Grid Meter » December 2006

December 22, 2006 | Comments: (0)

Grid Apps - Ever Hopeful

TAGS: None

Earlier in the year there was some traffic in the grid press regarding who was more advanced in their Grid use, the US or Europe. I forget the conclusion, although it might have been decided that in fact Asia was more advanced...

But in this week of the focus being on the consumer, where are we in getting Grids to the masses, well at least to the Enterprise masses.

This week I saw the following press release regarding the European Network of Excellence CoreGRID, and the 3rd "GRIDS@WORK" event and Grid Plugtests. From the release:

Over 200 delegates, representing 25 countries from all over the world, came to Sophia Antipolis, France, to discuss research projects and industrial requirements towards the implementation of Grids as a key driver of tomorrow’s 'Network-of-networks'.

I think I'll lump 'Network-of-networks' in with terms like 'cyber infrastructure'. A lot of talk about what the roads will look like, but no talk about the cars (applications) that will drive on them.

And like many before them, they have turned to a contest to figure out what exactly this 'Network-of-networks' is to be used for. See: Grid Plugtests - A Contest to Demonstrate the Power of the Grid.

Now earlier this year Amazon.com thought they had the killer Grid application with their Simple Storage Service. In a nutshell, the world's biggest bookstore wanted to become the world's biggest data store. Now they plan to venture out to providing computing power, something Amazon calls EC2. As the article points out they are certainly not the first to do this, and getting it up and running seems to be a bit "open ended". From the article:

In order to use the service, you create a server image (called an Amazon Machine Image, or AMI), based on an Amazon spec. Ultimately, the server image will be able to have whatever operating system, applications, configuration, log-ins and security that you want. At the moment, it only supports the Linux kernel. Amazon also has prebuilt AMIs built that you can use as well, so that you don't have to configure them from scratch.

Sure we've got applications, you just have to load them and configure them yourself.

I'm not sure if I'll get another post in next week before the new year, so happy holidays to all. I don't know about you, but I've asked Santa for more talk about Grid applications next year.

Posted by Greg Nawrocki on December 22, 2006 07:47 AM

December 14, 2006 | Comments: (0)

Turn-key Transactional Web Apps on a Grid

TAGS: None

Earlier this week I lauded the wonderful GRIDtoday special feature, Grid Gets Transactional. I keep going back to it as I think there are some really insightful observations, especially from Massimo Pezzini, vice president and distinguished analyst at Gartner.

I like the encapsulation of the definition of Grid 2.0 as an evolution from the HPC nature of Grid to one of a platform for transactional web applications.

The French pilot / author Antoine de Saint-Exupery said, "As for the future, your task is not to foresee it, but to enable it." And this GRIDtoday article introduces us to several companies that are indeed serving as the enablers for this evolution of Grid.

There is another company out there thinking along the same lines and taking it one step further to a complete hosted turn key solution, 3tera. I've mentioned them before in this blog but recently caught up with them for an interview in the Globus Consortium Journal.

The 3tera approach is called AppLogic, a "grid operating system" designed to run and scale transactional Web applications with a slick graphical drag and drop front end that the application architect can use to define the topology.

From the article:

We think the world is ready for a simpler way to run Web applications without having to think about all the hardware that actually is providing the horsepower to it.

Over the last ten to 15 years we have gone from using large SMP-based systems to using more commodity-based Intel and AMD-based servers. As a result, we've developed what some people have termed server sprawl. Instead of running an application on one server, we now run it on a distributed architecture of 20, 30, 60 or even more servers. And of course, there are some, like Google, that run on tens of thousands of servers.

We need a way to easily control applications that span physical resources, allowing people to stop managing hardware. That's what our grid operating system does. It is not an operating system in the sense of requiring people to write code specifically for it. Rather, it is an operating system in the sense that it abstracts and manages the distributed hardware grid and enables people to focus on their applications without having to make them "grid-aware".

In technical terms, AppLogic is a meta-operating system. It actually uses Linux within it. This allows people to run any Linux software they want on top of AppLogic without rewriting their code.

Posted by Greg Nawrocki on December 14, 2006 08:39 AM

December 11, 2006 | Comments: (0)

Transactional Grid, Beyond HPC

TAGS: None

One of the big criticisms about Grid is that if it wants to become "more than a science project", it is going to have to break out of its traditional "HPC wheelhouse" and become a practically applied technology for other aspects of enterprise computing. Transactional Grid applications may be just what the critics ordered.

In a fantastic GRIDtoday special feature, Grid Gets Transactional, we see that there are companies out there not just thinking this way, but actually deploying this way.

This is truly a glimpse of what the future holds for Grid.

Posted by Greg Nawrocki on December 11, 2006 08:31 AM

December 05, 2006 | Comments: (0)

Of Grid and Men

TAGS: None

I recently tried to contact someone via phone with my Vonage VoIP service. I received a "this number not in service message" for a number I knew was working. I picked up my mobile phone and sure enough the call went through. Back to Vonage and the "not in service message" again.

With the exception of the odd network glitch (which is no fault of Vonage) my service has been rock solid. While speaking with my contact they mentioned that they had Comcast VoIP service. Hmmmmmm....

There are plenty of conspiracy theories regarding Comcast and Vonage and my intent is not to fuel the fire as I have no real proof of what exactly was causing my connection problems. However, it got me thinking back to Grid again.

Regardless of whatever spiffy Grid enabled universal applications we come up with. Regardless of how interoperable and explicit our standards are. If we don't overcome sociological barriers to Grid when they present themselves, the technology means very little.

Perhaps this is the most prominent area where our friends on the academic side of Grid can teach those on the enterprise side of Grid.

Posted by Greg Nawrocki on December 5, 2006 09:35 AM