 |
 |
Hi Alessandro, Since this is a SubCluster property, each Subcluster can publish this info for itself properly. It is understood that SubClusters are organized as homogeneous sets of WNs (at least in in architecture, but preferably in all properties). So, this can be done exactly in a way experiments would want. In terms of site-info.def, future of YAIM should see (according to the documentation) inheritance of variables in a sense that you can have top-level site-info.def where default values of all variables are defined, and one sub-directory per node with site-info.def containing just the variables that have non-default values for this each particular node. If the variable is defined in the per-node site-info.def, it will have this value, if not, it will inherit the value from the top-level site-info.def. Or so I imagine :)) Best regards, Antun ----- Antun Balaz Research Assistant E-mail: [log in to unmask] Web: http://scl.phy.bg.ac.yu/ Phone: +381 11 3160260, Ext. 152 Fax: +381 11 3162190 Scientific Computing Laboratory Institute of Physics, Belgrade, Serbia ----- ---------- Original Message ----------- From: Alessandro De Salvo <[log in to unmask]> To: [log in to unmask] Sent: Thu, 26 Apr 2007 16:01:22 +0200 Subject: Re: [LCG-ROLLOUT] Distinguishing sites/clusters with different host CPU types > Hi, > this is very important when you have mixed architectures in the same > cluster: you should be careful of the value you put there because the > experiments will decide which software flavor to use, depending on > the value you put there. If you have a mixed 32/64 bits cluster then > you should use "i686" everywhere, taking care of the installation of > the 32 bits compatibility libraries in all your 64 bits WNs, > otherwise we could get lost. Cheers, > > Alessandro > > On Thu, 2007-04-26 at 14:50 +0100, Burke, S (Stephen) wrote: > > LHC Computer Grid - Rollout > > > [mailto:[log in to unmask]] On Behalf Of Antun Balaz > > said: > > > GlueHostArchitecturePlatformType: `uname -m` > > > > I'd just like to emphasise that it needs to be the result of uname -m on > > the WNs, not the CE! (I'm sure you know that, but ...) > > > > Stephen > > > -- > ========================================= > Alessandro De Salvo > I.N.F.N. - Sezione di Roma1 > Dip. Fisica, Universita' degli Studi di Roma "La Sapienza" > Piazzale A.Moro 2 - 00185 Roma Italy > e-mail [log in to unmask] > voice +39 06 4991 4325 > ========================================= ------- End of Original Message -------