Grid Meter » May 2007

May 21, 2007 | Comments: (0)

The Grid and the Web - Open Standards and Open Source

TAGS: None

Last week I was asked why open source was important to Grid. I'll admit, had to stop and think for a while.

We've all heard the reasons or ensuring interoperability and avoiding vendor lock-in, but these are benefits usually applied to open standards as opposed to open source. So I started thinking about the difference between open standards and open source, which lead me to the following phrase.

Open standards dictate what is possible today, open source is the enabler
for what is possible in the future.

I came to that phrase while drawing analogies between the Web and Grid. "The Web" and "the Grid" are similar in the sense that the terms themselves really mean nothing, they simply represent a technology. The meat behind those terms is taking action and building things with those technologies.

In the case of the Web, all web server software must implement open standards in order to play in the standard protocol environment of the Web, but they don't have to be open source. However, the most popular web server is Apache which happens to indeed be open source. In fact I suspect it is the most widely used collection of open source code in use.

I argue that the Web would not be where it is today if closed source server software was the norm. PHP, database interface plugins and maybe even Ruby on Rails (all those interesting new technologies that enable the collection of things we've taken to calling Web 2.0 possible) would not exist today if folks weren't able to get at the functional guts of the web server and make all those cool things happen behind the scenes via standard http requests.

There is still discussion regarding what Grid is and isn't, and that's not necessarily a bad thing, especially when open source software still makes up the majority of the driving force behind the technology.

Posted by Greg Nawrocki on May 21, 2007 09:21 AM

May 08, 2007 | Comments: (0)

Ground Swell for Grid - Where it May Come From

TAGS: None

CIO Insight recently gave some nice press coverage to a company I've been following for a while now, RevStor.

In a nutshell the RevStor product SANware builds a niftly little data grid seamlessly, with a very easy turn-key install, to provide the benefits of a Storage Area Network. The target market is small to medium size businesses that need a data replication solution for distribution and redundancy.

I've covered RevStor in this blog and in the Globus Consortium Journal and it is nice to see them getting more press play.

It is no secret that the mass uptake of grid computing in enterprise is not where many of us would like it to be, but that may be due to the "boil the ocean" goals that were initially set. Many tried to overlay the model that has worked in research and academic circles of large wide scale grids like the TeraGrid. The fact is that enterprise is too cautious for an "all in" approach and really needs to start smaller and work towards the larger enterprise grids.

This is certainly not an original observation and indeed there are companies using that observation in product development and business strategy.

I think we need to give credit to the smaller companies like RevStor. They have selected a tighter niche in which to ply the technology of grid. Success at this more focused level will no doubt bubble up and create that ground swell of interest that we need at the larger enterprise levels.

Posted by Greg Nawrocki on May 8, 2007 11:49 AM