Newsgroups: comp.parallel.pvm
From: heller@attila.imo.physik.uni-muenchen.de (Helmut Heller)
Reply-To: heller@nirvana.imo.physik.uni-muenchen.de
Subject: Re: PVM 3.3.4 won't start on **BLACK** NeXT!
Organization: Leibniz-Rechenzentrum, Muenchen (Germany)
Date: 29 Sep 1994 10:25:41 GMT
Message-ID: <36e4n5$2ua@sunserver.lrz-muenchen.de>

In article <3668tr$cvq@sunserver.lrz-muenchen.de>
heller@attila.imo.physik.uni-muenchen.de (Helmut Heller) writes:
| Hello,
| 
| * The exact version: 3.3.4
| 
| * The machine type(s) you're using, hardware and software: NeXTstation  
(black  
| hardware), NeXTSTEP3.2.
| 
| * The machine architecture PVM chooses for you: NEXT.
| 
| * A short description of the problem (what happens, when, etc.).
|   Include error messages (don't edit them to summarize).
|   Can't start pvm at all:
|   t2101ft 107 cpu| pvm
| libpvm [pid3614]: /tmp/pvmd.1413: No such file or directory
| libpvm [pid3614]: Console: Can't start pvmd
| 
| 
| * A record of what you did to make it happen.
| type pvm to the shell (tcsh).
| 
| t2101ft 108 cpu| echo $PVM_ARCH
| NEXT
| t2101ft 109 cpu| echo $PVM_ROOT/
| /Users/heller/pvm3/
| 
| both is correct!
| 
| I compiled the same source on a WHITE (intel-based) NEXT and there PVM runs  
| just fine.
| Any advice?
| Thanks,
| Helmut
| 

I got some response, but nothing helped so far. Here another piece of  
information:
when I try to start pvmd directly I get the following error:

t2101ft 128 cpu| pvmd
[pvmd pid11832] mksocs() bind netsock: Can't assign requested address
[pvmd pid11832] pvmbailout(0)


does this suggest what might be wrong here? 
Needless to say that previously pvm3.2.3 worked like a charm on that machine. 

Reading all these "Can't start pvm..." messages on the net from people who  
try pvm3.3, makes me think something might have been changed not for the  
better in the transition to 3.3, or what do you think??

Any advice welcome,
Helmut


