility of a thought leader in cloud computing (like Amazon) could duplicate American Airlines effort of the 1960s.” However, the Sabre story is fraught with controversy over AA’s ownership and control, he adds.

Lounibos says customers that consume a massive amount of compute power would benefit from such an exchange because they would have “access to an amazing amount of capacity at the lowest available cost (think Best Buy). Suppliers would benefit as well by the lowering their cost of operations (specifically in sales and marketing) if the field was leveled.”

For such an exchange to work, partners would need to be open with each other regarding their systems availability and pricing, Lounibos writes. The quality of service may vary between vendors.

Commoditization in a Differentiated Environment?
The latter issue was also raised by Andi Mann, VP of industry research firm Enterprise Management Associates. “If we think of this as codesharing for IT, where airlines may hand off a portion of your flight to another carrier, who knows whose cloud plan you will end up on,” Mann says. “If you have an agreement with a supplier who is SAS 70-compliant, would the other carrier also be SAS 70-compliant?”

Mann also questioned the commoditization of a nascent market in which vendors are keen to build differentiation, such as offering desktop as a service. “Customers may end up not buying the same model as another vendor,” Mann says. He also questioned whether customers would be willing or have the resources to constantly monitor the exchange to get a good deal.

A chart from CloudKick looking at latency for resources running on Amazon EC2.

One of the key selling points for cloud computing is scalability: the ability to handle traffic spikes smoothly without the expense and hassle of adding more dedicated servers. But this week some users of Amazon EC2 are reporting that their apps on the cloud computing service are having problems scaling efficiently, and suggesting that this uneven performance could be due to capacity problems in Amazon’s data centers. 

The reports emerge as rival services focus on Amazon’s performance in the battle for cloud computing mindshare and customers.

Amazon says that if customers are experiencing performance problems, it isn’t because EC2 is overloaded. “We do not have over-capacity issues,” said Amazon spokesperson Kay Kinton. “When customers report a problem they are having, we take it very seriously. Sometimes this means working with customers to tweak their configurations or it could mean making modifications in our services to assure maximum performance.”

Performance Issues Prompt Instance Upgrades
Kinton said Amazon has reached out to Alan Williamson of the cloud consultancy AW2.0, who is also editor of the Cloud Computing Journal and has been using EC2 for three years. In a blog post Tuesday, Williamson wrote that he has experienced growing performance problems running a sizable EC2 installation for a customer, which he believes are tied to growing load on Amazon’s servers. Williamson said he has needed to buy larger instances to maintain the same performance, increasing his client’s costs. 

“The problems that we are starting to see from Amazon are more than just the overhead of a virtualized environment,” he wrote. “They are deep rooted scalabilty problems at their end that need to be addressed sooner rather than later.

“Has Amazon become over subscribed?” Williamson wondered. “Sure feels like it, as we are being ‘taxed’ by being forced to move up their offering stack to just get the same level of performance we are currently enjoying. It appears that even Amazon have a limit to what they can scale to.” 

Cloudkick Weighs In
The post prompted Cloudkick to post charts showing periodic ping latency between EC2 nodes. “Alan Williamson’s post on EC2 oversubscription seems to make a lot of sense,” Cloudkick observed. “The network behind EC2 appears to be experiencing very sporadic latency issues.” UPDATE: Cloudkick has run additional data over a longer time frame and says the latency issue started just after Christmas.  

Cloudkick provides management tools for cloud infrastructure at both Amazon and Rackspace, and is hosted on the SliceHost service owned by Rackspace.

Amazon says the issues seen by Williamson and Cloudkick are not capacity related, and has pledged to work with customers experiencing problems. “I have been contacted by Amazon regarding this issue now and hopefully the data we provide them can help them diagnose any problems that may exist,” Williamson reported in a follow-up post.

Cloud Rivalry
The cloud rivalry between Amazon and Rackspace was underscored by this week’s announcement that cloud startup Encoding.com had shifted its primary operations from EC2 to Rackspace Cloud Servers, while retaining some operations at Amazon.

