lunedì 29 ottobre 2007

ssh tunneling to CERN

Two methods can be used:

1. modify ssh client configuration file
.ssh/config

#
#Usage:
#ssh -N -f -q lxtunnel; ssh pccmsto02

Host lxtunnel
HostName lxplus.cern.ch
User argiro
LocalForward 7777 pccmsto02.cern.ch:22
LocalForward 7778 pccmsto03.cern.ch:22
LocalForward 7779 lxcmsf3.cern.ch:22

Host pccmsto02.cern.ch
HostName localhost
User argiro
Port 7777


Host pccmsto03.cern.ch
HostName localhost
User argiro
Port 7778


Host lxcmsf3.cern.ch
HostName localhost
User argiro
Port 7779

then do :

ssh -Nfq lxtunnel this opens the tunnel

then you can normally ssh to lxcmsf3, or the other machines you have configured

2. Manual tunneling

ssh -L 7777:lxcmsf3.cern.ch:22 lxplus.cern.ch

meaning "forward requests to localhost:7777 to lxcmsf3".
In another window do

ssh -p7777 localhost

venerdì 19 ottobre 2007

ECALTB06 analysis and submission howto

CMSSW 1_1_0, plus ECALTBH4_0_1_0
https://twiki.cern.ch/twiki/bin/view/CMS/ECALTBH4

Never versions coming sometime

Packages

User/EcalTBAnalyzer
User/EcalTBAnalyzerData
RecoTBCalo/EcalTBRecProducers

should be extracted from http://cmssw.cvs.cern.ch/cgi-bin/cmssw.cgi/UserCode/Torino/Ecal/TestBeam/

or CVS


go to the production area:

User/EcalTBAnalyzer/data
run `eval scramv1 runt -csh`
run the sumbission script

./tbsubmit_single -c tbanalysis-sm16.cfg.tmpl -r files-sm16.txt -e lxcmsf3:/data/argiro/ecaltb2

where

-c file is the configuration template
-r file is the file containg the list of files (runs) to be processed
-e dir is the destination directory, to be created ahead of time

bjobs can be used to monitor, dummye and dummy will contain stdout and stderr
of the job submission, while processing logs will be in the destination dir

It is a good idea to stage files from castor first

csa07_stager.py -r filelist -s


Analysis macros are in User/EcalTBAnalyzer/data

runxtals.C -> produces files containing only one xtal. Must be modified to select which xtals
select_xtal.C(xtal, dir) -> select one single xtal (used by runxtals)
lin.C -> the do-it-all analysis macro

TBNtuple.h is needed from User/EcalTBAnalyzerData/interface

giovedì 4 ottobre 2007

Work - 4 Ott

TB: energies of some runs were missing in enerUff.txt file, that were discarded by select_xtal
Warning added, have to check that all useful runs have energy value

PhiSim: running onlin phisim on MinBias sample w/ raw and L1 info

DAQ: converting matric to xdata::vector

lunedì 1 ottobre 2007

Work - 1 Oct 07

Testbeam

Bug found in analysis code: TBNtuple.h was present in dir too, and was the one linked -> posx was
always zero. Have to reprocess

./tbsubmit_single.py -c tbanalysis-sm16.cfg.tmpl -f files-sm16.txt -e lxcmsf3:/data/argiro/ecaltb2/sm16-v2


another bug fixed. Committed to cvs