GameTable Online (GTO) is an online game site featuring web versions of popular, real-world tabletop games, based in Northampton, Mass. GTO’s specialty is turn-based strategy games—and the site makes them look and feel like the real thing.

While users enjoy getting to play both free and premium games, game creators and manufacturers benefit from exposing a new generation to classic offline games, like Axis and Allies. And the audience for these games is big—more than 11,000 games of Axis and Allies have been played so far this month. GTO’s audience is highly motivated, and it turns out that their online behavior affects their offline purchases—a survey showed that a whopping 20 percent of the players who were introduced to the game 10 Days in Africa on GTO went on to purchase the game.

To support an audience that fluctuates throughout the day, GTO needed a flexible hosting solution and could cut costs where needed. Their traditional dedicated hosting environment was too costly and turning servers off and on was timely.  They needed the ability to expand to more servers quickly and on demand, to cater to a game’s spiking popularity or allow the site to host multi-player tournaments. Expanding to more servers on the cloud would allow them to offer their clients an affordable price, while scaling as needed. They also wanted to separate their web servers from their game servers so when they made changes to one, it didn’t affect the other. GameTable Online turned to the Rackspace Cloud.

The system GTO came up with using Rackspace Cloud is a perfect fit for their needs. Using Cloud Servers, GTO set up their web, database, and game servers separately. This layout lets them use private web servers for client development and testing, while using their other servers for alpha testing. They also keep a backup of their live servers on Cloud Files, using the Cloud Servers snapshot feature to clone the server images.

The results have been impressive: GameTable Online cut their hosting expenses by 30%, without compromising their bandwidth or service. They are able to easily support varying numbers of players—often 250 players are concurrently logged in and playing. But with the ability to adjust the number of servers according to how many active players are logged in, GTO is able to keep costs (and premium activation fees) as low as possible.

“The ability to start small and grow rapidly is essential to meeting our schedules with a limited staff and budget,” explains GTO Operations Vice-President Robert Eng.

Instead of dealing with server headaches, GTO’s staff is able to focus on adding more games to GTO’s list of offerings, and continually improve user experience. It was the performance and ease of use of Rackspace Cloud Servers that convinced GTO to also start using Cloud Files.

“It’s allowed us to increase efficiencies in how we distribute the game code,” says Eng. “Our Rackspace Cloud Servers are invaluable to us.”

We’re pleased to have GTO bring game strategy to users all over the world on our cloud. Game On.

As of 5:30 AM CDT, Rackspace Cloud engineers are investigating intermittent performance issues within our Dallas / Fort Worth (DFW) data center affecting sites hosted in this data center. Sites may experience intermittent slow load times or present "No Suitable... Ed Gonzales

At approximately 2:00 AM CDT, the Rackspace Cloud engineers identified an issue on MySQL cluster 50-26.wc2 in our DFW data center, and are actively investigating this issue. This cluster is currently running in a degraded state. More updates to come.... Ed Gonzales

As of 7:00pm CDT, Rackspace Cloud engineers are investigating degraded performance issues within our Dallas / Fort Worth (DFW) data center affecting sites hosted in this data center. Sites may experience intermittent slow load times or present "No Suitable Nodes"... Ed Gonzales

Lew Moorman did a keynote, Toward an Open Cloud, at OSCON 2010 in Portland, Oregon last week. It was a big week for Rackspace as we announced OpenStack which Lew discusses further in his presentation. He highlights how Rackspace got started and  has succeeded through the years which is by taking standard based technologies and making them into great services. We have progressed with our cloud computing development, serving over 100,000 customers, and we’re now committed to opening our cloud technologies to the world. We think this is the best way to bring services to our customers. He talks about our OpenStack commitment in 4 opens: open design, open development, open community and open source. Take a look to learn more.

