Usage
I’ve created a new task but, when trying to run it, it says it doesn’t exist.
You need to rerun law index --verbose
I want to reinstall all the software that is downloaded when running the setup.sh
script for the first time.
You can run CMT_FORCE_SOFTWARE=1 source setup.sh
Someone has updated a CMSSW module I’m using and I want to obtain the latest version.
The easiest (and slowest) way is to reinstall the whole software stack (see a previous FAQ).
This will download again all CMSSW modules and recompile, but keep in mind it will delete the whole
data folder, so make a backup of your changes. It can also be done manually going into the module’s
folder, running git pull origin branchname (usually branchname = main), going back to
the ${CMSSW_BASE}/src folder and recompiling (scram b).
I launched some jobs to htcondor and after one day or so they still appear as idle.
If you are running at CIEMAT, remember you should store your analysis setup under /nfs, as /afs is not mounted in the htcondor nodes.
A colleague has already run some PreprocessRDF, Categorization or
MergeCategorization tasks. Can I use his/her ntuples?
Yes, just set CMT_STORE_EOS_PREPROCESSING, CMT_STORE_EOS_CATEGORIZATION or
CMT_STORE_EOS_MERGECATEGORIZATION to the new path (e.g.
/eos/user/${CMT_CERN_USER:0:1}/$CMT_CERN_USER/cmt or /nfs/cms/$CMT_CIEMAT_USER/cmt)
It looks like I’m missing events after running PreprocessRDF for some datasets, or I’ve
different number of events after running PreprocessRDF with the same selection more than once.
Firstly, you need to check the PreCounter for those datasets. If the number of events is different
compared to what appears in DAS, surely the problem is that you’re taking a corrupted file within the dataset.
A possible solution would be to fix the prefix for the dataset in your config file, so that you force the
program to take it from a server where you know it is not corrupted. From DAS, you can check the sites where
a dataset replica is stored by typing site dataset=/your_DAS_dataset/. You can then look for the prefixes of
the different sites by typing on your terminal (here for the european redirector)
xrdfs xrootd-cms.infn.it locate -h /one_of_the_files_from_DAS_dataset.root. The ones where it says Server are valid.
TIP: If you’re running on lxplus, try to use the T2_CH_CERN prefix when possible. It also runs much faster.
The documentation doesn’t compile, It gets this error:
Could not import extension sphinx.builders.linkcheck (exception: urllib3 v2 only supports OpenSSL 1.1.1+, currently the 'ssl' module is compiled with 'OpenSSL 1.0.2k-fips 26 Jan 2017'. See: https://github.com/urllib3/urllib3/issues/2168) make: *** [Makefile:20: html] Error 2
Run cmt_pip_install urllib3==1.26.6 and try again.