Hi All...
I'm implementing the HEP06 benchmark values in the information system,
and I profit the opportunity for reviewing some of the information I'm
publishing. For you reference, I'm looking to:
https://twiki.cern.ch/twiki/pub/LCG/WLCGCommonComputingReadinessChallenges/WLCG_GlueSchemaUsage-1.8.pdf
I have 113 hosts (2 quadcore cpus each) offering (113x8=) 904 cores.
Here are my doubts:
1./ What's the correct value for GlueSubClusterPhysicalCPUs?
I have set it to 226 (113 x 2 quadcore cpus) but I'm wondering if it
shouldn't be 113. I'm doubting because one of the following variables is
SMPSize, defined as the number of cores per WN, which in my case is 8.
Using those variables for some simple math operation like
(GlueSubClusterPhysicalCPUs=226) x (SMPSize=8), you get a wrong value
for the total number of logical CPUs (Cores).
2./ What's the correct value for GlueSubClusterLogicalCPUs?
I've set it up to 904 (the total number of cores) but since we have
sufficient memory per core, we allow some overbooking, and I wonder if I
should publish the total number of slots (allowed jobs) instead...
3./ I've set up "GlueHostProcessorOtherDescription:
Cores=8,Benchmark=54.90-HEP-SPEC06". I assume that the benchmark value
should be published by server (and not by core), and that the number of
cores is the total number of cores per server, and not per CPU as stated
in the docs (which in my case would be 4 since each host has 2 quadcore
cpus). Is my assumption right?
Thanks in Advance
Cheers
Goncalo
|