At approximately 1:05 PM CDT, the Rackspace Cloud engineers identified an issue on MySQL cluster 50-46 in our DFW data center, and are actively investigating this issue. This cluster is currently running in a degraded state. More updates to come.... Ed Gonzales

We’ve talked about Standing Cloud, a Rackspace Cloud Tools Partner, and you may have heard about them on TechCrunch. Standing Cloud allows you to easily deploy and manage open source applications on Rackspace Cloud Servers in minutes! Most IT professionals use cloud computing to get their configuration online quickly – Standing Cloud takes it to the next level. They are taking the headaches out of learning how to configure a series of open source applications and have turned it into a simple 3 click process for you.

Standing Cloud currently supports over 50 open source applications, including all the usual favorites such as Wordpress, Drupal, and SugarCRM, as well as more special-purpose systems like Hudson and Mantis.  They continue to add several new applications each month, and not only LAMP-stack apps – there are also Rails and JBoss applications available, and they’re always taking requests.

Learn more about Standing Cloud and how they can help you manage your open source applications seamlessly in their upcoming webinar. In this webinar you will learn how to install and launch an open source application on a  Rackspace Cloud Server in under one minute as well as see how powerful features, such as backup and restore, can be presented in an easy-to-use manner. Details below.

Date: Thursday, July 29th
Time: 11am PST / 1pm CST / 2pm EST

Click here to sign up.

Blake Yeager is the Product Manager for Rackspace Cloud Servers.

It’s been a few months since we announced the Beta release of our Cloud Servers for Windows offering and we’d like to provide you with an update on the progress we’ve made.  With the feedback received from our customers using Cloud Servers for Windows Beta, we’ve been hard at work completing some of the most requested functionality that was not initially released with the Beta and have been carefully monitoring performance and stability.

Rescue Mode

We launched the Rescue Mode feature two weeks ago, which you may have noticed.  Rescue Mode allows you to mount the drive from a Cloud Server that is not booting as a secondary drive on a new temporary Cloud Server.  This provides the ability to repair the Cloud Server or copy any critical data off the drive.

Snapshots/Backups

This week we are launching two more features: On-Demand/Scheduled Images (aka Snapshots/Backups) and the ability to Resize Up your Cloud Server.  With the On-Demand image feature you are able to make a copy of your entire Cloud Server instance (called an image).  This image can then be used to create new Cloud Servers or it can be used to revert your Cloud Server back to the point in time when the On-Demand image was created.  Scheduled Images allow you to specify a daily or weekly time window where the system will create an image of your Cloud Server instance.

Resize Up

The Resize Up feature allows customers to increase the size of an existing Cloud Server.  If after creating a server and configuring your application you realize that you need more memory or disk, a simple click of a button will scale your server up to a larger size.  Right now, you are unable to resize down but it’s a feature we are working on for the future.  While we are continuing to look at options, we recommend that you start with the smallest size Cloud Server that can run your application and only resize up to a larger size when you are ready for the change to be permanent.

Volume Shadow Copy Service (VSS)

Modern versions of Windows include VSS, a standardized mechanism for taking volume-based snapshots.  To ensure consistent backups, VSS signals applications to flush data and pause while the snapshot is being taken.  Our intention is to support VSS pass-through (a process by which Cloud Server backups/snapshots coordinate with the Windows guest VSS to ensure data is consistent), but at the present time, this is not functional.  Since Microsoft does not support snapshots that have been created outside of VSS pass-through, they are unable to offer support for Cloud Server instances created from a snapshot, backup or any instance that has been resized up.  Rackspace will attempt to support any Cloud Server created using these methods, however, until we get VSS pass-through working, we recommend that customers who require full Microsoft support use a built in backup tool such as NTBackup or Windows Server Backup to create and manage their backups.  We will of course notify you once we have full VSS pass-through support.

Additional Changes

In addition to these new features, we’re also making the following changes to the offering based on Microsoft’s recommended specifications and to ensure optimal performance:

• In order ensure that our offerings meet all of Microsoft’s minimum requirements for all versions of Windows we will no longer be offering the 512MB size option.  Any existing 512MB Cloud Servers will be grandfathered in at their current size but all new Cloud Servers must be 1GB or larger.

