Have you read about using utility computing and how it offers scalable flexibility over traditional server architecture? IT conversations recently interviewed Scalr’s Sebastian Stadil who demystifies the cloud and talks up his open source cloud resource manager.

 

 ITC.TM-SebastianStadil-2010.06.14 [40:45m]: Play Now | Play in Popup | Download

Lets just get to the meat of it ok. Then more after the jump —

Google is planning to launch an experiment that we hope will make Internet access better and faster for everyone. We plan to test ultra-high speed broadband networks in one or more trial locations across the country. Our networks will deliver Internet speeds more than 100 times faster than what most Americans have access to today over 1 gigabit per second, fiber-to-the-home connections. We’ll offer service at a competitive price to at least 50,000 and potentially up to 500,000 people.

From now until March 26th, we’re asking interested municipalities to provide us with information about their communities through a Request for information (RFI), which we’ll use to determine where to build our network.

That’s from the website.

Now notice this is not some high speed to the head end sort of offer. They specifically say FTH. So they intend to go right to the curb. Their testing will test some 50-500k patrons. What is not clear, is that a single site or a mix of smaller sites.

Google goes on to say they will provide —

* Next generation apps: We want to see what developers and users can do with ultra high-speeds, whether it’s creating new bandwidth-intensive “killer apps” and services, or other uses we can’t yet imagine.

* New deployment techniques: We’ll test new ways to build fiber networks; to help inform, and support deployments elsewhere, we’ll share key lessons learned with the world.

* Openness and choice: We’ll operate an “open access” network, giving users the choice of multiple service providers. And consistent with our past advocacy, we’ll manage our network in an open, non-discriminatory, and transparent way.

What I find of particular interest is the commitment to an open transport layer. A place where anyone can play? ISP and Google? If true that would be a game changer in the data transport marketplace. Fact if true it would complete a vision that was the reason that this blog was created for — create an open backbone and permit service providers to battle it out in the marketplace of products and services.

We keep our fingers crossed.

Linky.

What’s old is new again, with a twist.  Cloud computing shares many similarities with 80’s mainframe computing. When mainframes ruled the IT universe, many owners of these expensive systems would sell otherwise idle time to outside parties to help offset costs. This was easy to implement when the jobs were run as batches and low priority jobs would be run at times of low or no activity.

It’s no secret that most of the servers in data center racks spend most of their lives in idle. So if you’re Amazon, why not sell those idle cycles? Technology now exists to schedule “batches” and to run them as systems are available in real time. The end user may see good performance most of the time, with service degrading only when there are spikes in processing demand.   If this catches on, look for a healthy market in idle cycles from many more providers.

Today the company announced its Spot Instances feature, which is a new pricing option designed to encourage users to buy and consume its idle computing resources.

Prices will change depending on supply and demand, Amazon said, but firms that bid the highest for its services will get access to the platform and be able to use it as and when they need to.

Amazon suggested that Spot Instances would be most suitable for customers that have flexible start and stop times for their rendering, data processing, modelling and analysis processing, adding that by working in this way they would be able to benefit from lower prices, and more importantly, only those prices that they are prepared to pay. (The Inquirer)

Carla Schroder writes in Linux Today —

As much as we warn about privacy, security, and reliability problems in cloud computing, it’s coming and we can’t stop it. So do we join the cloud party? Heck no.

…

Well here we are on the threshold of this very thing, and now the geeks are complaining and warning against it. Why? Because we like to be perverse? Well maybe that is part of it. But for me the biggest problem is trust. I don’t trust many tech vendors because they haven’t given me any reasons to trust them, and plenty of reasons to not trust them. Over and over and over and over and over and over and over.

Why would I entrust them with my data when they do not respect my privacy or the privacy of my data? In the US personal privacy is not protected, and vendors who mangle and lose your personal or business data pay no penalty or recourse, other than bearing the brunt of your peeve. Marketers are all about privacy invasion, as much as they can get away with, and collecting, mining, and buying and selling us. Even worse, service providers roll over at the slightest “boo”, releasing customer records at toothless DMCA takedown requests, and caving in to law enforcement without even making them go through due process. Where are all those attack lawyers when they can do some good for a change?

Carla has a point, and many times I have the same sentiment. But we should not cloud the Cloud as a technology vs that of a business practice using the Cloud. If Skype or Google has an outage it is from this observers point of view little from technology loss as simple human error. Small consequence, but an important one overall.