In conjunction with the Encoding.com announcement, Rackspace revealed results of data from The Bitsource comparing performance of EC2 and Rackspace Cloud Servers. The study, which was commissioned by Rackspace, favored Cloud Servers on most key metrics.

UPDATE: Shortly after we posted this, the Rackspace Cloud experienced an outage in which Encoding.com was briefly offline.

Amazon (AMZN) has forged a dominant position in the cloud computing market, and has reinforced its early leadership by consistently cutting prices and adding features. With the emergence of cloud performance monitoring tools and a growing field of competitors, it’s not surprising that the performance of Amazon Web Services will be closely watched and much discussed. A key point: Kinton says that thus far Amazon has met all of its service level agreements.

UPDATE: There’s been an active discussion of Amazon’s performance on Slashdot and in worthwhile blog posts by Reuven Cohen (Oversubscribing The Cloud) and Chris Hoff (Over Subscription vs. Over Capacity: Two Different Things). The distinction Chris is making is an important one, and will be familiar to those who’ve been involved in the historic debates about overselling and “unlimited” accounts in the shared hosting industry.

A screen shot of Cloud Exchange, which displays dynamic pricing for Amazon Web Services.

Earlier this week Amazon Web Services announced Spot Instances, which allows customers to buy computing power on Amazon’s EC2 cloud computing platform based on fluctuations in price. This dynamic pricing allows customers to save money by running applications when prices fall below a designated level, or outbid other customers for available capacity. How can you take advantage of these pricing trends?

Cloud Exchange provides a visual interface for the spot pricing data Amazon is making available to developers. The app was developed by Tim Lossen (who has made the code available via GitHub) and graphs pricing data on different instance types in three availability zones. We found this site via Matt Sherman, who argues that pricing is Amazon’s core competency.

Here are a couple of additional resources and data points regarding Amazon Web Services:

• Ryan Kearney’s post on Comparing CDN Performance that looks at how Amazon CloudFront’s content delivery performance compares to similar offerings from Rackspace, Go Grid and Simple CDN.
• Amazon’s CloudFront began offering streaming video, a development that CDN analyst Dan Rayburn predicts will disrupt the market for content delivery. “While I don’t see Amazon changing the CDN landscape over night, they are already starting to have an impact on the market and as they continue to add more functionality to CloudFront, their impact will only continue to grow,” Dan writes.  
• Guy Rosen at Jack of All Clouds has an early analysis of adoption of the new West Coast availability zone on EC2 with existing zones for the East Coast and Europe.  

This type of dynamic pricing will allow some customers to save money by running applications only when spot prices fall below a designated price point, while also allowing users to quickly run large jobs by outbidding other customers for available capacity.

Large Scale Market Pricing
“This is the first step on a large scale towards ‘market pricing’ for computing based on offer and demand,” writes Thorsten von Eiken on the RightScale blog. “I know many people have been dreaming about something like this and a few startups have started to offer a compute market of some sort. But with Amazon’s offering it is now available on a large scale to anyone.”

There’s a lot of focus in today’s announcements about how Spot Instances can help customers save money. But this refinement provides tools for Amazon to more effectively monetize capacity on its infrastructure. As spot pricing gains in popularity – as it surely will – Amazon could get more revenue and mileage out of its infrastructure, especially during hours where there currently is low demand (and thus lower spot pricing). It could also build in pricing power for Amazon at times when it runs shy on capacity, a scenario in which buyer competition for limited resources could drive prices higher.

But thus far Amazon’s strategy has been to reduce prices whenever possible to boost adoption of its cloud computing services, especially by enterprise customers looking for the most bang for their computing buck in a difficult economy. Amazon’s mighty infrastructure – backed by its growing network of data centers – is a competitive differentiator.

“Spot Instances are an innovation that is made possible by the unparalleled economies of scale created by the tremendous growth of the AWS Infrastructure Services,” said Werner Vogels, Amazon’s CTO, writing at All Things Distributed. “The broad Amazon EC2 customer base brings such diversity in workload and utilization patterns that it allows us to operate Amazon EC2 with extreme efficiency. True to the Amazon philosophy, we let our customers benefit from the economies of scale they help us create by lowering our prices when we achieve lower cost structures. Consistently we have lowered compute, storage and bandwidth prices based on such cost savings.”