• We have also changed our vCPU allocation for Windows.  Instead of allocating 4 vCPUs to every Cloud Server we are now scaling the number of vCPUs that we allocate with the size of the Cloud Server.  This change effectively increases the guaranteed minimum CPU time that Cloud Servers are allocated.  The table below details the new vCPU allocations.

It is important to note that the changes we are talking about will only apply to Cloud Servers for Windows.  Sizes and vCPU allocations for our Linux Cloud Servers remain unchanged. As before, Windows Cloud Servers are still allowed to ‘burst’ and utilize any spare CPU cycles on the host.  We will be rolling out the vCPU changes to existing Cloud Servers for Windows customers over the coming weeks.

We are very excited about these new features!  We anticipate being able to remove the Beta label and offering a full SLA around Cloud Servers for Windows in the coming weeks.  We can’t thank our customers enough for providing critical feedback needed to make Cloud Servers for Windows a success. As always we will continue to release new functionality and improved features we hope will help you be successful. If you have any questions or comments, feel free to post here and I will get back to you.

At approximately 10:15am CDT, Rackspace Cloud Engineers identified an issue with our Control Panel. While the CP may be accessible, functionality may be degraded. Our engineers are working diligently to identify and resolve the issues. We will update the STATUS... Ed Gonzales

At this time, the add / delete functionality for Cloud Sites DNS records has been temporarily disabled within the control panel. We are in the process returning this access and should have full functionality restored as soon as possible. UPDATE:... Ed Gonzales

At approximately 11:15 PM CDT our Engineers have identified an issue potentially causing latency and intermittent errors when accessing Webmail. We are working on resolving this as soon as possible. We apologize for the inconvenience. Update: 12:31 CDT Some customers... Ed Gonzales

Rackspace Cloud system administrators will be performing a routine scheduled maintenance on Wednesday July 21st, 2010 from 02:00 am to 03:00 am CST. This maintenance will be conducted on the MSSQL clusters listed below in our DFW (Dallas Fort Worth)... Ed Gonzales

Imagine a world where code used by the biggest clouds is freely available to any developer, anywhere.  A world where that code was a standard used to build private clouds as well as a variety of new service offers.  In this world, workloads could be moved around these clouds easily – you could fire your cloud provider for bad service or lack of features, but not have to rewrite the software to do it.  Imagine an open source cloud operating system that lifts IT to the next level of innovation, just as Linux drove the web to new heights.

Today, we at Rackspace launched an ambitious project called OpenStack that aims to make this new world a reality.

I want to lay out the thinking that got us here and why we think this moment will change computing forever.

“The cloud” at its most fundamental level is all about a massive supply increase in computing power.  The PC era was all about putting a computer on every desk.  The cloud era goes a step further, putting the power of supercomputing at the literal fingertips of every individual at anytime.   Whether it’s enabling a youth soccer coach to schedule practice across the online calendars of 18 families, or helping a scientist fold proteins to design new cancer drugs, or encouraging a frontline employee to instantly and cheaply test a new marketing campaign, the exponential growth in computing power and applications is changing every corner of our economy and society.  And, this era is truly just beginning.  We have seen only a tiny fraction of the potential gains that arise from cheap, ubiquitous computing power.

As this landscape has evolved, some have dismissed cloud computing as just a return to the mainframe era. This view is fundamentally wrong.  Mainframes were available to only the smartest employees at the richest companies.  The cloud is accessible to all, and usable by anyone, at low cost.  Its ubiquity is the source of its power.

However, there is one area where mainframe concepts are intruding into the cloud – the vertically integrated technology stack.  As hardware and software merge into services, the danger of locked down proprietary software stacks are emerging in the cloud space.  The cloud world changes everything, and that is not good to many entrenched interests of the old guard.  Core technologies from operating systems to hypervisors to databases are being used to tie cloud customers into an integrated view of the world.

