Ganga, grid usage, etc: Which practices are good and which are bad?

[ryuwd]

I've had differing responses to practices I've learned from the Starterkit, from what I've heard in Distributed Analysis, about a couple of things:

HTCondor usage

corresp. page: https://lhcb.github.io/starterkit-lessons/self-guided-lessons/htcondor.html

Apparently using HTCondor / LXBATCH is against LHCb fair usage policy? See https://mattermost.web.cern.ch/lhcb/channels/distributed-analysis/s7ckr86kwpyxubu3n6mfsxupia

Replication of grid files to CERN-USER storage elements for analysis purposes

corresp. page: https://lhcb.github.io/starterkit-lessons/second-analysis-steps/managing-files-with-ganga.html I've been discouraged from doing this, and told I should instead access my files by authenticating with a grid certificate.