Subject: NIS and {Net,Free}BSD
To: None <current-users@netbsd.org>
From: Brian C. Grayson <bgrayson@marvin.ece.utexas.edu>
List: current-users
Date: 08/13/1998 15:22:17
  We run a mostly-NetBSD group of machines here, but due to the
lack of SMP support (yet -- sniff!), we're trying out FreeBSD on a quad
box.  It's been rocky so far, and NIS is misbehaving, and I'd
appreciate any help with our NIS problem.

  The setup:  orac (128.83.52.138) is our NIS server, for the
domain ``pdslab.''  sim1 is a quad-PPro box running a recent
FreeBSD SMP snapshot (although I also had NIS problems with a
2.2.7-RELEASE uniprocessor snapshot).

  1.  If sim1 runs ypbind, it will never bind.  root logins won't
      even work!?

  2.  If sim1 runs ypbind -ypsetme, followed by ypset orac, it will
      bind, but after a few minutes, ypset exits saying it was
      unsuccessful.  At this point, NIS is working, users can log in,
      etc.  However, after some time, ypbind will start printing:

Aug 12 14:25:14 sim1 ypbind[128]: NIS server [128.83.52.138] for domain "pdslab" not responding
Aug 12 14:26:08 sim1 ypbind[128]: NIS server [128.83.52.138] for domain "pdslab" not responding
Aug 12 14:37:51 sim1 last message repeated 11 times

  I know for a fact that the server was okay during this time, as
our other machines (running NetBSD) were all happy.

  I'm at a loss to debug this further.  I can provide more
details, or tcpdumps, if someone tells me what to provide.
Everything is fairly standard -- domainname set to pdslab on all
machines, sim1 is listed in the ACL for the pdslab domain.

  Note that my reason for posting here, and not to a FreeBSD
list, is because we have had problems with a NetBSD NIS server and
{FreeBSD,Linux,Solaris} clients.  It may be that all of them have
flaky clients, for all I know.  On all three systems, a simple
ypbind fails, but a ypset (eventually, in the case of Solaris)
gets things happy.
  
  I'm sure other people use NetBSD NIS servers with other OS
clients, so I'm sure it is supposed to work!

  TIA

  Brian
-- 
"If you get 100 power engineers in a room, 101 of them will work this problem
 like this."  - R. P. Massey, ELEC 342