 |
 |
Hi, That's great, Dieter! Kaushik is here at CERN for a few days starting on Thursday. We will try to run some production jobs at SLAC. There will no doubt be some issues that require the attention of locals. For example, outbound connections from worker nodes to CERN is a continuing issue. We will need help setting up a proxy. Who should we contact about this? As a _short-term_ test to confirm there are no other problems, is it thinkable to allow say one specific worker node to connect to CERN? We will NOT run real production this way. Cheers. Charlie -- Charles C. Young M.S. 43, Stanford Linear Accelerator Center P.O. Box 20450 Stanford, CA 94309 [log in to unmask] voice (650) 926 2669 fax (650) 926 2923 > -----Original Message----- > From: [log in to unmask] > [mailto:[log in to unmask]] > On Behalf Of Dieter Best > Sent: Tuesday, March 21, 2006 9:54 AM > To: Best, Dieter S. > Cc: atlas-sccs-planning-l > Subject: New JT for local MySQL: Progress > > Hello, > this is just to let you know that I fixed some problems with > the new JT, the port number was missing and also mysql > database names are case sensitive. csc.simul.trf is now > running at SLAC! - I think that's an important milestone. > -- Dieter > > Dieter Best wrote: > > > Hello, > > I have installed the new JobTransformations at SLAC (cache > 11.0.41.6) > > and gave the CSC production a shot. There is a new environment > > variable that needs to be set, for example > > T_ATLASMYSQLSERVER=osgserv04.slac.stanford.edu > > But I have encountered problems. Not necessarily related to our > > database replica since if I replace > osgserv04.slac.stanford.edu with > > atlasdbdev.cern.ch I find similar problems (testing on an > interactive > > machine). If I unset this environment variable > csc.simul.trf works as > > expected. I emailed Alessandro di Salvo et al my findings, > maybe more > > tomorrow. > > -- Dieter > > > > Dieter Best wrote: > > > >> ... and Martin Woudstra et al, the people working on the > new python > >> Job Transforms, plan to add the same functionality. Looks > like this > >> feature is in the main stream now. > >> -- Dieter > >> > >> -------- Original Message -------- > >> Subject: Re: New JT service for local MySQL replicas > >> Date: Mon, 20 Mar 2006 17:14:35 +0100 > >> From: Martin Woudstra <[log in to unmask]> > >> To: Alessandro De Salvo <[log in to unmask]> > >> CC: Davide Costanzo <[log in to unmask]>, Mireia Dosil i > >> Bonmatí <[log in to unmask]>, Andrea Di Simone > <[log in to unmask]>, > >> Kaushik De <[log in to unmask]>, Dieter Best <[log in to unmask]> > >> References: <[log in to unmask]> > >> > >> > >> > >> Hi Alessandro, > >> > >> Can you tell me where I can find this new tool? > >> I will then also add the same functionality to the python > jobtransforms. > >> > >> Cheers, > >> Martin > >> > >> > >> Alessandro De Salvo wrote: > >> > >>> Dear all, > >>> I've just introduced a new JT tool to allow to use the > local MySQL > >>> replicas. The new service is called > "addLocalMySQLSupport" and will > >>> appear from JobTransforms 11.0.41.6 in the production > caches, as it > >>> will be ready. > >>> I've changed all your transformations (csc.*, ctb.*) in > that cache > >>> to add the new service, so you may want to take a look at them. > >>> This service is very useful for sites like SLAC, where we need to > >>> use local replicas to access the CondDB. In that case the > execution > >>> agent should only set the env variable T_ATLASMYSQLSERVER to the > >>> name of the local server with the replica(s). If the > variable is not > >>> set, the service addLocalMySQLSupport does nothing. > >>> I'm waiting a bit before creating the cache because I want to add > >>> some other changes in there too, but I'll soon put it in > production. > >>> Cheers, > >>> > >>> Alessandro > >>> > >> > > > >