Any company today has little excuse not to consider internalizing the Cloud. That is bring the Cloud in house. If you have a half dozen servers or more then the company can pull it off. Start with a simple virtualization effort sufficient to free up 2-3 servers. Then consider a Cloud service like Eucalyptus. (There are others as well.) Get those nodes working then port some of those virtualized services over to your own private Cloud.

Trust issues disappear. Or they should, otherwise you have personnel problems larger than your considerations of a Cloud provider. Your results on your level of Nines you require is limited by your pocketbook. Your ability to keep working is also limited to your local networking maintenance if you are in a single location.

But Clouds don’t stop at the company door. Dual core PC’s are generally the norm walking out the door even for individuals. We are almost on the cusp of that happening to the 4 core chips as the price curve continues to drop away on the Intel and AMD product lines. At that point a Cloud in a box is a reality for individuals as well. It won’t be quite as Nines capable as a couple of discrete machines due to single point of failure issues. But a micro Cloud would permit higher levels of service, SOA type backup/restore becomes possible, and harnessing 3-4 cpu’s to a single task is simpler.

So as always, there are means to the madness.

Linky.

Intuit has pretty much created a person and small business software monopoly with it’s Quicken and Quickbooks products.  Startup Mint blasted onto the scene with a free web based service that was easier to use and didn’t nickel and dime users to death like Quicken. While no one is talking about how much market share Mint took from Quicken,it’s obvious that Intuit saw it as a serious threat to its personal finance monoculture.

Intuit has announced that it is purchasing online personal finance rival Mint for $170 million. That is no small change for the free personal finance service named one of The 100 Best Products of 2008 by PC World.

Intuit has long dominated the world of personal finance software with Quicken in all of its flavors. Microsoft Money was its primary competitor, but Microsoft bailed out of the personal finance software market earlier this year. (PC World)

Third Pipe law of the cloud #1: The cloud wants to be open.

Rackspace, the web hosting company that is trying to morph itself into a utlity computing provider has just taken a step in the right direction to make it happen. By opening its API’s, Rackspace will make it’s cloud offering useful to the developer community. It’s kind of hard to imagine being successful the in the utility business without the support and involvement of developers.

Rackspace said today that it will release the APIs for its Cloud Servers product, which provides on-demand, per-instance-based computing. Releasing the APIs means a variety of companies — such as RightScale — will be able to build products and development platforms using Rackspace’s cloud without going through its control panel to manually allocate the necessary servers. This helps bring Cloud Servers, which Rackspace acquired through its purchase of Slicehost last year, into the same league as Amazon’s EC2, which already allows developers to build platforms and management products on Amazon’s cloud through its own APIs. (GigaOM)

I’ve been told that getting customers to pay by subscription for Microsoft’s wares has been a dream of  Steve Ballmer’s since the time he was the company’s lone salesman. It’s a strange turn of events that he’s  getting his wish from a technology he has fought so hard to undermine: utility computing in the cloud.

Microsoft’s Azure of cloud platform doesn’t seem to be positioned well to win many new converts, but it could be enough to keep cloud bound true believers in the temple of Windows. That may be just enough to help Microsoft keeps its strangle hold on the corporate IT world a little longer.

Microsoft will offer Windows Azure under a pay-as-you-go pricing model where customers will pay only for the services they consume. Microsoft plans to charge 12 cents per hour for computing, 15 cents per gigabyte stored, and one cent per 10,000 storage transactions.

The Azure platform includes a Web-based relational database in Microsoft SQL Azure together with connectivity and interoperability with .NET services. Microsoft said it will charge $9.99 for the basic Web edition of its SQL Azure database, which includes up to a one-gigabyte relational database; and $99.99 for the business edition with up to a 10GB database. By contrast, .NET services will be 15 cents per 100,000 message operations, including service-bus messages and access-control tokens, the company said.

And when it comes to network bandwidth, Microsoft said it will charge between 10 and 15 cents per gigabyte for Windows Azure, SQL Azure and .NET services. The software giant also rolled out an enterprise-class guarantee backed by a service-level agreement that covers service uptime, connectivity and data availability for all three services. (Yahoo)

Ted Codd, that is, father of the relational database concept. The concept that made DB2 possible and Oracle a company. The concept is still a great one. The thing is tho, Codd’s idea was two dimensional — rows and columns. You could select what those two spatial entities meant but no more. But our world today is multidimensional in scope. Enter Google –

Under the hood of Fusion Tables is data-spaces technology, which will make conventional databases go the way of the rotary phone, according to Stephen E. Arnold, a technology and financial analyst who is president of Arnold Information Technology.

