LHC Computer Grid - Rollout
> [mailto:[log in to unmask]] On Behalf Of Gonçalo Borges said:
> CE_OTHERDESCR="Cores=3.84286,Benchmark=9.45-HEP-SPEC06"
>
> may solve the GSTAT 2.0 issue but won't it bring additional other
> problems specially in the future accounting reference?
The installed capacity accounting will calculate your total CPU capacity as LogicalCPUs*Benchmark, so you should make sure that both of those values are right, i.e. LogicalCPUs should be the total number of cores in the subcluster, and Benchmark should be the *average* HEPSPEC for those cores. (And hence Cores*Benchmark is the average power per physical CPU, although the accounting doesn't use that directly.) That should be the real power of your system, regardless of any scaling in the batch system.
On the other hand, for job accounting APEL doesn't look at any of those values. Up to now it has always used the SI00 value (still in specint units, but scaled to HEPSPEC with the standard factor 4). However, there is now a version of APEL (about to be released I think) which will use the CECapability CPUScalingReferenceSI00 value if you publish it. Once that's out then you can separately publish the physical power in SI00 and the scaled power used in the batch system in ScalingReferenceSI00.
Stephen
|