As cloud computing dominates more and more aspects of the tech world, similar-sounding but confusingly different something-as-a-service acronyms keep piling up. You've probably heard of SaaS (Software as a Service), since it applies mostly to cloud services delivered to end users. 

But what about IaaS (Infrastructure as a Service) and PaaS (Platform-as-a-Service)? Even many tech professionals can't explain the differences without babbling incoherently. 

Fortunately, this new infographic from IaaS provider ProfitBricks does a good job of explaining the differences and who uses which one for what. Enjoy.

If you're looking for a U.S.-based cloud provider to host your business applications and data, you might want to take note of one new report that ranks the performance levels for providers in 2012. The results might surprise you, because it turns out that size and location may make a big difference in overall performance.

The data was provided to ReadWrite via Compuware's CloudSleuth service, which monitors the on-going availability and response time of providers around the world. The CloudSleuth application is hosted by cloud providers just like any other customer's app and is constantly polled for metrics from 17 locations around the U.S.

Global Cloud Availability, Past 30 Days

The data for the rank list was aggregated over the 2012 calendar year, measuring average response time, availability, and consistency. Consistency is the standard deviation of the response time, indicating the range of values for the reported response times over the course of the year, according to Compuware APM Technology Strategist Stephen Pierzchala.

To come up with the overall ranking, Pierzchala explained, the company combined the individual ranks for all of the three categories.

In 2012, based on overall rank, the top five providers were:

1. Connectria (US Central - Missouri)
2. Layered Tech (US Central - Illinois)
3. Qube (US East - New York)
4. Layered Tech (US South - Texas)
5. Layered Tech (US Central - Missouri)

Not exactly widely known providers, are they? The players with more notoriety - like Rackspace, Google, Windows Azure and Amazon Web Services - are conspicuously missing from the top five listing. Looking at the rest of the list, Rackspace's Texas facility rolled in at number 6 overall, with Windows Azure's Illinois data center tying for ninth. Google App Engine was ranked number 15, and the highest Amazon Web Service rank was the EC2 facility in California, which came in at number 26 in a list of 38.

Fly-Over States Rock

There's another interesting pattern, one that leaped out as soon as Pierzchala pointed it out.

"If you look at the first 10-15 providers, they're almost all located in the center of the U.S.," he noted.

Location, it seems, makes a big difference in where you choose to host your cloud applications. If you are one of the coasts, there's a certain logic to hosting your cloud applications and data in a facility that is in the middle of the country, where all users, wherever they are, can roughly traverse as the same amount of network to get to the actual cloud machine. Balancing your data and app use geographically can provide real benefits.

North America Cloud Response Times, Past 24 Hours

That the top-ranked providers are all smaller providers may not be an accident, either. Larger providers may have slower response times ands availability by virtue of the fact they are working with so many customers and web applications. If that's the case, there may be something to using a smaller provider that's not located on one of the Coasts.

The full overall rank list is below, followed by the top 10 providers in terms of average response time and availability for 2012, respectively.

2012 Top U.S. Cloud Providers

1. Connectria (US Central - Missouri)
2. Layered Tech (US Central - Illinois)
3. Qube (US East - New York)
4. Layered Tech (US South - Texas)
5. Layered Tech (US Central - Missouri)
6. Rackspace (US South - Texas)
7. GMO Cloud (US West - California)
8. PeakColo (US West - Colorado)
9. SoftLayer (US East - Virginia)
9. TekLinks (US South - Alabama)
9. Windows Azure (US Central - Illinois)
12. SoftLayer (US South - Texas)
13. SoftLayer (US West - Washington)
13. US Signal (US Central - Michigan)
15. Google App Engine
16. ElasticHosts (US West - California)
17. BlueLock (US Central - Indiana)
18. CloudSigma (US West - Nevada)
19. Dimension Data (US East - Virginia)
20. ElasticHosts (US South - Texas)
20. TekLinks2 (US South - Alabama)
22. SoftLayer (US West - California)
23. GoGrid (US West - California)
24. Cartika (US South - Texas)
25. Dimension Data (US West - California)
26. Amazon EC2 (US West - California)
27. Terremark (US East - Florida)
28. eApps (US East - Georgia)
29. Amazon EC2 (US East - Virginia)
29. Green House Data (US West - Oregon)
31. eApps (US East - Virginia)
32. GoGrid (US East - Virginia)
33. Voxel (US East - New York)
34. Claris Networks (US South - Tennessee)
35. Green House Data (US West - Wyoming)
36. Voxel (US West - California)
37. Bit Refinery (US Central - Colorado)
38. PhoenixNAP (US West - Arizona)

2012 Top 10 U.S. Cloud Providers (By Average Response Time)

1. Windows Azure (US Central - Illinois)
2. Layered Tech (US South - Texas)
3. Rackspace (US South - Texas)
4. Connectria (US Central - Missouri)
5. SoftLayer (US South - Texas)
6. Layered Tech (US Central - Illinois)
7. Qube (US East - New York)
8. Layered Tech (US Central - Missouri)
9. PeakColo (US West - Colorado)
10. BlueLock (US Central - Indiana)

2012 Top 10 U.S. Cloud Providers (By Availability)

1. Layered Tech (US Central - Illinois)
2. Connectria (US Central - Missouri)
3. SoftLayer (US East - Virginia)
4. Terremark (US East - Florida)
5. US Signal (US Central - Michigan)
6. SoftLayer (US West - Washington)
7. GMO Cloud (US West - California)
8. Green House Data (US West - Oregon)
9. Amazon EC2 (US East - Virginia)
10. Qube (US East - New York)

Story images and data courtesy of Compuware.

Yesterday, Facebook bought Parse, a San Francisco startup with a service designed to greatly simplify the process of creating mobile and Web apps. Today, Parse's rivals are doubtless celebrating because the (reportedly) $85 million acquisition effectively puts a big seal of approval on their techniques for automating some aspects of app development.

(See also: What Happens When Almost Anybody Can Build A Mobile Business App?)

In tech jargon, outfits like Parse are often called "backend as a service" (or, worse, BaaS) companies. But they could be better described as mobile cloud-service companies. They offer services designed to easily tie mobile apps into the cloud, providing a host of automatic "backend" functions such as data storage and connections to social networks. That allows developers to focus on the core elements that make their apps sing instead of doing a lot of complicated integration with cloud systems.

One of the companies paying closest attention to Facebook's move is Boston-based Kinvey, one of Parse’s biggest rivals and a startup eager to see this cloud-service market really hit the big time.

What This "Validation" Means

Over the space of a few months in 2011, three startups effectively created this app-service automation market. Parse, StackMob and Kinvey promised easy cloud integration to mobile developers, but lookalikes quickly surfaced. Cocoafish (acquired by Appcelerator, Tiggzi (now Appery.io), FeedHenry, Applicasa from the startup end, new services from the likes of Sencha (Sencha.io) andeven Apple (iCloud) joined the fray. IBM and SAP now also offer similar cloud solutions.

(See also: Parse Offers "Backend as a Service" to Mobile Developers)

Some critics wondered if the industry segment had become too crowded and if all the outside entrants would doom the three original backend-service providers. They were small, their business models were unproven and their stories (i.e., “we provide backends so you don’t have to") were quickly in danger of being drowned out by competition claiming the same thing. There was a time in 2012, before it raised its first venture funding, when Kinvey had serious doubts if it would make it.

A map of the BaaS ecosystem from Kinvey from February 2012

Then these companies, which initially had started as developer tools, started turning into actual businesses. StackMob and Kinvey found that big companies were really interested in their services. Parse started attracting brands like the NFL's Green Bay Packers, Hipmunk, Armani and the Food Network.

(See also: Bringing Enterprise Data To Your Mobile Workers)

It's no coincidence that Facebook named Parse, StackMob and Kinvey (along with the likes of PhoneGap and Sencha) as preferred technology partners last week. These are companies with useful skill sets. All three have done extensive work with Facebook in the past. 