Pricing History Provides Guidance
Vogels said Amazon will provide a history of the Spot Price for each instance type in each availability region via the AWS Management Console. “Spot Price history is a valuable tool in helping customers use what-if scenarios to determine right pricing level for a particular workload,” he said.

Amazon cloud evangelist Jeff Barr noted that heavy use of spot instances could require more sophisticated management of a customer’s EC2 capacity to account for pricing fluctuations. “From an architectural point of view, because EC2 will terminate instances whose bid price becomes lower than the Spot Price, you’ll want to regularly checkpoint work in progress,” Barr writes. “Many types of work are suitable for this incremental, background processing model including web crawling, data analysis, and data transformation (e.g. media transcoding). It wouldn’t make much sense to run a highly available application such as a web server or a database on a Spot Instance, though.”

Craig Balding at Cloud Security points out one potential weakness: the potential for the dynamic pricing model to be exploited by denial of service attacks, which could lay the groundwork for black hats to manipulate pricing on AWS or launch a resource-wasting economic DOS.

Amazon’s press release included examples of the use of AWS spot pricing in applications for pharmaceutical companies like Eli Lilly, Varian Inc. and Pfizer and in batch processing of market analytics for financial services clients of Clarity Solutions.

“As customers continued to expand their use of AWS, they started asking if additional pools of capacity were available, even if only for a few hours at a time. Some customers were looking to reduce costs in exchange for being flexible as to when they run their application; others told us they were willing to pay more when they had urgent, high volume needs,” said Peter De Santis, General Manger of Amazon EC2. “Because of the dynamic nature of supply and demand in the Amazon EC2 environment, we developed Spot Instances to let customers take advantage of our unused capacity while specifying a price they are willing to pay.”

Will the market for pricing of cloud computing capacity look like the financial markets? Joe Weinman of AT&T looks at potential pricing innovation and strategies in Hedging Your Options for the Cloud over at GigaOm.

“This incident impacted a subset of instances in a single Availability Zone,” said Amazon spokesperson kay Kinton. “Most of that subset of instances were back online in 45 minutes.”

The issues started at 4 am East Coast time Wednesday, and affected one of the three availability zones in Amazon’s East Coast operation. The zones are designed to provide redundancy for developers by allowing them to deploy apps across several zones.

“A single component of the redundant power distribution system failed in this zone,” AWS said in its status report. “Prior to completing the repair of this unit, a second component, used to assure redundant power paths, failed as well, resulting in a portion of the servers in that availability zone losing power. Impacted customers experienced a loss of connectivity to their instances. As soon as the defective power distribution units were bypassed, servers restarted and instances began to come online shortly thereafter.”

Amazon is known to operate a major data center in Ashburn, Virginia. EC2 previously experienced brief downtime in both June and July.

• Amazon Web Services Asia Expansion.  Amazon Web Service announced Thursday that an expansion of their services into an Asia-Pacific region will take place in the first half of 2010, when developers and businesses will be able to access infrastructure services from multiple availability zones in Singapore, with other zones in Asia following in the second half of 2010.  AWS services included at launch will be Elastic Compute Cloud (EC2), Amazon Simple Storage (S3), SimpleDB, Relational Database Service (RDS), Simple Queue Service (SQS), MapReduce and CloudFront.  Pricing for web services in Asia will be announced when launched in 2010.
• Hurricane Electric Expands Infrastructure at Equinix.  Internet backbone and IPv6 provider Hurricane Electric will extend its presence to additional Equinix data centers outside of the United States. Hurricane Electric will expand into Equinix Tokyo-2, Hong Kong-1 and Zurich-1 facilities.  Citing an increasing demand for IPv6 content in Asia and Europe as a reason for global expansion, the Hurricane Electric presence in global Equinix data centers will also allow other Equinix customers to easily exchange IP traffic with more than 500 associated IPv6 backbones.  Equinix chief marketing officer Jarrett Appleby said “operating also within our TY2, HK1 and ZH1 centers will put Hurricane Electric in the middle of an existing community of international and local networks and carriers for its next generation IP access service.”
• Level 3 expands in Atlanta. Level 3 announced Thursday an expansion of operations and enhancing local presence in the Atlanta area.  The initiative will provide mid-market enterprises with greater access to Level 3’s services via its extensive backbone network, metro fiber-optic footprint, and a locally focused sales and customer support team.  Level 3 will expand the network in the Atlanta area that already passes nearly 15,000 businesses today.  Following their “link globally and connect locally” mantra, the move will allow Level 3 to provide a competitive alternative for Atlanta area businesses.  Level 3 has announced similar focused expansions in Chicago, Miami and New York in recent months.

