NFS, RFS and the meaning of life

[Ed Mackenty]

In article <2225 at brl-adm.ARPA> bzs at bu-cs.bu.EDU (Barry Shein) writes:
[A request for comments on RFS and NFS, and NFS yellow pages.]

At this site we have two Suns using NFS to share each other's disks, and
about 20 AT's using PC-NFS that can access the Suns disks as well.
The suns can't get to the AT's disks since MS-DOS is brain-damaged :-).
It gives us a really nice development environment for our AT-based code.
We keep all the source under SCCS on a Sun, edit on the Sun (some folk
edit with PC editors), and run make on an AT to put it all together.
The Sun's disk looks like just another hard disk to the AT.  When we want
to distribute a new version of some program, we just put it in the right
place on the Sun, and every AT gets it.  They can even have remote
directory specs in their PATH variable.  Although there are some problems,
they have to do with PC-NFS, not NFS in general (such as not being able
to execute jobs on the Sun from an AT without using telnet and logging in).

We do not use the Yellow Pages functionality at all here.  Since all our
NFS machines really only share two disks, maintaining the /etc/hosts files
on the two Suns is trivial.  Since the AT's can't share each other's disks,
there is no reason to update their hosts file when a new AT is added.

We really like NFS.  Without it, our software development on the AT's
would be slowed incredibly and we'd have all kinds of source control
problems.  The folks at Sun said the next version of PC-NFS will have
an RPC service to invoke remote shells which will allow us to issue SCCS
commands directly from an AT.  I might actually start using mine if it
could do that. :-)
	- MacK.
-- 
- MacK		Edmund R. MacKenty                                         
USENET:		...{seismo,allegra}!harvard!humming!mack             
DISCLAIMER:	Who, ME?
QUOTE:		"Nothin' to do but smile, smile, smile." - The Grateful Dead