Many in the tech community associate “validation” of a new technological or business approach with startup venture funding or outright acquisition. But the likes of Parse arguably found validation much earlier, with the arrival of big, high-profile customers. 

“Lots of people are saying [the Parse acquisition] ‘validates’ the space,” said Sravish Sridhar, CEO and co-founder of Kinvey. “I disagree. The space was validated when brands like J&J, Aetna GSN and Cadillac began trusting their data and apps to BaaS."

Sridhar has a point. But until Parse sold out, none of these startups had entered the "big money" realm of tens of millions in funding, revenue or acquisition. In that sense, the Facebook-Parse deal has definitely lifted the prospects of StackMob, Kinvey and the rest of their competitors.

Where Do These Startups Go From Here?

Parse reportedly had a long line of suitors. Facebook won the bidding, but Dropbox, Google and Yahoo also all apparently had interest, according to Wall Street Journal reporter Evelyn Rusli. A variety of other companies have also shown interest in the backend-service startups, including Salesforce (customer relationship management), Intel (chip manufacturing and developer tools) and classic enterprise service providers like IBM and SAP, which have acquired mobile enterprise application platforms (MEAPs) in the past. 

Dropbox had also bid for Parse-- didn't meet fb's offer though.. google & yahoo also expressed interest ---

— Evelyn Rusli (@EvelynRusli) April 25, 2013

When parsing what the Parse acquisition means to companies like Kinvey and StackMob last night, Kinvey’s head of marketing Joe Chernov turned to Sridhar and said, “Do you know of any other tech space that has so many different kinds of big companies wanting to acquire its vendors?"

It's a good question and one that should have the likes of Kinvey and StackMob hi-fiving, jumping in their seats and making plans for happy hour. 

Twitter will be the next company to watch. It recently bought Boston-based Crashlytics and Bluefin for a total a little less than $200 million. Twitter is beefing up on its own application ecosystem (see: Vine and Twitter Cards) and could very easily find a place for backend services in its app efforts.

Guest author Robert Rizika is CEO of ProfitBricks USA.

Cloud computing is so hot right now that the cloud pretenders are crawling out of the woodwork. You don't have to look very far to find consultants who know nothing and "cloud services" that consist of little more than tacking the word "cloud" onto an otherwise ordinary service.

Ironically, though, some of cloud computing's biggest issues come from the paragon of the technology: Amazon. For the past several years Amazon Web Services (AWS) has been synonymous with cloud computing in the startup world. Startup CEOs hand Amazon their credit cards as a "checkbox item" - just like our grandparents called Bell Telephone when they needed a phone.

But times are changing. Danny Rimer of Index Ventures recently commented on how Amazon is losing its grip on the cloud computing space, even as cloud computing continues its massive growth.

(See also Amazon: Can It Stay King Of Cloud Computing Forever?)

Amazon And The Promise Of Cloud Computing

Is Amazon moving away from cloud computing's core values? The promise of cloud computing lies not only in its technology but in its pricing and packaging structure. The idea is pretty simple: pay for the compute power you use. If you need more, you buy more. If you need less, you buy less.

This is in obvious contrast to past models in which you buy hardware for your on-premise data center or your colocated space, then live within whatever means you own. That capital expense was slow to react to shifts in demand and put you at risk of over- or under-capacity. Hourly billing freed us from that - most users saw a clear progession toward per minute billing to even more closely match the demand curve.

Why, then, when you use an Amazon server for 62 minutes do you still get charged for 120 minutes? Multiply that by a couple hundred servers over a year and the difference adds up. How come AWS gives its best prices only with 1-3 year contracts, with up front fees? That sure seems a lot like that step-wise CAPEX spend cloud was supposed to avoid.

AWS’s pricing page includes several terms that seem anti-cloud, for example:

It's A Technology Issue

If you look at the technology stack, things haven’t progressed much in the last 5 years there either. Have you ever wondered why you get offered stock sizes that Amazon selects for you instead of the exact CPU, RAM and block storage you want?

Because AWS uses Ethernet-based networks, once a virtual machine has been provisioned on physical hardware, it’s stuck there until you turn it off. That's why Amazon users occassionally get emails asking them to turn off their virtual machines.

As Amazon plays a Tetris-like game of fitting VMs on hardware, you end up paying for capacity you didn’t want just to make Amazon’s life simpler.

There’s got to be a better way, right? Cloud Computing 2.0 providers are built on technologies like InfiniBand, that allow for 80-gigabit speeds within the network - making it easier to separate CPU cores, RAM and storage to provide infinite options. It also allows a similar flexibility in billing.

That can make a big difference. Cloudability, says the 100 companies it surveyed waste 40% of their Amazon spend buying too many resources.

Cloud computing is supposed to let you buy your core setup and then become elastic, expanding as you need it and contracting when you don't. If you need computing power for 65 minutes you should pay for 65 minutes. If you need it for three hours, you should pay for three hours. That's the way we can truly achieve the promise the cloud offers.

Image courtesy of Shutterstock.

Guest author Jeetu Patel is general manager of EMC Syncplicity.

As enterprise software moves into the world of Software-as-a-Service (SaaS) and consumer technology innovations invade the workplace, how companies evaluate enterprise software vendor changes dramatically in many way. Yet, in some ways, it remains the same.

First, five things that SaaS changes forever:

1. Quality User Experience Drives Adoption

Gone are the days when IT could mandate software solutions with a less than stellar user experience. Today, users will go rogue and adopt consumer apps over enterprise-approved software if it makes them more productive and more mobile. Before you select an enterprise SaaS solution, put yourself in the users’ seat and compare the experience to leading consumer apps. Do they match up? If not, you better keep looking.

2. Simplicity Trumps Feature-Rich

For decades, enterprise software providers have jammed features into their products to meet every IT and user need. The mobile first, cloud-computing world is all about apps that do one thing really, really well. A portfolio of simple, elegant products that are easy to use and easy to implement makes more sense than a complex, comprehensive solution with a long roll-out time and a steep learning curve.

3. Continuous Improvement Is Expected

The 18-month product release cycle is a thing of the past. Today's users demand constant improvements to the way they work - without radical changes that require retraining or disrupt productivity. And you’ll find it’s a great advantage to have your vendor improve features without having to install any software updates. Ask your SaaS providers how they maintain their products with regular releases that streamline and bring the best to the top. What is their pace of innovation?

4. You’re In The Driver’s Seat

One big change in enterprise software is the shift from perpetual licensing that hits capital expense budgets to subscription-based pricing that hits the operational expense budget. Software in the cloud requires no capital investment, expensive roll out or prolonged training. With relatively low initial investements, if a service doesn't solve the problem or users don't adopt it, cancel your subscription and move on.

5. Your Success Is Critical To The SaaS Vendor's Success

Because there are no huge upfront costs, SaaS vendors have to keep customers happy on an ongoing basis. Enterprises have no problem paying good money for software that delivers value, they just have a problem paying upfront for technology that they are not likely to use.

And now, three fundamental ways your relationship with your vendor does not change.

1. You Still Want To know And Trust Your Provider

No matter how much digitization permeates our lives, people will continue to make large software or Software-as-a-Service purchases from people they know and trust. But this is a marathon, not a sprint. You need to partner with vendor in it for the long haul and are accountable beyond any one product or service.

2. Security, Compliance And Management Still Matter

IT technology restrictions may not seem logical to users, but the need to mitigate risk and comply with requirements remains and enterprise reality. A SaaS provider may be highly secure and have a terrific consumer following, but if it doesn’t meet the compliance bar, it doesn’t belong in the enterprise.

3. You Still Need To Know What’s Coming

SaaS companies that cater to the consumer market often introduce new features by rolling them out to users even before they tell them. Enterprises need predictability and a transparency about upcoming changes. Updates may have important implications for security, compliance, compatibility and workflow. Make sure your SaaS vendors communicate proactively.