Amazon says the new service provides a fully featured MySQL 5.1 database, so any code, applications, and tools that developers use today with their existing MySQL databases will work with Amazon RDS. The service will automatically handle administration, patch management, and backups.

“Many (AWS) customers have told us that their applications require a relational database,” said Adam Selipsky, Vice President, Amazon Web Services. ”That’s why we built Amazon RDS, which combines a familiar relational database with automated management and the instant scalability of the AWS cloud.”

Here’s a roundup of notable resources, analysis and commentary on Amazon RDS and its competitive impact on the cloud ecosystem: 

• Amazon CTO Werner Vogels discusses Amazon RDS in a post at All Things Distributed. “The service takes much of the hassle out of setting up and managing relational databases, such as backups and code patching, freeing up its users to focus on their applications and business,” Werner writes.
• AWS evangelist Jeff Barr provides examples of how to get started with Amazon RDS, saying it ”enables a lot of really enticing development and test scenarios. You can set up a separate database instance for each developer on a project without making a big investment in hardware. Once you’ve deployed RDS for production use, you can easily scale up to larger instance sizes, add additional storage space and make backups with ease. You can easily snapshot a production database and then bring it back to the lab to dig in to a problem.”
• At TechCrunch, Nik Cubrilovic writes that Amazon RDS “makes the task of creating and starting new DB instances easier, but does not mean that your resource allocation will automatically grow and scale with resource requirements. There are existing third-party services, such as Fathom, that are built on AWS and use EC2 to create and manage DB instances.”

• Krishnan at Cloud Ave says the announcement of RDS preempts Microsoft’s rollout of a hosted SQL service for its Azure developer cloud. But he also predicted that Amazon RDS will “crush the Y-Combinator startup FathomDB that offers database as a service that is run on top of Amazon EC2. … Probably, this announcement should also serve as a warning bell for the companies that build their entire business on Amazon ecosystem. They are just one announcement away from complete destruction.”
• Cloud management specialist Rightscale addresses how the new offering relates to its services. “If you run three XL RDS instances the extra cost is already more than a RightScale subscription,” it notes, adding that its “interesting to see how the per-hour price surcharge compares with a more flat-fee subscription to a broad management service. But our core conviction is that we want to offer our customers the broadest choice possible and we’ll support RDS instances in the RightScale dashboard.”
• Mark Bao writes that Amazon RDS “could be very disruptive” but raises questions on performance. “RDS opens an important question while on the topic of database speed optimization: it scales up well, great: but how is the performance between webservers and the new RDS storage location?,” Bao writes. “It’s an open question and we’ll see how it plays out when people mess around with it.”

Now comes an interesting nugget that offers some insight intoAmazon’s data center operations: Randy Bias of CloudScaling says that Amazon is running 40,000 servers to support EC2, citing two sources for the number. Randy uses that as the basis for a calculation of how much revenue Amazon may be generating from EC2 (his ballpark is $220 million to $260 million annually).

Note that the 40,000 server estimate is just for EC2, and doesn’t include the S3 cloud storage service or the servers supporting Amazon’s massive retail operations. But it’s safe to say the new numbers clearly affirms our estimate that Amazon is in the 50,000 server club. See CloudScaling for additional details and Randy’s analysis.