1) there are "virtual modules" that do nothing but pull over a stack of dependencies. But fairly rare.
2) the correct way to simplify adding modules to the same type of machine — assembling packages in the usual your system format. For centos || fedora search Google perl2rpm, cpan2rpm. Collect your set of packages for modules which do not exist in the repositories.
2.1) just In case, let me add that the packages of Perl modules usually have the name of a type of perl-*.rpm, where instead of stars you want to substitute the package name with span. For example, the desired module URI::Escape. Looking at search.cpan.org find it in the package "URI-1.55". So try to find already assembled package perl-URI if you do not find — collect themselves.
3. Once again, use the packages. First, it will be easier to track versions of the modules and upgrade them if necessary simultaneously. Secondly, you will not be faced with the fact that since the last installation of the modules on cpan there is a new version. This can lead to quite interesting effects. For example, Storable checks when unpacking the data, what version they were serialized, and in case of discrepancy send quite far.
\r
Success!
