[comp.protocols.nfs] PCNFS 3.0.1.a hangs

adam@metaware.metaware.com (At these prices, I can't NAME names) (12/12/89)

	We are running PC-NFS 3.0.1.a on some IBM/IBM compatibles.
We are using 3com 3c505 cards on thin ethernet. PCNFS is running
with a 3com network coresident.

	My problem is this. I have installed everything. 3com works
great. Telnet/PC-NFS works great as well. However, if I do not log into
one of our unix machines (and stay logged in) within 10 minutes of booting
my machine, PCNFS will hang if I later try to invoke it. This leads to lots
of reboots and irritation for all.

	Does this sound familiar? Is there a fix, have I misinstalled
something?

any and all help appreciated...
-- 
adam margulies                                    metaware incorporated
                                                  INTERNET: adam@metaware.com
                                                  UUCP:     uunet!metaware!adam
                                                  ATT:      (408)429-META x3016

bryan@intvax.UUCP (Jon R Bryan) (12/12/89)

From article <159@metaware.metaware.com>, by adam@metaware.metaware.com (At these prices, I can't NAME names):
> 
> 	My problem is this. I have installed everything. 3com works
> great. Telnet/PC-NFS works great as well. However, if I do not log into
> one of our unix machines (and stay logged in) within 10 minutes of booting
> my machine, PCNFS will hang if I later try to invoke it. This leads to lots
> of reboots and irritation for all.
> 
Aha!  So it wasn't my imagination.  I'm running 3.0 and have had the
same thing happen.  It's not consistent, but it does occur.  Also, in my
case Telnet will run from Desqview on a 33Mhz '386 clone but not on a
Compaq 386/20 or a 16Mhz Intel Inboard 386 in an old XT.  Desqview has
trapped some illegal instructions a couple of times on my machine, at
start-up, but that's not consistent either.  Interestingly, Telnet will
run from Windows on the Compaq.  From either Desqview or Windows,
pushing Telnet to the background eventually locks both machines up.

-- 
Jon R. Bryan	<=>	bryan@intvax.UUCP
Sandia National Laboratories
Intelligent Machine Principles Division
Albuquerque, New Mexico

geoff@hinode.East.Sun.COM (Geoff Arnold @ Sun BOS - R.H. coast near the top) (12/13/89)

Quoth bryan@intvax.UUCP (Jon R Bryan) in <2070@intvax.UUCP>:
#From article <159@metaware.metaware.com>, by adam@metaware.metaware.com (At these prices, I can't NAME names):
#> 
#> 	My problem is this. I have installed everything. 3com works
#> great. Telnet/PC-NFS works great as well. However, if I do not log into
#> one of our unix machines (and stay logged in) within 10 minutes of booting
#> my machine, PCNFS will hang if I later try to invoke it. This leads to lots
#> of reboots and irritation for all.
#> 
#Aha!  So it wasn't my imagination.  I'm running 3.0 and have had the
#same thing happen.  It's not consistent, but it does occur.  Also, in my
#case Telnet will run from Desqview on a 33Mhz '386 clone but not on a
#Compaq 386/20 or a 16Mhz Intel Inboard 386 in an old XT.  Desqview has
#trapped some illegal instructions a couple of times on my machine, at
#start-up, but that's not consistent either.  Interestingly, Telnet will
#run from Windows on the Compaq.  From either Desqview or Windows,
#pushing Telnet to the background eventually locks both machines up.

Operation of PC-NFS toolkit applications (including telnet) within
window systems of any kind is problematic to say the least. In the
case of MS Windows, this is due to the fact that there seems to be no way
to lock an application into memory, which makes it impossible for
lower-level code to upcall or post an event to the application with any
certainty that Windows won't have moved or remapped stuff. 

A number of Toolkit customers have implemented Windows applications
in conjunction with the Toolkit, so it IS possible. However these
are dedicated Windows applications, not bimodal (DOS or Windows)
apps.

This whole area is the subject of continuing study within the PC-NFS
group. For now, however, we are unable to offer a satisfactory
solution. Disclaimers, etc.

[I've come to the conclusion that in many respects it's best to
treat Windows as a distinct OS, along with DOS, OS/2 and Unix. It
happens to share some lower level elements with DOS (which is good,
because we can at least provide file and printer sharing with PC-NFS),
but in other respects it is a quite seperate beast. If the gossip
in comp.sys.ibm.pc and PC Week are true, it seems likely that DOS and
Windows will continue to diverge over the next couple of years.]


Geoff Arnold, PCDS Group,     | Quote of the week: "Shut up and mind your
Sun Microsystems Inc.         | Canadian business, you meddlesome foreigner."
Internet: geoff@East.Sun.COM  | (Theodore A.Kaldis, <kaldis@topaz.rutgers.edu>
Disclaimer: Obviously....     | on Nov. 22 1989, in reply to Joe Nunes.)

henk@cs.eur.nl (Henk Langeveld) (12/13/89)

adam@metaware.metaware.com (At these prices, I can't NAME names) writes:

>	My problem is this. I have installed everything. 3com works
>great. Telnet/PC-NFS works great as well. However, if I do not log into
>one of our unix machines (and stay logged in) within 10 minutes of booting
>my machine, PCNFS will hang if I later try to invoke it. This leads to lots
>of reboots and irritation for all.

We've encountered similar problems with MKS toolkit on PC-NFS machines.
The problem manifests itself only when we use MKS' setup with 
SHELL=c:\etc\init.exe  in config.sys.   I ran etherfind on the server
to find out what was happening on the net, and the pc does a call to
nfs RFS_LOOKUP, after which the server issues an ip arp request for
the pc.

The arp request is strange, as the only authority on our net concerning
host names is the server itself.

-- 
Henk Langeveld, Unix SysAdmin	| domain: <henk@cs.eur.nl>
Department of Computer Science	| phone:  +31 10 4081346
Erasmus University Rotterdam	| also:   langeveld@hroeur5.bitnet
Room H5-05, P.O.Box 1738, NL-3000 DR  Rotterdam, The Netherlands.