Gone are the days when IT could mandate which tools were used where. People want to work as efficiently as possible, anywhere, on any device. That dramatically affects how enterprises choose and buy their software, but some things never change.

Image courtesy of Shutterstock.

Guest author Lucas Carlson is the founder and CEO of AppFog.

2012 was a huge year for the cloud, providing developers with a bounty of both public and private Infrastructure-as-a-Service (IaaS) providers, Platform-as-a-Service (PaaS) offerings and Software-as-a-Service (SaaS) options.

Enterprise developers quickly understood the benefits of the cloud, and sunk their teeth into porting projects onto public infrastructure, experimenting with PaaS, developing cloud-based dev/test solutions and incorporating all kinds of SaaS offerings into their daily workflow.

While this is all very exciting for developers, rampant cloud adoption can be a headache for enterprise IT departments. They need to manage, govern and control cloud adoption or deal with chaos.

Uncontrolled Cloud Growth Creates Cloud Sprawl

Cloud sprawl is what happens when enterprise IT and the lines of business it supports all simultaneously but independently employ a variety of cloud services in an uncoordinated fashion.

The business groups could be running apps across a half-dozen different runtimes and using an equal number of different database technologies - some on-premise on a VMware vSphere cloud, others experimenting with OpenStack and some on Rackspace OpenCloud across various regions and zones. The rest could be using Amazon’s public cloud - in Virginia or perhaps Singapore.

In some cases, the enterprise pays for the cloud services. But in many cases, individuals are simply using their personal credit cards and expensing the costs. It’s the Wild Wild West!

Not surprisingly, IT leaders are conflicted. On one hand, they’re inspired by the drive and creativity of their development teams. And they love the velocity of development and releases, and the innovation that results. But they are positively horrified by the thought of making it all work together: “If my department has an app in development on Amazon Web Services (AWS), we need it on our private cloud. Who’s going to make that happen? How much time are we talking here?”

From Cloud Sprawl To Cloud Sanity?

Cloud sanity means having one deployment and management solution responsible for the wide range of app lifecycle processes, including app deployment, deletion, starting/stopping, database service provisioning and tunneling, cloning and re-deploying apps across infrastructures, memory management and more.

It also means having one solution for deploying to AWS, and for migrating from Azure to OpenStack, and for creating new PostgresQL databases, and for syncing up distributed systems through RabbitMQ… well, you get the idea. Right now, sadly, most companies have a single tool (or more commonly each employee has their own version of a script) for each individual task.

The old way of taming cloud sprawl was through vendor lock-in. If your company’s data and apps ended up scattered across a variety of clouds, then you were compelled to pull them all under a single umbrella, be it a private cloud or a single public cloud provider. IT gained predictability from doing this, but lost the ability to experiment outside of the chosen vendor’s bubble. Lock-in means degraded velocity and an elimination of flexibility. Lock-in means an end to innovation.

It doesn’t have to be this way.

Horizontal Hybrid PaaS = Cloud Sanity

A horizontal hybrid PaaS provides a single solution to handle app deployment and management without having to give up on an agile, experimental way of doing things. With a horizontal hybrid PaaS, the enterprise gains freedom from both vendor lock-in and cloud sprawl.

CTOs, CIOs, IT directors and other technology decision makers regain governance of cloud services across the enterprise, as well as single-point-of-truth insight into where apps and services are housed. If they don’t like what they see in terms of performance or agility or any other metric, the workload portability gained from using a horizontal hybrid PaaS makes it easy to switch between clouds, turn services on and off, and switch apps from development to production and back.

Meanwhile, developers gain agility and speed above and beyond anything they’ve experienced in the enterprise. They gain the ability to use and experiment with a variety of cloud services – public, private, Rackspace, AWS, OpenStack, Azure and any language they want – in a way that doesn’t give their bosses headaches about cloud sprawl.

Business leaders are happy because IT is humming along, quickly implementing new ideas and doing more with less. All with reduced cost complexity.

Heterogeneous Clouds In The Enterprise

In 2013, heterogeneity is the name of the game in the cloud. But cloud heterogeneity is still scarier than it needs to be. But heterogeneity can be freeing and enabling, if it's properly managed.

IT should want to have it all: flexibility and control, experimentation and insight, the cost and efficiency advantages of the cloud and a no-surprises approach to cost apportionment.

Image courtesy of Shutterstock.

IBM's decision to throw its considerable weight behind OpenStack has some folks declaring victory for the open source cloud consortium. The hitch? Amazon already claims a considerable lead, with more than 70% of the market, according to Forrester's James Staten, and as one prominent Amazon backer declares, the lead grows daily.

Can Amazon Withstand Sustained Cloud Computing Competition?

The question is whether Amazon can withstand a sustained, concerted attack by nearly everyone else in the cloud computing industry.

So far, according to Adrian Cockroft, director of architecture for Netflix, Amazon Web Services' biggest customer, the answer is an emphatic "Yes." In a Twitter exchange last week, Infoworld's Eric Knorr asked "Is the choice really between OpenStack and AWS? What about other cloud solutions?" Cockroft's response was clear:

@ericknorr that's not the choice. A barrier for people currently using AWS is the feature gap to anything else. OpenStack not closing gap.

— adrian cockcroft (@adrianco) March 14, 2013

When asked which features, in particular, AWS had that OpenStack couldn't match, Cockroft insisted that there is "too much to list," but "Availability Zones and Autoscale Groups" are two features that stand out. Furthermore, Cockroft indicated that OpenStack's six-month release cycle cripples its ability to catch up with Amazon, which rolls out new features (and price drops) on a continuous basis.

In other words, "Abandon all hope, ye who enter here."

Has Amazon Already Won? Yes and No.

But has Amazon already won? If you look at the absolute number of jobs being created, relative to OpenStack or Cloudstack, the answer is yes. Ditto if you go off general interest, as measured by Web searches:

But if you look at relative job growth, suddenly OpenStack has a fighting chance.

Perhaps even more tellingly, a foray into the code contributions for OpenStack suggest a truly dynamic, growing entity, one that is no longer Rackspace's pet project, but rather a true community effort. The most recent data I could find is a year old, but already shows growing influence by Red Hat and others.

Does this matter? Absolutely.

Big Players Will Make A Big Difference

However, I suspect that OpenStack will gain prominence in tandem with a few of its primary supporters gaining outsized influence due to these code contributions. Linux took off as IBM invested $1 billion (and then much more) and Red Hat, in particular, invested armies of engineers to make it into an enterprise-grade operating system standard.

The same will hold true of OpenStack. Right now it's making waves by being the open, community standard. That's nice, but insufficient and somewhat misleading, as Gartner's Kyle Hilgendorf has established. Ultimately, enterprises don't care about community and openness unless the product itself is rock solid.

Which is one reason that Microsoft and Google also can't be counted out. Microsoft holds sway with CIOs, and has been actively welcoming open-source technologies to its Azure cloud service. Google, for its part, was building clouds long before it was cool, and has so many hooks into developers with its various software and services, from Android to Maps to YouTube to Apps, that it cannot help but be a major player. Google Compute Engine's performance compares well against Amazon, too, but it's the ease developers will have tying into Google's services that truly favor it.

Has Amazon won Round 1 of the Public Cloud wars? No question. But some serious competitors are looming, each with attributes (Microsoft, enterprise fealty; OpenStack, community; Google, popular developer services) that give them a real chance to cut into Amazon's significant lead.

Image courtesy of Shutterstock.

Google's cloud-storage service, Google Drive, went down early Monday morning, and returned about two hours later.

Officially, Google Drive went down as of 7:17 AM Pacific time, according to Google's status page. At 9:35 AM PT, Google said it had been resolved. During the outage, however, the other services within the Google Apps cloud - including Gmail, Calendar, Google Talk, and Google Documents remained online.

