 |
 |
On 15 Oct 2009, at 09:59, John Gordon wrote: > Graeme, so if sites have SL5 nodes now you will use them? Of course - we do already at RAL, RALPP and Glasgow. > Will you use > both at a site? Generally no. We'd prefer to move production to SL5 in one go. We do have some SL4 nodes still at Glasgow and that's ok if they are behind a different CE (it's explicitly in the SL5 migration page). > Do you use the SL5 native kits now? No, not yet. Cheers Graeme > > John > >> -----Original Message----- >> From: Testbed Support for GridPP member institutes >> [mailto:[log in to unmask]] On Behalf Of Graeme Stewart >> Sent: 11 October 2009 13:49 >> To: [log in to unmask] >> Subject: Re: User Analysis Test (UAT) in Oct., 2009 >> >> On 9 Oct 2009, at 18:03, Elena Korolkova wrote: >> >>> Hello >>> >>> I have two questions: >>> 1. Should we have Squid server before this test? >> >> It depends upon the analyses users might like to run. >> However, it's unlikely that much will depend upon DB access >> as people will likely run over MC rather than cosmics. >> >> However, the squid is a definite priority, so I would aim to >> get that installed in the next week (it only takes a few >> hours) so that you have it. >> >>> 2. Will this tests run on SL5 nodes only? I mean if we want to >>> participate should we move to SL% or we can have SL4 WNs? (In >>> Sheffield we can't afford to have both) >> >> It doesn't matter - there's nothing today which requires SL5 nodes. >> However, again we'd like to see the SL5 migration happen soon, so >> that >> SL5 native built kits can start to become the default. We >> gain 15-20% CPU with SL5 gcc4.3 code, so your site works >> better for ATLAS on SL5. >> >> Both things are important, but squid is definitely faster to >> deploy so a priori I would deploy that first. >> >> Cheers >> >> Graeme >> >> -- >> So we made our own computer out of macaroni pieces And it did >> our thinking while we lived our lives >> > -- > Scanned by iCritical. -- So we made our own computer out of macaroni pieces And it did our thinking while we lived our lives