If the web has taught us anything, it is that open systems, portability, and choice drive innovation.  The open Linux system brought us a mountain of software and tools to help accomplish almost any task.  And, each component, whether a database or a widget could be moved in and out freely based on the job getting done.

We at Rackspace have long talked about an “open” cloud.  And as a service provider built on our Fanatical Support difference, we have never had an interest in creating technical walls around our service.  But, given that no standards tools have emerged to build massively scalable clouds, we too have had to build custom software that creates some level of wall around our cloud offerings.  For months we have debated how to drive greater standards and increase the velocity of cloud technologies in general.  We finally converged on the obvious answer:  open source our cloud technology.

Today, we announced a new open source project that includes those core technologies: OpenStack.  And, we are not alone.  As we looked at all the projects that already existed to drive standards we saw that other efforts were underway that complemented what we have done.  We saw a ton of promise in the Nebula computing project built by NASA and are making it a core part of the project.   Taking the contributions of Rackspace and NASA as a starting point, OpenStack forms a powerful foundation of technologies including, a scalable compute provisioning engine – OpenStack Compute – and a fully distributed storage engine – OpenStack Object Storage.

The community, which we plan to actively support and drive, is live today at openstack.org with code available for download.

Last week we assembled a strong group of cloud community leaders and developers to meet and review the architecture, engage on technology direction and contribute code.  The effort attracted more than 100 participants from 25 companies including hosting companies, telecom providers, hardware manufacturers, cloud ecosystem companies and beyond.  This enthusiasm and collaboration around OpenStack has laid the foundation for a vibrant and innovative approach to building the core software to power the future cloud world.

What do we expect OpenStack to mean for the cloud community?  Some pretty major things.  One, anyone will be able to run this cloud and do it anywhere.  Enterprises and governments will be able to build private clouds.  Service providers will have the same technology used by Rackspace and NASA to build new offers.  Choice and portability are inevitable in this world.  Two, the whole tech ecosystem can build around this foundation.  With wide adoption, there will be a market for new services all around this core engine.  From storage systems to monitoring tools to management systems, there is no end to what can be attached to the core project.  Three, the cloud will advance faster than ever.  Between just NASA and Rackspace, an army of developers are committed to the continued advancement of OpenStack.  With our emerging supporters in the project, we expect to dramatically expand that army.  Finally, a core set of standards will be freely available and totally open.  New technologies can be attached.  Better solutions will be driven into the product.  And, the use of this powerful technology will not tie you to the use of any other technologies.

For our customers, we think there are many benefits that flow from these community gains.  Not only will this help our offers develop faster and more transparently, but our customers can run private editions of our core systems in house or in our managed hosting operation.

We could not be more excited about the launch of this project and the enthusiasm around it.  As a company that has invested a great deal in the development of cloud technologies, we did not take the decision to open source lightly.  We think this decision will serve our interests and those of our customers.  While we at Rackspace hire top developers and engineers to make sure our technology is second to none, seeking a technology advantage has never been our approach.    We have our own vision about how to deploy this technology and serve customers – by giving them seamless access to scalable computing with the trusting partnership that comes through Fanatical Support.  But, there will be many approaches and winning formulas.  We think by welcoming those approaches and driving standards and more rapid innovation we will all win.

We hope you join us in this cause.  We know there are many parties who might want to join us in the effort, please reach out to us on this blog or through openstack.org.

We look forward to updating you as we make progress. Stay updated by following OpenStack on Twitter, joining the IRC channel #openstack, joining the OpenStack mailing list or following the OpenStack blog.

Rackspace Cloud system administrators will be performing a routine scheduled maintenance on Monday July 19th, 2010 from 02:00 am to 03:00 am CST. This maintenance will be conducted on the MSSQL clusters listed below in our DFW (Dallas Fort Worth)... Ed Gonzales

At approximately 3:27 PM CDT, our DFW datacenter experienced a brief service interruption. During that time, Cloud Sites customers would have experienced intermittent site accessibility issues. Update: As of 3:36 PM CDT, the brief networking instability at our DFW Data... Ed Gonzales

This post was written and contributed by Randy Hayes, CEO and Co-Founder of Capacity Calibration, a Rackspace Cloud Tools Partner.