So far, Google hasn't provided an explanation, but has acknowledged the problem. "We're aware of a problem with Google Drive affecting a significant subset of users," Google posted at 8:10 AM PT. "The affected users are unable to access Google Drive. We will provide an update by 3/18/13 9:10 AM detailing when we expect to resolve the problem. Please note that this resolution time is an estimate and may change."

Update 9:19 AM PT: Google looks like it's on its way to fixing the problem. "Google Drive service has already been restored for some users, and we expect a resolution for all users within the next 1 hours., it added at 8:55 AM. "Please note this time frame is an estimate and may change."

Update 9:49 AM PT: "The problem with Google Drive should be resolved," Google said. "We apologize for the inconvenience and thank you for your patience and continued support."

What this meant for users is that if you wanted to create a new document or spreadsheet, you can - you must simply save it to your own computer for offline editing. Unfortunately, that requires a user to have previously enabled offline editing.

Google-formatted documents couldn't be saved to Google Drive, and if you have a document saved on Drive that you or others wish to edit, you're out of luck. Naturally, other files stored on Drive weren't accessible. However, photos backed up automatically with Google's Google+ service were accessible. It's just part of the complex interdependencies within Google's cloud services, and what happens when one service goes down while the others stay alive.

Clouds Fail

Although we often talk about the cloud in idealized terms, the fact is that cloud services do fail, and when they do, they often fail in a very high-profile manner. The world's largest cloud service, Facebook, stumbled in early January, and last week, Microsoft suffered an outage to Hotmail.com and its complementary Outlook.com service, which last week it blamed on an overheating datacenter. In 2011, RIM's BlackBerry email service was down for days, and outages at Amazon's cloud services have taken down customers including Netflix and others. (Of course, it's also true that in-house services also fail, they just tend to make fewer headlines.)

Now that Google solves the problem and restores service for its users, we'll be forced to wait until Google discovers the root cause of the outage, and hopefully, publishes it. But the outage is yet another reminder that no one service is bulletproof, even for a company with the scale and knowledge of Google.

Image courtesy of Shutterstock.

Guest author Seth Payne is a senior product manager at Skytap.

The rapid emergence of cloud computing is part of a fundamental shift in how IT services are developed and delivered. There are plenty of upsides: The cloud is faster to deploy, allows businesses to be more agile, flexible and innovative, and can often lower overall costs.

But cloud computing also introduces new risks for technical managers. Questions regarding security, data ownership and reliability of the cloud are justifiably top of mind for CIOs and other IT managers.

Time To Stop Hugging Your Servers

When enterprises embrace cloud computing they are, in essence, entrusting a portion of their IT operations to a third party – an outsider. Relying on a cloud provider makes them an integral part of various IT operations, so it is imperative that cloud providers and the enterprises they work with find ways to transcend the traditional, transactional relationship and enter into a truly cooperative partnership.

So, how can cloud providers and enterprises build a cooperative, trust-based partnership?

Communication Comes First

To set the stage, clear lines of communication should be in place so that when issues inevitably arise, enterprises and cloud providers can work together to find resolutions – without pointing fingers and playing blame games.

Cloud providers must hold up their end by offering complete transparency when it comes to their policies, underlying infrastructure and support/maintenance procedures.

While operational risk is inherent when adopting any cloud platform, most enterprise consumers recognize that no IT implementation, whether cloud or on-premises, can be undertaken without some degree of risk.

Cloud providers can help mitigate any concerns by clearly outlining potential risks and the procedures in place to deal with them. Key areas for discussion should include

• High-level storage, compute and networking architectures
• Datacenter specifications
• Service Level Agreements (SLAs) for outages, and recovery plans
• Disaster mitigation processes

What Enterprises Need To Do

Enterprise cloud consumers must also play an active role in building a cooperative partnership with cloud providers. The importance of clearly communicating the planned use of the cloud services they’re buying cannot be overstated. Sharing a realistic and accurate plan allows cloud vendors to provide the best possible performance and user experience - as well as mold their offerings to meet the enterprise’s present and future needs.

There is no doubt that IT will continue its march toward cloud computing – the benefits are just too compelling to ignore. But enterprises moving to the cloud have to understand that they are not simply purchasing a specific product or service. Rather, they are entering into a core partnership with their cloud provider.

With up-front communication and transparency on both sides, cloud providers and enterprise IT can adjust to this new paradigm together as they work to design and deploy better solutions to increasingly complex technology problems.

Image courtesy of Shutterstock.

Guest author Daniel Lopez is co-founder and CTO of BitNami.

It is difficult to avoid a weird feeling of Deja-vu when looking at the current cloud-computing landscape. Microsoft is once again battling for the future of the technology industry.

For years, Microsoft dominated the IT landscape with its Windows operating system, providing an industry-standard platform that others built on top of. Regardless of any pricing issues or technical shortcomings, the vast ecosystem of Windows applications and service providers ensured the continued success of the platform for many years and was an insurmountable barrier for competitors. It was not until the Web came along that this dominance was seriously challenged. The book High-Stakes, No Prisoners chronicles the story of the Frontpage acquisition and does a good job of providing a peek into the ruthless ‘battle for the Web’ against Netscape.

Microsoft Is A Cloud Computing Underdog

Microsoft is now waging another platform war: the battle for the cloud. The difference is that this time, Microsoft is the underdog.

Amazon has built not only an automated way to spin up new servers and databases, but an entire platform for building and running a whole new generation of applications. Where in the past you had to write apps using Win32 APIs and third-party OCX controls, you can now write applications using Amazon’s cloud APIs for file storage, database access, message queues and dozens of other services. The launch of the AWS marketplace further solidified Amazon’s move up the stack. If Amazon acquires a critical mass of users and vendors to build on top of its platform, the network effect will make it very difficult to displace that ecosystem.

Microsoft has not been sitting idle. The original version of Windows Azure was architected around a Platform-as-a-Service (PaaS) offering and was very Windows-specific. It had many shortcomings and attracted little developer and partner support.

Making Windows Azure More Competitive

However, in 2013 Microsoft has refreshed its Azure offering, providing a Virtual-Machine-centric offering modeled after Amazon’s Elastic Compute Cloud (EC2). The company went out of its way to make sure Linux and open source were first-class citizens. Microsoft has even demoed Azure using Apple MacBook Pro laptops and launching Ubuntu images. Microsoft finally “got it” - the launch of Azure Virtual Images was the first step towards fighting AWS head on.

About a month, Microsoft unveiled the Windows Azure Community portal, which provides dozens of popular open source applications and language runtimes contributed by partners. Even more recently, Microsoft took this a step further and made the images from the portal available directly in the Azure console, so they can be easily deployed onto Azure infrastructure. By making it easier to deploy third-party apps on its cloud, Microsoft is helping to grow its own ecosystem while increasing the utilization of its infrastructure. It also provides a counterpart to the AWS marketplace that, while limited, it is in many aspects simpler and easier to use.

Not Better, But Maybe Good Enough?

Microsoft still offers only a fraction of the functionality of Amazon, but it has a much bigger established user base among small and medium businesses and the enterprise. Coupled with its willingness to aggressively compete on price, Microsoft does not necessarily need to be better than Amazon to win. It just needs to be “good enough” to prevent its own users from switching.

It is incredibly refreshing to finally see viable competition to Amazon in the public cloud arena. Together with Google Compute Engine, Microsoft should be able to give Amazon a good run for its money.

Who will be the big winners of this war? For one, end users, who will benefit from lower prices from increased competition, as the cloud giants fight for market share.

Image courtesy of Shutterstock.

Guest author Mike Pav is engineering vice president of Spanning Cloud Apps, a provider of data protection solutions for the cloud.

When Amazon Web Services crashed on Christmas Eve (which brought down Netflix among other high-profile sites), Amazon offered this explanation: its elastic load balancers failed. Load balancers, as the name implies, distribute the network's workload. Among their most important functions is protecting the system's components from becoming overburdened and shutting down. 

