Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!usc!rpi!batcomputer!theory.TC.Cornell.EDU!christos
From: christos@theory.TC.Cornell.EDU (Christos S. Zoulas)
Newsgroups: comp.sys.apollo
Subject: nfs/rpc question
Message-ID: <1991May23.055131.14273@batcomputer.tn.cornell.edu>
Date: 23 May 91 05:51:31 GMT
Sender: news@batcomputer.tn.cornell.edu
Organization: Electrical Engineering, Cornell University, Ithaca NY
Lines: 33
Originator: christos@theory.TC.Cornell.EDU
Nntp-Posting-Host: theory.tc.cornell.edu


Hello,

I have managed to compile most of the rpc and rpcsvc 4.0 stuff successfully
on our 400s's but I am having a couple of problems with rpc.rexd, and
in general with any daemon that tries to run execve() or chdir() on
an nfs directory. For example execve, it seems to hang forever if it tries
to execute an NFS directory, or randomly if the nfs path is too deep.
If I stat each path component before I finally execute execve it works fine.
[Painful, but at least there is a workaround]

Chdir is worse. I cannot seem to be able to change deeper than two levels of
nfs mounted directories! 

Are these known problems with NFS 2.1? Is there some magic environment
variable the executable or some magic system call the daemon must do for these
systems calls to work properly?
[I set all the ones mentioned in environ(7)...]

Other than that, I've gotten rpc.rstatd, rpc.rusersd and the 4.3BSD+NFS inetd 
to work perfectly... 

Of-course with irritating slowness of NFS, I am really starting to question
if it is worth all the trouble. It feels that the remote filesystems are 
made out of floppies :( Is 2.2 any better?

Thanks in advance,

christos
-- 
Christos Zoulas         | 389 Theory Center, Electrical Engineering,
christos@ee.cornell.edu | Cornell University, Ithaca NY 14853.
christos@crnlee.bitnet  | Phone: (607) 255 0302, Fax: (607) 255 9072