On 12/3/18 5:22 AM, g4sra wrote:
> >From my perspective, this topic has had some very interesting
> contributions. Thank you all whom have contributed.
>
> To pick out just one as an example, I had considered NIS\YP to be (or
> rather didn't consider because) all but defunct, and not taken it's
> simplicity and reliability over other methods into consideration.

NIS/YP is especially interesting for me as something long unused.

At one point in my career I had to restore a plant that use a semi centralized NIS/YP.  I got the bright idea of putting a YP slave on the all the hosts and syncing those to the
master.

It took me a week but I found that upstream had a bug in the slave scripts such that they would never sync.  The bug didn't exist in sunos or solaris so it was unique to Linux.

I've found that AD is VERY sensitive to time differences, even in a pure windows environment.  How Windows admins tolerate it I have yet to figure out.

The pam module, oddjob makes it somewhat better, but a bit weird.

The stated use of AD for resource access might be better served by full on Samba 4, but AD and GPOs can perform that kind of limiting

PXE boot is well known for the type of lab/classroom environment... Long ago, I used bootp for doing mass installs/reinstalls of OS/2.  It was pretty well documented in the IBM
Redbooks.