After Amazon's outage, the Web became a virtual fount of suggestions for avoiding more such glitches. Some said Amazon's cloud customers should write their own load balancers. Others said service providers like Netflix should deploy multiple data centers as insurance against another PaaS failure.

(See also: Why Netflix's Christmas Eve Crash Was Its Own Fault)

Failure Is An Option

A month later, it seems clear to me: Cloud outages, while rare, will continue to be a fact of life.

Here's why: Perfection is simply too expensive. To achieve uptime of more than 99.99% requires an investment of  money, machine and human resources that - given the rarity of failures - just isn't worth it. The extra cost inevitably would be passed along to customers, all but negating the cloud's cost advantage. 

Instead, customers should expect PaaS providers to provide them with a well-reasoned plan for handling any disruptions.

PaaS providers should be the first to know when an outage has occurred:

• They should be able to estimate when service will be restored.
• They should know and be willing to report who was impacted, and whether data was irrevocably lost.
• After an outage has been reported and until service is restored, PaaS providers should supply customers with regular status updates.
• Once service has been restored, they should offer a detailed post-mortem as well as a plan for avoiding future interruptions.

Here's where it gets tricky: PaaS providers are understandably reluctant to offer gory details for fear that they will lose current or prospective customers. If the PaaS company in question is publicly traded, those fears will be compounded by the worry that its stock price will tumble.

The real reason to sign onto a PaaS has nothing to do with whether it claims to offer 100% uptime. You choose a PaaS provider because it offers scalability and elasticity, and the same efficiency and user experience regardless of the level of system usage. Applications can be built and delivered on a PaaS an order of magnitude faster when compared with non-cloud-based systems.

Using a PaaS not only reduces a customer's total cost of ownership - they operate on a pay-per-use model - it allows them to delegate tedious and time-consuming IT chores like system monitoring and maintenance. With that stuff out of the way, PaaS customers can focus their resources on truly adding value for their constituencies.

Even after the well-publicized outages, the reason so many high-profile companies - including Netflix - still use Amazon as their PaaS provider is because it does a great job of providing ready-to-use features. AWS isn't 100% reliable, but it can be used with very little up-front investment and scaled as needed. And that is an enormous improvement over the Information technology systems of the past.

Image courtesy of Shutterstock.

It took Amazon Web Services (AWS) eight years to hit $650 million in revenue, according to Citigroup in 2010.  Just three years later, Macquarie Capital analyst Ben Schachter estimates that AWS will top $3.8 billion in 2013 revenue, up from $2.1 billion in 2012 (estimated), valuing the AWS business at $19 billion.  It's a lot of money, and it underlines Amazon's increasingly dominant role in cloud computing, and the rising risks associated with enterprises putting all their eggs in the AWS basket.

The Enterprise Is Cloud Computing's Next Big Target

Ultimately, it's the enterprise that will crown a winner in cloud computing. And as Macquarie Capital posits, Amazon is gearing up to take on the enterprise:

We highlight that much of AWS‟s past growth has come from SMBs [small and medium-sized businesses] and start-ups. While these will continue to drive growth globally, in 2013 we expect a more concerted effort to focus on growing the AWS business at large enterprises. Importantly, the broadening platform offered by AWS is helping to drive adoption in larger enterprises (moving up the tech stack, with RedShift and other recent product launches). Additionally, we think emerging markets offer a significant opportunity, as many enterprises in these regions will likely bypass traditional technology infrastructure and go straight to the cloud.

Not that winning the enterprise will come easily. Outside the SMB market, Amazon has real challenges.  

One is reliability, which has Amazon scrambling to root out "unknown unknowns" before they bring down AWS. In this the company largely has to go it alone. Even if Amazon were to open source the infrastructure it uses to run its operations, as Facebook has, it's unlikely that the company would derive much benefit. Amazon runs at a scale that few developers have experienced, rendering its "many eyeballs" just that: eyeballs that don't necessarily know what they're looking at, and impotent to dig in and help make AWS more resilient. Sure, there are engineers at Google and other Web giants who can appreciate Amazon's scale (450,000 Linux servers and counting), but what's their incentive to help by contributing code or insight?

It's unlikely that the enterprise is going to toss out AWS any time soon, however. It's just too convenient, and relatively cheap. Amazon is happy to run at ridiculously low margins, according to its CEO Jeff Bezos in a Harvard Business Review interview, because the company believes the potential of the Internet is still largely untapped. In Bezos' mind, we're in the midst of a land grab for the future of retail and computing. Building and owning a market is more important than collecting profits from it. At least for now.  

Amazon Is A Tough Competitor

This translates into serious customer benefit. As for competitors, well, Businessweek's Brad Stone calls the company's approach "the Bermuda triangle of business."  

Amazon also has the benefit of having been first to market, with its interface (API) now dominant among developers. In responding to a podcast question about why his company chose to embrace the Amazon API instead of creating its own, Basho CTO Justin Sheehy highlights the developer inertia cutting against API innovation:

I've never heard a single customer say, "Don't give me the Amazon API."... If everyone were starting to build these things at the same time then innovating at the API might be interesting. When the majority of users are already using the same one, however, you don't pick it because it's the best. You pick it because it's the API that gives you the least friction to people adopting your product.

None of which is to suggest that AWS has already won the battle for the enterprise. There are plenty of excellent alternatives, particularly for enterprise workloads, in Google Cloud Platform, VMware CloudFoundry, Microsoft Azure, Rackspace and others, not to mention private/hybrid cloud solutions from Citrix (Apache Cloudstack), Eucalyptus, OpenNebula and more. Many of these have technical and other advantages, as well as price discounts in some cases. But right now it is Amazon that is setting the terms, and with its aggressive pricing and product roll-outs, it's making those terms as onerous as possible for competitors to surpass.

Ultimately, it's a race to see whether competitors can improve their ease-of-use and cost structure faster than Amazon can improve its reliability and security. Either way, customers win.

Image courtesy of Shutterstock.

After an ill-timed outage on Christmas Eve zapped popular video provider Netflix - the popular refrain has been clear: Blame the cloud. But when there's a car crash, do we blame the highway or the humans driving the vehicles? Is Netflix really the victim here, or did it drive off the road all on its own?

Reports of Netflix crashing again on Christmas Eve day started trickling in about an hour after my youngest daughter, stuck inside on a bitter cold Minnesota day, complained that the service wasn't working on my iPad. That problem was alleviated by slapping a password on the device and sending her into the kitchen to help the rest of the family prep for dinner like she should have been doing in the first place. But the inconvenient timing of the outage was enough to cause a bump of coverage on the national news.

As the postmortems came though, it appeared that - once again - Netflix's problem lay within the cloud on which the service is hosted: Amazon Web Service's Ashburn, VA, data center.

Virginia? Again?

Neither AWS or Netflix have released a detailed report on what actually happened, but reports indicated that it was the elastic load balancers at the Virginia data center that somehow dropped the ball and led to significant traffic loss for Netflix viewers trying to watch their favorite Christmas movies. The service was back up by Christmas Day, but dropping the ball on Christmas Eve didn't make Netflix many friends.

Meanwhile, as many people noted during the Netflix outage, Amazon's own Instant Video service had no reports of problems. That raised a few eyebrows for customers wondering how Amazon managed to keep its own service going while its competitor was kaput.

No one is accusing Amazon's business units of collaborating to bring down Netflix. But the very fact that Netflix relies on a competitor's infrastructure to deliver its services seems to generate a conflict of interest.

A lot of those same industry observers are also calling for Netflix to get the hell off of Amazon's cloud. This is not the first time, after all, that AWS problems have smacked around Netflix and other popular Web services, and that Virginia data center specifically seems to be cursed.

I think a service like Netflix (of which I am obviously a customer) should keep its destiny in its own hands. But if you think that moving to it's own cloud will be the sure-fire cure-all for Netflix' reliability issues, think again.