One of the key benefits of cloud computing is the on-demand capacity it provides for either expected or unexpected surges in traffic.  There’s a Catch-22 here, which is that nobody knows when they will need extra capacity or how they will handle it without testing, and fortunately the cloud itself provides the solution – the ability to handle extra capacity and the ability to generate it go hand in hand.

Capacity Calibration, or CapCal, was designed for the cloud of the Internet itself, and for several years it used the seti@home model where people from all over the world downloaded the CapCal agent and received $0.30 per hour that their computer was used for testing.  While this allowed some very large tests to be run, $0.30 is 10 times the cost of a 256MB Rackspace Debian Linux instance and far less reliable (since agents on the Internet can drop out or check in at any time).  So now, rather than having a network of over 15,000 computers with no more than 10% to 15% being online at a given time, we can spin up as many as we need on the Rackspace Cloud in less than 3 minutes and tear them down when finished testing.

Clearly the emergence of cloud computing gave CapCal a superior platform for load generation, but the fact that is was designed for the cloud in the first place has given us a decided advantage – testing the scalability of a solution designed to test scalability is another Catch-22, and fortunately we were able to encounter and solve these issues early on.  Because of its distributed, Web-based design we have been able to run tests in excess of 300,000 users delivering several million hits per minute requiring over 500 agents.

The CapCal agent running on the Rackspace Cloud takes less time to start up than the Windows desktop I use for development and its available bandwidth is far greater than what 99% of all the older CapCal agents ever had.  This means we can generate way more virtual users per agent without affecting the test results and this helps to keep costs and complexity way down.  For example, NASA used CapCal to prepare for the first live broadcast of its shuttle launch in 2001 and the average agent bandwidth was only 56k, so what took over 1000 agents to generate in those days can now be done with 50 or less.  But 50 is still a big number when it comes to physical boxes, and once again the cloud solves the problem simply and inexpensively.

Using the Rackspace Cloud API, a new instance of the CapCal agent is fired up in a minute or two at the most and reports directly to the CapCal server, which can handle up to 1,000 agents in its base configuration.  Each agent can generate anywhere from 200 to 1,000 virtual users depending on the bandwidth requirements of the application under test, so the kinds of tests that most people need to run (500 to 5,000 users) will require just a handful of agents.

Thank you, Rackspace Cloud, for helping to reinvent the industry and making the sky the limit of what we can do in web application testing!

Paul Ford, from Rackspace Corporate Development, is Your Connection to the Rackspace “Cloud Tools Partners” Ecosystem . To find out more about how CapCal and other tools can increase your productivity, satisfy your IT needs, and generally make your life easier, contact him any time at paul.ford@rackspace.com

This post was written and contributed by Nati Shalom, CTO & Co-Founder of Gigaspaces, a Rackspace Cloud Tools Partner.

The NoSQL movement faults the SQL query language as the source of many of the scalability issues that we face today with traditional database approach.

I think that the main reason so many people have come to see SQL as the source of all evil is the fact that, traditionally, the query language was burned into the database implementation. So by saying NoSQL you basically say “No” to the traditional non-scalable RDBMS implementations.

This view has brought on a flood of alternative query languages, each aiming to solve a different aspect that is missing in the traditional SQL query approach, such as a document model, or that provides a simpler approach, such as Key/Value query.
Most of the people I speak with seem fairly confused on this subject, and tend to use query semantics and architecture interchangeably. So I thought that a good start would be to provide a quick overview of what each query term stands for in the context of the NoSQL world. Then, I’ll try to break some common misconceptions — which led me to come up with the YeSQL term.

Common Query Semantics in the Post Only-SQL world
The following are some of the common query semantics in the NoSQL world. For those that are interested in code examples i’ve linked each category with the relevant GigaSpaces reference API.

