|
Comment: |
Hi,
The document describes a CIM job submission and management model for grid (batch) scheduling systems. To what extent has this model been vetted
against current batch scheduling systems such as LSF, PBS, Grid Engine, Condor, DQS, etc? I see this as necessary to substantiate the claim in the
second paragraph of page 5 that it "contains the standard information and relationships..."
I appreciate the presence of the disclaimers in the same paragraph to not address particular software or methods, but as GGF Recommendations serve a
purpose of facilitating interoperable implementations, and where one or more of these batch scheduling systems would be likely to be used in a typical
system, I think it is important to establish that this specification reflect at least a consensus of a reasonable and sufficient subset of these
suppliers, especially given that the most desirable parties to implement and support this model might be the batch system suppliers themselves.
Now, I may completely misunderstand the application of this proposal as it relates to current batch system products, and my question here may be off
the mark. If so I would welcome correction.
Thanks,
John Tollefsrud, ISP AD
|