The fault for the Netflix outage, the company would like us to believe, lies solely with AWS. But does it really?

Or does the problem lie with misuse of AWS tools? If the elastic load balancers were indeed the reason for the Christmas Eve outage, who was ultimately responsible for configuring those balancers?

Winning The Blame Game?

The highway analogy applies here, too. AWS is the highway, a shimmering ribbon of concrete, on-ramps and bridges that enable cars to get from point A to point B. Most of the time, the highway's operations run smoothly. But when someone misuses the highway, chaos will most certainly ensue - no matter how good the infrastructure is. 

If you don't like the highway example, pick another brand of infrastructure, like a building or a ship or a bridge. It's all the same: Use the infrastructure the wrong way, and bad things happen.

Netflix would (and can) argue that sometimes, no matter how well you're operating within the infrastructure, that infrastructure can break. That's true. Tragically, things fall apart and people and businesses can get caught in the wreckage. Such is life in an entropic universe.

But even if AWS has a faulty infrastructure, doesn't Netflix still have ultimate responsibility to create the solution? After all, customers are "renting" their movies from Netflix, not Amazon. And as pointed out, this is not the first time there's been problems at this particular data center. Why, after getting slapped off the Internet this summer, didn't Netflix make sure such an occurrence would happen again?

Netflix shares were down slightly on Thursday (about 1% as trading drew to a close). Maybe some shareholders are asking themselves why Netflix hasn't done more to shore up te reliability of its service. I know this customer is.

Image courtesy of Shutterstock.

Now that Oracle plans to gobble up yet another company for its growing portfolio of cloud-based software, the question is whether the company can chew everything it puts in its mouth.

The business applications maker said Thursday that it would buy Eloqua for $871 million. Eloqua's apps measure the effectiveness of marketing and sales projects. Oracle plans to make the company's software the centerpiece of its "Marketing Cloud" product line.

The acquisition will fill out the cloud-based services offering Oracle hopes will grab business from rival Salesforce.com. Earlier this year, Oracle closed a $1.9 billion deal for Taleo, a maker of online human-resource management software, and a $1.5 billion purchase of RightNow Technologies, which makes software for managing customer service.

Oracle's Challenge

Oracle suddenly has a lot of cloud software to make work well together if it wants to give customers a reason to lease multiple applications. The task won't be easy, given that the multi-vendor software was never designed "to fit together like Legos," said Andrew Frank, analyst for Gartner, said.

Taleo is a customer of Eloqua, so that portion of the integration challenge may be easier. But the extent of Oracle's success will be determined when its services go up against products from Salesforce.com, IBM and SAP. "Whether it fits together well is a question for the market to determine," Frank said.

On paper, the acquisition gives Oracle an apples-to-apples comparison with Salesforce.com, which provides integrated marketing, sales and service management applications. Oracle needs the same all-in-one portfolio, because its rival "continues to have a strong win rate against Oracle in CRM (customer relationship management)," said Richard Sherlund, a managing director of equity research for investment bank Nomura, in The Wall Street Journal.

Oracle is certainly willing to pay for what it wants. The company offered Eloqua shareholders a 31% premium over the Wednesday closing price. The transaction is expected to close in the first half of 2013.

Based in Vienna, Va., Eloqua was founded in 2000 and went public in August. Customers include Sony, Johnson & Johnson, VMware and Siemens.

Cloud-Based Marketing Apps Are Hot

Marketing applications are favorites of vendors building out cloud-based services, because the data-intensive software is a good fit for cloud deployments, Frank said. Data centers that run cloud services are built to handle lots of data and to provide the computational power needed to analyze that data.

In addition, marketing software needs to be integrated with other cloud services, such as social media. An added value is the fact that the applications are accessible to the mobile devices now expanding in the enterprise.

As a result, researchers like Gartner see an increase in spending on cloud-based marketing services, and Oracle doesn't plan to be left behind in the race to grab those dollars.

Image courtesy of Shutterstock.

It's a dirty secret that ad providers want you to ignore: The vast majority of website failures are caused by online-advertising services. They're choking the very sites from which they're trying to generate revenue.

This is a symptom of the distributed Web with which we work. Sites today depend on a sheet of services - like social media, security and advertising - to bring the full Web experience to their users.

And the problem is growing even more complex with cloud-based services. Naturally, sites can be impacted by data-center outages, but problems with the services themselves can also slow or bring down your own site. And it doesn't take a natural disaster to create the first tear that rips apart other connections, either. Sometimes it's just one service getting hammered that starts a chain reaction that knocks your site off the Web.

Catch-22 In Action

Compuware, which monitors cloud and site performance, says the No. 1 culprit is ad services. 

The graph below highlights errors found in Compuware-monitored sites over 24 hours in early October. As Stephen Pierzchala, APM Technology Strategist, explained, the failures signify any event that returned an HTTP error or reflected a serious timeout period, caused by either an application failure or slow down, or connectivity problem.

Ad services were the prime cause of failures, according to the survey, followed by Web-analytic and Web-page component services. What's telling about these errors is that both categories are ubiquitous for any site trying to generate revenue through ads. Analytics can be even more pervasive, because even admins not dealing with ads need to understand their traffic.

For something that seems to comprise such a large portion of third-party failures, there is little anecdotal evidence of ad-service problems reported on the Internet. Some quick searches reveal little first-page information on various search engines. This would seem to suggest that either the problem is not as big an issue as the Compuware data would suggest, or there may be some active reputation management going on.

Reasons For Problem Vary

The causes of the third-party errors can vary: a failed server, lost connectivity, or even a bad script that wreaks havoc on the client's initial load of the site. Front-end optimization vendor Strangeloop surveyed third-party scripts that affect the top 200 Internet retail sites and found an average of 6.7 third-party scripts on each. That's 6.7 ways a site can be slowed to a crawl or stopped altogether.

Strangeloop's report includes some telling video demonstrations about how some sites do better than others with third-party scripting.

While completely avoiding ad services isn't an option for most websites, site administrators should use this data to more closely examine the terms of the service-level agreements they sign with each of their service providers. If you rely heavily on ad services and analytics, make sure your site is designed to work around slowsdowns and breakdowns - and that the vendor in question has its own work-arounds.

Site delivery is increasingly about supply chain management, where the supply is content. Website owners need to make sure their chains are as strong as possible.

Title image courtesy of Shutterstock.

While cloud computing giants Amazon, Google and Microsoft scramble to cut prices to lure customers to their cloud-computing infrastructures, their smaller rival Rackspace Hosting is heading in the opposite direction.

"We're not going the route of a race to the bottom," says John Engates, Rackspace's chief technology officer. Instead, Rackspace is betting that corporate customers will pay more to make the San Antonio-based company their IT department in the cloud - a strategy that's sure to face challenges as the number of competitors rises.

Rackspace's Success

Over the last several years, Rackspace has competed mostly with Amazon in the business of renting data center servers for pennies an hour to run websites and business applications. Since its initial public offering in 2008, Rackspace has grown its cloud infrastructure business to nearly 24% of revenue. The company has more than 180,000 business customers and topped $1 billion in revenue last year for the first time.

But the competitive landscape is growing more crowded. Besides Google and Microsoft, heavyweights Dell, Hewlett-Packard and IBM have also checked in for battle. Rackspace is banking that analysts are right when they say price alone isn't going to win business. Besides pitching better services at a higher price, Rackspace is touting OpenStack, an open source, cloud-computing platform it launched with NASA in 2010. Roughly 150 companies have joined the initiative, including Intel, Dell, HP, IBM, Cisco and AT&T. In fact, on Tuesday, HP released HP Cloud Compute, making it the first non-Rackspace vendor to build an offering with OpenStack (see HP Switches On Public Cloud, Thanks To OpenStack).