• Key/Value query: Key/Value query, as the name suggests, is probably the most basic form of query. Each data item is associated with a unique identifier (key). In the NoSQL world, memcache is one of the most common implementations of such an interface. A common pattern to perform complex queries with memcache is to defer them to an underlying database which is used as a search engine. The result of these queries is a key or set of keys that is then used to perform subsequent fetching of the values through the memcache data store. The reason they gained new momentum in the post-SQL world is because they lend themselves fairly natively to the concept of partitioning and distribution, which is a key piece in making a data store scalable. In other words, people were willing to trade the rich query functionality provided by most traditional RDBMS for scalability with only basic query support if that was their only choice.
• Document-based query: The roots of this model are in the search engine world, where it is very common to store different types of documents, even if each one represents a completely different object. In the NoSQL world, a document is not the typical Word or PowerPoint that you would see in search engine, but rather objects in the form of Jason or XML, or binary objects associated with a set of key/values, as in the case of Cassandra. In SQL terms, a document can be seen as a blob that is associated with a set of keys, each indexed independently and maintaining a reference to this blob. Each blob can be a different type (tables), each blob can have different set of associated indexes (keys). Matching is done through the associated indexes. The result-set often includes multiple types, each containing a different set of data. Because the indexes and the blob don’t need to conform to a strict structure of rows and tables, it is referred to as “schemaless” — i.e., it can have different versions of the same type, and add fields to new types without having to modify any table or update older version of the data. Examples that support the document model are CouchDB and MongoDB.
• Template query: Template queries were common in JavaSpaces and even in later versions of Hibernate. With template-based matching, you can fetch an object based on class type or inheritance hierarchy, as well as values of the attributes of that object. In more object-oriented versions of template matching you can also perform matching based on specific items within a graph attribute. GigaSpaces is one of the better-known implementations that support the JavaSpaces template query model.
• Map/Reduce: Map/Reduce is often used to perform aggregated queries on a distributed data store. A simple scenario would be Max, or Sum. In such scenario the query request need to be executed independently in each partition (Map) and then aggregated back to the client (Reduce). An implicit Map/Reduce can take a certain query request and spread the execution of that query implicitly. The client gets the aggregated query as if it was a single query. The explicit model allows you to execute code in free-form, where you can control the mapping model — which call goes to which data, the code to run in each node (aka tasks), and the results. In a typical Hadoop implementation, Map/Reduce is often done through the explicit model. Frameworks like Hive and Pig provide an abstraction model that can handle that process implicitly.
• SQL query: If you think about it, SQL is yet another form of dynamic language that was specifically designed for complex data management. With SQL, data is often ordered in tables and rows. Some of the query semantics in SQL, such as Joins, distributed transactions, and others are known to be anti-pattern for scalability. It is mostly this aspect of SQL semantics that associates SQL with scalability limitation. Examples of NoSQL implementations that support SQL are Google Bigtable using JPA, Hive/Hadoop, MongoDB and GigaSpaces. I will discuss in further details below what that actually means.

YeSQL — There’s Nothing Wrong with SQL!
Now that we’ve covered some of the concepts behind query formats, it becomes more apparent that there is nothing really wrong with SQL. Like many languages, SQL gives you a fairly long rope with which to hang yourself if you choose to, but that is true of almost any language. If you design your data model to fit into a distributed model, you may find that SQL can be a fairly useful format to manage your data. A good example is Hive/Pig/Hbase and Google JPA /Bigtable. In both cases the underlying data store is based on a scalable Key/Value store, but the front-end query language happens to be SQL-based. MongoDB aims toward a similar goal with the main difference that it provides SQL-like support and doesn’t fully comply with any of the existing standards.

It’s About the Architecture, Stupid!
NoSQL implementations such as Hive/HBase as well as JPA/Bigtable can be a good example of how next-generation databases can support both linear scaling and a SQL API.
The key is the decoupling of the query semantics from the underlying data-store as illustrated in the diagram below:

Supporting a SQL API on top of a NoSQL data store in Google Bigtable

Convergence is Underway
Last week I spent some time at the Hadoop summit. Hadoop created a fairly generic substrate that led to an innovative ecosystem behind it. There are already many new frameworks today that provide different levels of abstraction to the way Hadoop manages data in both query and processing, such as Hive, Cascading, and Pig. Many of them provide tools that the original creator of Hadoop never even thought of.

