ManicDeity writes "'Forget the Matrix. It's time for the Grid-' From the good people at CERN who brought us the WWW comes a new network designed for scientific use named GriPhyN (Grid Physics Network)...
Not true - CERN is neither responsible for GriPhyN (which is an NSF project), nor the invention of the data grid concept. The experiments at CERN (and other places) "merely" drive the need for something like GriPhyN.
That said, there is a European project similar in scope to GriPhyN, which CERN is a part of: the DataGrid Project.
For a book about Grids, you can look for "The Grid: Blueprint for a New Computing Infrastructure", edited by Ian Foster and Carl Kesselman. Both are on the GriPhyN project: Dr. Foster is a principle investigator on GriPhyN, and Dr. Kesselman is one of the Senior Personnel.
It's nice to see that someone finds a use for perfectly good existing technology instead of rushing to push everyone into the next new thing. By using a peer-to-peer or other existing technologies the scientists are giving themselves credit for knowing what they need instead of believing in hype.
As much as I'd like to say that scientists are naturally skeptical of the gee-whiz hype associated with new technologies, it really isn't true.
I'm not associated with GriPhyN, but my boss is one of the principle investigators. There was a distributed computing project called Nile which did suffer (IMHO) from pushing some new technologies where it wasn't really needed. The result was slowed development, the initial prototype came out a lot later than it should have, and the whole project suffered.
So I think the decision to go with proven tech was due to a lesson learned the hard way.
I couldn't even set up a nationwide backbone capable of routing gigabytes of data with 11.9 million dollars. Let alone petabytes of data which not only have to be routed, but have the additional overhead of abstraction of jobs that have to be completed, and output returned to the original user?
The 11.9 million bucks is for R&D only - it will pay mostly for people. There will be some clusters purchased with this dough, but just for development. The money to purchase the big iron will come from a separate grant, which is just beginning to be negotiated.
Even if this happens, what if some scientist from geneva wants to use all this distributing power to give his distributed.net scores a boost? It's supposed to be as easy as using the electrical grid. "When scientists submit a processing job to this worldwide network of computers, the only thing they care about is that the job gets done. They don't know which machine (or machines) the work gets farmed out to." Who's going to pay for all this processor time? I guess they will have to install meters on the side of the scientist's building, and someone will have to come check it monthly.
I believe there is no answer yet to the questions "how do we guarantee that only `grid` jobs run on the network?" or "how do we bill the grid customers (if at all)?".
... Fold proteins, find aliens, break crypto... what else?
How about exploring the fundamental forces of nature and structure of the universe? That's what the GriPhyN project (Grid Physics Network) will be supporting. Some other related grid projects and forums can be found on the Links page.
Another good description of GriPhyN, maintained by one of the principle investigators, is
here.
Slashdot Mirror
Not true - CERN is neither responsible for GriPhyN (which is an NSF project), nor the invention of the data grid concept. The experiments at CERN (and other places) "merely" drive the need for something like GriPhyN.
That said, there is a European project similar in scope to GriPhyN, which CERN is a part of: the DataGrid Project.
For a book about Grids, you can look for "The Grid: Blueprint for a New Computing Infrastructure", edited by Ian Foster and Carl Kesselman. Both are on the GriPhyN project: Dr. Foster is a principle investigator on GriPhyN, and Dr. Kesselman is one of the Senior Personnel.
As much as I'd like to say that scientists are naturally skeptical of the gee-whiz hype associated with new technologies, it really isn't true.
I'm not associated with GriPhyN, but my boss is one of the principle investigators. There was a distributed computing project called Nile which did suffer (IMHO) from pushing some new technologies where it wasn't really needed. The result was slowed development, the initial prototype came out a lot later than it should have, and the whole project suffered.
So I think the decision to go with proven tech was due to a lesson learned the hard way.
---
#include "disclaimer.h"
The 11.9 million bucks is for R&D only - it will pay mostly for people. There will be some clusters purchased with this dough, but just for development. The money to purchase the big iron will come from a separate grant, which is just beginning to be negotiated.
Even if this happens, what if some scientist from geneva wants to use all this distributing power to give his distributed.net scores a boost? It's supposed to be as easy as using the electrical grid. "When scientists submit a processing job to this worldwide network of computers, the only thing they care about is that the job gets done. They don't know which machine (or machines) the work gets farmed out to." Who's going to pay for all this processor time? I guess they will have to install meters on the side of the scientist's building, and someone will have to come check it monthly.
I believe there is no answer yet to the questions "how do we guarantee that only `grid` jobs run on the network?" or "how do we bill the grid customers (if at all)?".
Cheers,
Craig
How about exploring the fundamental forces of nature and structure of the universe? That's what the GriPhyN project (Grid Physics Network) will be supporting. Some other related grid projects and forums can be found on the Links page.
Another good description of GriPhyN, maintained by one of the principle investigators, is here.