Rackspace and OpenStack supporters are hoping businesses will buy into their claims that the platform provides more flexibility than the various proprietary systems offered by Amazon and others. Theoretically, whatever you are running on one OpenStack implementation can be moved to any other OpenStack implementation, making it easier for customers to switch vendors.

Fighting Vendor Lock-In

"The challenge with proprietary (cloud technology) is that customers feel like they're going down a path that's sort of a one way street," Engates tells ReadWrite. "They're locking themselves in and there's no way out. If they choose to go somewhere else, they have to re-architect, rebuild or retool, and that's a challenge."

Over the last several weeks, Rackspace has rolled out the complete suite of OpenStack-based products, including servers, databases, infrastructure monitoring, backup, storage and networking. "We're in full production with all these services," Engates says.

Is OpenStack Immature?

OpenStack has its critics. Research firm Gartner says Rackspace and other supporters are not really interested in building an open cloud platform. Rather, they have pooled resources in order to battle Amazon's dominance in providing cloud-based Infrastructure-as-a-Service (IaaS). They also do not want to pay license fees for commercial cloud technology, such as VMware's vCloud stack.

Rather than being ready for prime time, OpenStack is an "early-stage project whose future, though promising, is still uncertain," Gartner says.

"Some people have been led to believe that because OpenStack is open source, it is an open and widely adopted standard, with broad interoperability and freedom from commercial interests," Gartner analyst Lydia Leong wrote in a recent analysis. "In reality, OpenStack is dominated by commercial interests, as it is a business strategy for the vendors involved, not the effort of a community of altruistic individual contributors."

Big Data

Nevertheless, Rackspace is plowing ahead with plans to use OpenStack early next year in helping companies manage the huge amount of data created each day. Rackspace has partnered with Hortonworks, which is using the Apache Hadoop data platform to power products for storing, managing, processing and analyzing large amounts of data. "Big data is an area that we think is still in the early days, but there's a lot of interest and a lot of excitement around it," Engates says.

Amazon also sees money in big data. Last week, the company introduced its first big data product at the first Amazon Web Services (AWS) re:Invent customer conference. Amazon is calling big data platform Redshift Data Warehouse as a Service, and business intelligence vendor Jaspersoft has announced support for the technology.

Like Amazon, Rackspace is taking its cloud services outside the U.S. Roughly a quarter of its revenue comes from overseas, mostly from Europe. It has data centers in London and Hong Kong and plans to open one in Australia "very soon," says Engates, who didn't know the exact date. The company is also considering a data center in Latin America.

With expansion overseas, a major platform redesign and new products on the way, Rackspace will need higher prices to fund its ambitions. The question is whether potential customers will still find them a good deal in a highly competitive market

Lead image courtesy of Shutterstock.

The race to commoditize the OpenStack cloud operating system has a new entry, and Hewlett-Packard, one of the first backers of the open-source project, could soon be enjoying the spoils of victory. If it doesn't screw it up.

There is little evidence that yesterday's announced general release of HP Cloud Compute, which is based on OpenStack, is going to have problems, but it's a little hard to have confidence in HP in general given the stumbles the company has had lately. But this is one of the best signs in a while that the company is on a stable path.

Here's why: despite the hype surrounding OpenStack cloud computing, it is important to remember that there is no real commercial product available from OpenStack yet. OpenStack is an open-source software project, and even though everyone and their brother (like Red Hat, SUSE, Canonical and HP) are contributing to OpenStack, there has been, until now, only one other commercial implementation of OpenStack -- Rackspace's Open Cloud service.

Rackspace, though, had a distinct advantage over HP: it pretty much invented OpenStack and therefore had a big headstart. But that's all changed now.

HP's Cloud Compute software is based on just one part of the broader OpenStack ecosystem, the aptly-named OpenStack Compute, formally named Nova. Compute is the primary element of any infrastructure-as-a-service (IaaS) model, and is the part of cloud computing that competes directly with Amazon Web Service's Elastic Cloud Computing (EC2) platform.

HP Cloud Compute is a public cloud IaaS offering, much like AWS' EC2 and Rackspace's Open Cloud, and will compete for the same clientele. It's appropriate that HP is the first outside vendor to launch the commercial HP Cloud Compute platform -- after all, it was one of the earliest non-Rackspace vendors to jump on board the OpenStack bandwagon after Rackspace announced it would be opening the code.

An argument could be made that the OpenStack race is about over and it really is a three-way battle between CloudStack, AWS/Eucalyptus and OpenStack now that HP has pushed out this general release. But there are some key issues to consider.

First, as Sony learned with Betamax and Atari and Palm learned in general, being first to market isn't always the path to victory. How HP provides service and how it prices this new service will be a big factor in its success.

Right out of the gate, the prices look pretty good, starting at four cents per hour. More tellingly, the service level agreement promises 99.95% uptime. This is something AWS can do, too, but you have to configure your machines a certain way in order to qualify for this level of service.

Second, there are other players in this game that are coming up fast. Rackspace is already pushing OpenStack-based products, and Linux vendors Red Hat and SUSE are heads-down running toward the finish line. As operating-system vendors, could they have a better advantage than hardware-oriented HP in offering cloud services?

Not to mention that there's still a need to deploy a private-cloud option based on OpenStack. If you want to run OpenStack software inside your own firewall now, you have to spend days getting the systems configured, because there's no unified package to be downloaded and installed yet. Without that option, HP is just another AWS competitor in an increasingly crowded public-cloud sector.

Finally, and not to put too fine a point on this, there's still the very real consideration that this is HP we're talking about.

Even putting aside recent missteps, it's going to take some work to convince customers that it can be a public-cloud company. If it works, it will be very good, because then they'll be able to sell hybrid- and private-cloud services a whole lot easier. But that transition of perception could still run aground with more fumbles by HP.

Still, HP is now in the game with a public-cloud offering, made possible by OpenStack. Even if they do nothing else, HP has just given AWS one more thing to worry about.

It's a great time to jump into cloud computing. If your business runs on the Web, you can reap the benefits of a good old fashioned price war among Google, Microsoft and Amazon. The trio is lowering prices for their Infrastructure-as-a-Service (IaaS) cloud-computing platforms in order to attract developers. For startups and other companies not tied to legacy platforms, that means the time is right to get a good deal on building and running websites and client-side applications on any one of three state-of-the-art data centers.

A New Round Of Price Cuts

Less than six months after launching Compute Engine, Google dropped prices this week by about 5% for its baseline services, even though the cloud service is still in preview mode and available only to select customers. Amazon and Microsoft are expected to bring their prices in line with Google's for their Elastic Compute Cloud (EC2) and Azure services, respectively.

That's because the rivals have cut prices in unison before. In March, Amazon lowered pricing first, and Microsoft and Google followed suit within days. Google had Cloud Storage at the time and unveiled Compute Engine in June at its Google I/O developer conference.

"All other factors equal, price isn't going to get you to success," warned Melanie Posey, analyst for IDC. "But having pricing out of line with competitors is certainly going to set you back in the market."

Is Anyone Making Money?

No one on the outside knows whether Microsoft, Google or Amazon is making money on their increasingly important side businesses of renting computing power via the cloud. None release the financial details needed to determine whether their cloud platforms are profitable.

Because these are not any of the companies' core businesses, there's little pressure to make money right away on cloud services. "It may very well be the case that these cloud platform divisions don’t have any set mandate to make a profit," Posey said. Therefore, pricing could go down much lower, if the companies need to prime the customer pump even further.

Amazon has the first-mover advantage. The online retailer opened its data center to developers in 2002, making Microsoft and Google the challengers. Amazon has done well in building a customer base of startups and established businesses. Among the biggest cloud-based sites it supports is Tumblr.

But the Infrastructure-as-a-Service market remains young, and Google and Microsoft have plenty of cash to throw at their fledgling platforms. Also, because the goal isn't yet about stealing Amazon customers, but more about attracting the growing number of companies building Software-as-a-Service (SaaS) businesses in the cloud.

