Slashdot Mirror
Towards a World Wide Grid?
Roland Piquepaille writes "In recent months, the concept of 'cloud computing' was all the buzz. European researchers think about another name, the World Wide Grid, which could run on top of the Internet. In an article to appear soon, ICT Results will report about the g-Eclipse project. As the scientists said, 'the g-Eclipse project aims to build an integrated workbench framework to access the power of existing Grid infrastructures. The framework will be built on top of the reliable eco-system of the Eclipse community to enable a sustainable development.' The project started in July 2006 and was successfully completed in June 2008 for a total cost of €2.5 million, including a EU contribution of €1.96 million."
I'm a comp sci nerd, but I didn't get anything from the summary. I'm definately not going to read the article.
These Italian schemes are a threat to America and everything Americans hold dear: our children, our base-ball parks, and our Hot Dogs. I say that these "nets" should be banned from the American intern net for security and moral purposes.
If you google "integrated workbench framework" (in quotes), all of the 250-something results seem to refer to this project.
Have you read my blog lately?
I am not a cloud expert, but: Anything with "grid" in it makes me think "designed by committee" and "sucks"... and the fact that the effort described in TFA was funded by the EU doesn't make me feel any better about it. Maybe it would make more sense to wait until something like Hadoop takes over the world, then just standardize existing practice. (Apologies to my friends in the grid world.)
This will not work for most consumer applications. You want to play a video game -- you can't 'outsource' that processing to a grid because of latency -- in the time it takes to submit the raw data and get a result back, your system could have done it locally. It might work for complex photoshop filtering where the user might have to wait a few seconds to a minute. It would certainly be nice for transcoding video. In short, "grid" computing is good for non-interactive (batch) tasks. Most consumers have little need of this. It's far more useful for commercial enterprise.
Not only is there a latency issue, but there's a bandwidth issue -- a really big one. Very few people have a fiber link to the internet and unlimited bandwidth. And there's a lot of businesses out there that want to have it stay that way. Comcast comes to mind as internet equivalent of OPEC -- except instead of barrels we've got gigabytes. It's an artificial market, but until the infrastructure is radically modified, grid computing is only going to be happening between large data centers made for and run by commercial business. And by the time the bandwidth issue is "solved", grid computing might be meaningless because the hardware will be so much faster and storage space so much more plentiful that there's little justification for Joe Average.
#fuckbeta #iamslashdot #dicemustdie
Imagine a beowulf cluster of those.
(Man, isn't everyone tired of that joke by now? Oh, you are? Sorry.)
Could this be Web 3.0?
Sig this!
From the article:
Where the internet is a communications channel between computers, the grid goes beyond this by not just using the internet for communications but also as a means of sharing computing resources. Every computer and user can access and make use of the combined resources of the grid.
And just how long will it be before someone decides to create a WWG application that uses it as one vast storage pool of copyrighted material with distributed indexing of the contents and the RIAA, MPAA, ... of the world sue the whole thing into non-existance or buy laws to make it a criminal offence to run it?
Structured correctly you wouldn't know who was adding to it or downloading data from it. After all a download would be just be a request to replicate a bit more data making a vastly distributed virtual filesystem a bit more redundant. You may not even be able to tell if it was someone making a request to make a local copy or the software automatically increasing storage redundancy of static data (assuming that there's no logging).
Now granted its a really cool idea and its even quite handy for organizations who need more power. There have been times when I'm sitting at work running an FPGA simulation and I click the go button and go for a walk.
At home reading some news doing some simple development and hell even compiling a new kernel here and there I'm generally happy with the speed of my laptop. Hell I'm even contemplating giving up some speed in exchange for portability so long as I don't have to skimp on resolution.
Much of the world is already in a power crunch. Not only a power crunch but a fuel crunch in general. So when we decide that gas is too expensive and we all start plugging our cars into the grid the problem gets even worse. Can you imagine if ALL of the world's computers hit 100% usage all the time so they could constantly be servicing who ever needed resources at the moment?
Laptop's probably wouldn't want to opt into this network for the huge hit they'd take in battery life. Not to mention the increased bandwidth requirements needed for what most users consider background tasks.