Closed DennRoy closed 1 year ago
I am fine with this, but just FYI: If you have to frequently submit jobs split on an event-basis, you can save a lot of time by storing the total number of events per NANOAOD file (whether it is on the GRID or on a local storage system) into a text file with
pico.py write -y UL2018 --nevts
You can also add the option --try 6
to try again 6 times before giving up, and specify with e.g. -s DY*Jets
what samples you want to get the filenames & event number for.
Once you have the text files, you can set the filelist
in the sample config file, e.g. here.
For more instructions, please see https://github.com/cms-tau-pog/TauFW/tree/master/PicoProducer#samples
@DennRoy, apologies for not following up on this. Was the problem solved with my previous suggestion in the end?
Hi Izaak, yes, that's right. It's better with your suggestion.
Sometimes the script crashes for me with an error because a file couldn't the read in DAS, but then it works the second time, only to crash again at a later time. Doing multiple attempts would fix this.