In a recent survey of 556 organizations across 10 countries in North and South America, Europe and Asia/Pacific, Gartner found 71% have been using SaaS for less than three years and 77% are expecting to increase spending on the application-delivery model.

Cost Is Only One Factor

While low prices certainly help get potential customers' attention, it's not enough to seal the deal.

"I would say that everyone continues to have room to cut prices," said Gartner analyst Lydia Leong. "However, customers generally say that agility, not cost, is their primary driver. There are use cases where the cheapest possible compute is valuable, but we generally see customers buying based on capabilities, not price."

That's true even for cash-strapped startups. "Startups are far more interested in time to market and capabilities than cost, since the former two things translate to competitive advantage," Leong said.

So with every round of price cuts, the vendors will also be adjusting the amount of system memory, processing power and other features they offer, depending on market demand.

No matter which company edges ahead in features, any advantage will likely be short lived. Amazon, Google and Microsoft are all fierce competitors and they are unlikely to let a rival enjoy an advantage for very long. That mean that the real winners of the cloud computing price war will be startups and other potential cloud customers.

Image courtesy of Shutterstock.

In the age of cloud computing, it's an archaic thought: The livelihood of some popular websites currently rests on bucket brigades carrying diesel fuel up multiple flights of stairs just to keep generators running. But that's the reality as Lower Manhattan struggles with massive power outages and flooded understructures in the wake of Hurricane Sandy.

It's a decidedly 17th-century solution to a 21st-century problem. But it's the only way to get fuel to generators that are located in powerless high-rises when elevators are out of service and basement fuel pumps are incapacitated due to flooding.

For site admins at Fog Creek Software in New York, the solution is clear, albeit arduous: When diesel is delivered to their building at 75 Broad Street, the fuel is hand-delivered to the generators on the 17th floor via bucket brigade, giving the online service provider more hours of uptime while basement flooding is cleared and the building awaits power restoration from ConEd. Meanwhile, users of services like Trello, FogBugz and Copilot are basically in a holding pattern, knowing that one missed fuel delivery or downed generator is all that stands between uptime and downtime right now.

[Disclosure: ReadWrite is a Trello user.]

A Too-Common Tale Of Woe

Fog Creek is not alone in its situation. ISPs Peer1 and Internap both maintain facilities in the same building. Fog Creek and fellow ISP tenant Squarespace are working with Peer1 on the bucket brigade to keep servers going. Internap is reporting that since their fuel pumps were swamped, the facility was shut down after the generators' fuel ran out.

The story is being repeated across the New York Metro area, as data centers switch to back-up power on diesel generators as expected. But what no one counted on was the massive flooding that would compromise fuel systems and generators across the area hit by Sandy.

ISP Datagram was shoved off the grid Monday night when its 33 Whitehall facility had its basement flooded. This shut down high-volume sites like Gawker, Gizmodo and Buzzfeed. Gawker and Gizmodo are still running "emergency" pared-down sites, but Buzzfeed managed to get itself back up on its feet as early as Tuesday.

Clouds Can Help, Too

How was this done? Probably by the one method that every single Web customer in Lower Manhattan should be thinking about just as soon as the lights come on: BuzzFeed got its website on the cloud.

"In a nut, our engineering team moved everything over to Amazon Web Services, with one developer working overnight despite a tree falling through his roof," wrote BuzzFeed's Matt Buchanan.

Granted, it's not easy to just up and replicate your site to a cloud-based system. BuzzFeed's team was aided by the fact that much of the site's content was already replicated on Amazon, so they were halfway there.

"You might wonder why, of course, BuzzFeed, Gawker, and others aren't already all aboard the cloud train, ready to switch to different servers at the drop of a hat," Buchanan concluded. "The fact is, Amazon cloud service and other services like it weren't around when BuzzFeed and sites like Gawker and Huffington Post were architected years ago. If the site was built today, the architecture might look a bit more cloud-like than having a huge data center based in downtown Manhattan."

Hindsight has the benefit of being 20/20, it's true. But after the waters from Sandy recede, look for a lot more websites to gain a renewed focus on the cloud.

Image courtesy of Shutterstock.

Although Sandy continues to be primarily a flooding event, High, sustained winds and some large fires are complicating life on and off the Internet.

Service issues for YouTube were widely reported Monday at 7:30 p.m. EDT, and were intermittent throughout the rest of the night, until they ended around 6 a.m. today. Tumblr's problems started at about the same time, with intermittent outages through the night.

Tumblr is based in Manhattan, bit it is likely that the social network's systems are co-located elsewhere. Other New York-based firms, like BuzzFeed, Huffington Post and Gawker likely have suffered Sandy's wrath. BuzzFeed was down last night through 8:30 a.m. today, HuffPost is in station-keeping mode due to a power outage at its offices, and Gawker is down.

The infrastructure of the Internet is solid so far, although edge-of-network problems continue to plague the East Coast, sending ripples  across the continent. According to the Internet Traffic Report (ITR), North American traffic was rated 67 out of 100 at 8:45 a.m. EDT today, with several Internet routers across the country reporting 100% packet loss. While many of the affected routers are on the Eastern Seaboard, several are not, which could mean there are systemic issues other than Sandy affecting traffic.

According to the ITR, traffic speed on the Internet dropped sharply from a rating of 69 just after 4:50 p.m. EDT Monday, just as winds from Sandy grew dangerous on the coast. Because of the size of the storm, high winds range as far west as Chicago, with scattered power and telecommunication outages reported throughout the immense affected area.

Sometime around 3:30 a.m., packet loss spiked briefly, indicating a significant disruption of Internet service, the cause of which was not clear.

Data centers in the affected region seemed to fare well.

Outage Analyzer indicated a few scattered outages, including a suspected outage at pro-market.net that affected 174 domains, and an earlier outage at ad provider ContextWeb. KDDI's Northeast U.S. servers also went down for two hours at 9:12 p.m., potentially impacting 675 domains. All of these servers were located in northern New Jersey, and as of post time, the problems were resolved.

Downrightnow is reporting some issues with Twitter, Facebook, Gmail, Yahoo! Mail and Tumblr. While downrightnow is a useful tool, it gathers some of its data from anecdotal complaints on social-media services like Twitter. Given the scattered router problems on the East Coast, it is likely that customers' access to those services are only degraded.

Those who rely on Amazon Web Services will be pleased to note that their services are smoothly running in North America, with no problems reported. Google Apps and Google's App engine cloud service are also showing green lights across the board.

As Sandy moves north, dumping rain and pounding with winds, there is still a chance that some data centers will be affected by the storm. Certainly local and regional Internet service on the edge of the network will be affected, either through power or direct disruptions. For now, major service disruptions seem to have been avoided.

[Updated]

Equinix, one of the maintainers of a New York City-area Internet exchange, has released details on how it is weathering the storm. According to Sam Kapoor, Chief Global Operations Officr, "All sites in New York/New Jersey, and several in Philadelphia and Washington, D.C., experienced power outages and customer loads were transferred to generator power.  All of these sites have at least 48 hours of fuel onsite with fuel vendors on standby to deliver more as needed. As of this morning, some of these sites are back on utility power, while some remain on generators.

"NY9 experienced a failed generator that impacted service to several customers. We made repairs and service was returned this morning. The site remains on generator power," Kapoor added. "Sites in New York and Washington, D.C. experienced water leaks. While most were minor and quickly contained, at least one leak at a New York site impacted a customer. We are currently working onsite with the customer to contain the issue."

Equinix clearly isn't taking their responsibilities lightly, either.

"To minimize the storm’s impact and prepare wherever possible, we tested generators, filled fuel tanks and made arrangements for backup fuel, arranged to have full time staff onsite at each facility to maintain operations, secured hotel rooms near the sites, and brought  in cots, MREs, and other emergency supplies, so that our Operations staff could comfortably support customers even in extreme conditions," Kapoor explained.

Image Courtesy of NASA