Data spaces as a concept has been around since the early 1990s, and Google, realizing its potential, has been developing it since it acquired Transformic, a pioneer of the technology, in 2005, Arnold said.

Data-spaces technology seeks to solve the problem of the multiple data types and data formats that reside in organizations, which have to scrub the data and make it uniform, often at great cost and effort, in order to store and analyze it in conventional databases.

Data spaces envisions a system that creates an index that provides access to data in its disparate formats and types, solving what Arnold calls the “Tower of Babel” problem.

In the case of Fusion Tables, the technology should allow Google to add to the conventional two-dimensional database tables a third coordinate with elements like product reviews, blog posts, Twitter messages and the like, as well as a fourth dimension of real-time updates, he said.

So one could build a multiarray database that factors in names, geography, Twitter conversation and time period in a slice and dice fashion with simple system calls. A far cry from what one would have to do in a N+1 relational database to get to the data. Good bet this is what is driving the backend of Google’s Wave.

This has some impacts on corporate futures. As data becomes multidimensional the needs to provide this functionality will increase. For the likes of Google/Amazon who can provide these level of services they expand their futures. For the likes of Oracle/IBM if they do not respond, will be left behind. And there is a future for non-cloud N dimensional databases, for many companies that operate in a ’secure at all costs’ environment. (eg. NYSE, any company with a HIPPA requirement)

One other thing to consider. Google owns the company that came up with idea. So it would not be beyond the realm of possibility that they could be very aggressive on the whack-a-mole in the courts against other entrants. It also indicates that the cloud business has entered a new phase — kill the competition.

Linky.
Fusion Labs.

I know that in evaluating a Stephen Wolfram production, my meager intelligence quotient may not be sufficient to grasp the gravity of what I’m dealing with. So I don’t feel all that put down that I can’t figure out how Alpha is useful to anyone outside of a small audience of college professors and professional engineers.

Alpha is really good at telling you all sorts of information about mathematical expressions, showing you publicly available data about populations and geography, and comparing stock quotes. However, I am a software engineer by trade, and the information I need is about Python module documentation, Apache configuration, and why some f@#$^% snippet of CSS won’t @#$^% render in @#$^% Internet Explorer @#$^% 6. For someone like me (and in the web market, there are a lot of people like me), Alpha is breaking ground in a New Kind of Uselessness.

Maybe I’m being too hard on Wolfram. After all, when industrial recruiters want to hire people, they don’t cruise the local science fairs, but a blue ribbon for your three-panel magic marker and pipe cleaners display on natural language processing and artificial intelligence really sets you apart from all of the other nerds. So, let’s see how Alpha does on a relative scale. I have some experience in building a search-related web product, so I can tell you for sure how most users’ first interactions with a new search engine go.

Which just goes to show that you can’t get love from everyone all the time. Wolfram-Alpha if you think about it is an answer machine in numbers for a world flooded in verbs. We need more of the former but we get mostly the latter. Sad to say.

Is the author right? Hard to tell. But as an attraction getter I think it just blew its whole wad of interest cycle this weekend. What Woflram can generate is interesting. But only an actuary would have a love sequence over the thing.

Linky.

When App Engine was first introduced, almost a year ago to the day, it stuck with Python, a favorite among code-happy Google Oompa Loompas. But after countless request from developers outside the Mountain View Chocolate Factory, the platform has now embraced Java as well.

“We wanted to give developers something that they could be ecstatic about, but we knew we would have to marry the simplicity of Google App Engine with the power and flexibility of the Java platform,” Google engineers Don Schwarz and Toby Reyelts wrote in this evening’s post to the official App Engine Blog. “We also wanted to leverage the App Engine infrastructure - and by extension Google’s infrastructure - as much as possible, without giving up compatibility with existing Java standards and tools.

“And so that’s what we did. App Engine now supports the standards that make Java tooling great.”

According to Cnet, the service’s new incarnation runs version 1.6 of the Java Virtual Machine (JVM). So, in theory, it can also handle code written in such languages as Ruby on Rails and JavaScript.

With that Google opens up a new world of possibilities. Java either partially or totally is a component of so many common tools. That lack of support on GAE has been an issue for many with GAE as a python only platform. One should expect to see a new bevy of applications show up on GAE as a consequence.

There is one other plus out of this. GAE can now give AWS some competition. As Google rounds out the support matrix of other languages we can expect a minor price war to be engaged between the two. Can’t wait.

On a GAE related note, be advised that GAE now has a Cron module available for use. Details here on the Google Labs site.

Linky.