[Pvfs2-developers] Active Storage?

[Rob Ross]

I'm sure we'll all get sync'd up soon. No worries.

Rob

Jarek Nieplocha wrote:
> David,
> 
> We have an early prototype work using a different approach than the 
> original
> kernel space implementation of Active Storage. It is done purely in user 
> space.
> One advantage of the approach is that improved portability and simplified
> porting to PVFS2. Working against the SC deadline so far we only had time
> to evaluate both implementation approaches using Lustre. Other than
> testing that it works we have run only a few experiments with PVFS.
> 
> Because this development is so recent and you never know about the fate
> of the SC papers submissions, people other than Rob and Evan do not
> know about it yet.
> 
> Jarek
> 
>>  > I stand corrected; the PNNL team has integrated components into 
>> PVFS and
>>  > is testing these components. I don't believe that they are using the
>>  > FUSE approach that you mention.
>>  >
>>
>> This concerns me a little.
>>
>> /me gets on his open source soap box
>>
>>  From my experience in the open source world if there's any chance for
>> a third party group (pnnl) to produce code for an open source project
>> it has to be done in a very open and public manor with the input and
>> collaboration of the current developers of the original project. Since
>> design decisions are being made and new code is being developed it
>> should be in line with the original intent of the project and should
>> adhere to the coding standards put in place by the original project
>> itself. Usually this means sharing code, prototypes, and being
>> involved with design discussions.
>>
>> If communication isn't happening between the third party group and the
>> original project developers on all aspects of the development. Then it
>> becomes very easy for the open source project developers to dismiss
>> the efforts of the third party group (I would rather not see this
>> happen again). Usually siting, poor coding, poor design decisions, and
>> not being in line with the original project goals.
>>
>> Has this sort of communication been happening? Have you seen any of
>> the papers written up by people at pnnl on this project? and efforts
>> involving your code?
>>
>> Thanks,
>> - David Brown
>>
>