[Halld-offline] computing model, alpha release

[David Lawrence]

Thanks for the response Graham.

     I think we clearly need to discuss this within the collaboration to 
clarify what it is we expect to use the grid for. If we only need JLab 
to provide GRID access to data files is one thing. If we want it to 
provide CPU resources from the JLab computing farm so jobs can be 
submitted remotely through the GRID, that's another thing entirely.

     Ideally, we would discuss this at the Offline Software meeting next 
week. However, I think it is critical that Richard be involved in the 
discussion and highly desirable that some of our remote institutions who 
have direct GRID experience (IU, CMU, Regina, ...) be involved as well. 
So Richard, are you available next Wednesday at 13:30?

Regards,
-David

On 2/15/11 4:14 PM, Graham Heyes wrote:
> I agree with a lot of what you are saying and have nothing against a GRID based solution to accessing off-site resources.
>
> The issue at the moment is that we are tasked with providing computing requirements and existing resources, for example your four universities. Using these numbers we can look for implementations that will meet the requirements using all of the available resources including existing JLab and off-site resources and projected future funding. If GRID is the best fit then that is the way we go. The problem is that, at the moment, everything is anecdotal with no hard numbers. What offsite resources are there? Are they already part of a GRID? If they are then what would be the cost to JLab of joining that GRID? If we can answer that AND it is lower in cost than any sensible proposal that anyone else can come up with then we back the winner.
>
> Graham
>
>
> On Feb 15, 2011, at 3:17 PM, David Lawrence wrote:
>>      Certainly we should be able to put forward a set of requirements. I'd have to point out though that we are already developing expertise using certain GRID technologies between at least 4 of our universities (if I'm counting right). The thing that has me confused though is that this idea of it taking so much manpower to support it. I know of a certain collaborator in GlueX who, in addition to maintaining a full teaching schedule, maintaining his own computing farm, and heading a group responsible for one of our major detector systems in GlueX, has still managed to get his farm on the GRID and use it as well as other GRID resources for simulation studies. I can't imagine him spending more that 15-20% of his time on this. Why do we expect it to take a JLab IT person (with a CS degree) to have to spend so much more time? I must be missing something.