Which brings me to the point that we can apply the same decoupling pattern I mentioned above to support a document model in connection with SQL. In other words, I believe that most of the leading databases will support all of the semantics listed above, and we won’t have to choose a database implementation just because it supports a certain query language.

We’ve already seen a similar trend with dynamic languages. In the past, a language had to come with a full stack of tools, compiler, libraries, and development tools behind it, making the selection of a particular language quite strategic. Today, a JVM in Java or a CLR in .Net provides a common substrate that can support a large variety of dynamic languages on top of the same JVM runtime. Good examples are Groovy and Java or Jruby.

Final Words
As I pointed out throughout this post, SQL is actually a fairly good query language and will continue to serve a major role in the post only-SQL world. However, the concept of one size fits all doesn’t hold up. The data management world is going to be built from a variety of tools and data management languages, each serving a particular purpose. Ideally, we should be able to access any data using any of several query languages, regardless of how it was stored. For example, I should be able store a JSON object using a document model and then, at any time, query that JSON object using SQL query semantics or a simple Key/Value API.

Paul Ford, from Rackspace Corporate Development, is Your Connection to the Rackspace “Cloud Tools Partners” Ecosystem . To find out more about how GigaSpaces and other tools can increase your productivity, satisfy your IT needs, and generally make your life easier, contact him any time at paul.ford@rackspace.com

Last month Bitcurrent issued the findings of a comprehensive study on cloud computing performance. The Rackspace Cloud was among the five top providers targeted for rigorous testing and we’re proud of how we came out.  The other cloud platforms were Salesforce’s Force.com, Google’s App Engine, Terremark’s vCloud, and Amazon’s Elastic Compute Cloud.

Given the growth of the cloud market and the fact that the term “cloud” has come to mean many things—from hosted web applications to fully-fledged platforms—Bitcurrent wanted to assess the current state of cloud performance.  They wanted to see if the promise of pay-as-you-go economics and elastic capacity could live up to the demands of large-scale enterprises.

As part of their testing regimen, Bitcurrent teamed up with Webmetrics, a web monitoring firm. They tested four aspects of performance across the five cloud networks. The test agents are described below:

• A simple web request, to measure the responsiveness of the system for a trivially small, static object.
• A request for a large (2 Megabyte) object, to test the network throughput.
• A request that triggered a million mathematically-intensive calculations, to test computing power.
• A request that searched 500,000 rows of a database for a string, to test the back-end I/O of the system.

The tests were executed, using the agents, from multiple locations around the world. The results of the tests can be illustrated in the diagram below.

The Rackspace Cloud performed well across all categories—responsiveness (1-pixel GIF), network throughput (2MB GIF), computing power (CPU), and I/O (database search).  While no provider was a clear winner in every category, Rackspace Cloud was in the top tier for each test. We even managed to surprise the testers on the CPU front:

“The PaaS providers did well, largely because of their shared storage model that is optimized for large data sets across many machines.  We were surprised by how well Rackspace’s cloud handled the simultaneous traffic despite being a PaaS-based solution.”

The summary of the Bitcurrent report as well as the complete report, with all the testing details, are available for download.

At approximately 8:50 AM CDT, the Rackspace Engineers identified an issue causing degradation on the Cloud Sites control panel. We continued to investigate this matter until further information was available. Update: As of 9:20 AM CDT our Cloud Sites control... Ed Gonzales

At approximately 8:50 AM CDT, The Rackspace Cloud engineers identified an issue on MySQL cluster 50-39 in our DFW data center, and are actively investigating this issue. This cluster is currently running in a degraded state.UPDATE: As of 9:35 AM... Ed Gonzales