[mpich-discuss] MPICH2 1.1a2 - problems with more than 4computers

[Jayesh Krishna]

Hi,
 From your reply it looks like your problem is the same as the one
described in ticket 355. Try upgrading to a server edition of windows.
 
(PS: We will include some code for retrying the network share conns -
however upgrading to a server edition is your soln.)
Regards,
Jayesh

  _____  

From: lccostajr at gmail.com [mailto:lccostajr at gmail.com] On Behalf Of Luiz
Carlos da Costa Junior
Sent: Tuesday, September 22, 2009 1:29 PM
To: Jayesh Krishna
Cc: mpich-discuss at mcs.anl.gov
Subject: Re: [mpich-discuss] MPICH2 1.1a2 - problems with more than
4computers


Hi Jayesh,

My configuration is very similar to the Tina's (user in this ticket).
I have 4 machines

Intel(R) Core(TM)2 Quad CPU
Q6600 @ 2.40GHz
8 GB of RAM

running

Microsoft Windows XP
Professional x64 Edition
Version 2003
Service Pack 2

over a Gigabit Network

I run mpiexec with the following command:

>> mpiexec.exe -configfile myconfigfile.txt

myconfigfile.txt:
-map j:\\computer1\paralell -n 1 -host computer1 -wdir j:\ j:\master 
-map j:\\computer1\paralell -n 4 -host computer1 -wdir j:\ j:\slave
-map j:\\computer1\paralell -n 4 -host computer2 -wdir j:\ j:\slave
-map j:\\computer1\paralell -n 4 -host computer3 -wdir j:\ j:\slave
-map j:\\computer1\paralell -n 4 -host computer4 -wdir j:\ j:\slave

And the error I get is:
*********** Warning ************
Unable to map \\computer1\paralell. (error 71)

*********** Warning ************
*********** Warning ************
Unable to map \\computer1\paralell. (error 71)

*********** Warning ************
launch failed: CreateProcess(slave) on 'computer2' failed, error 2 - The
syst
em cannot find the file specified.

launch failed: CreateProcess(slave) on 'computer3' failed, error 2
- The system cannot find the file specified.

*********** Warning ************
Unable to map \\computer1\paralell. (error 71)

*********** Warning ************
launch failed: CreateProcess(slave) on 'computer4' failed, error 2 -
 The system cannot find the file specified.

launch failed: CreateProcess(slave) on 'computer3' failed, error 2
- The system cannot find the file specified.

*********** Warning ************
Unable to map \\computer1\paralell. (error 71)

*********** Warning ************
launch failed: CreateProcess(slave) on 'computer2' failed, error 2 - The
syst
em cannot find the file specified.

launch failed: CreateProcess(slave) on 'computer3' failed, error 2
- The system cannot find the file specified.

*********** Warning ************
Unable to map \\computer1\paralell. (error 71)

*********** Warning ************
launch failed: CreateProcess(slave) on 'computer4' failed, error 2 -
 The system cannot find the file specified.

launch failed: CreateProcess(slave) on 'computer4' failed, error 2
- The system cannot find the file specified.

unable to read the cmd header on the left context, Error = -1.

Error posting readv, An existing connection was forcibly closed by the
remote ho
st.(10054)

If I can provide you any further information, please let me know.

Regards,
LC


2009/9/22 Jayesh Krishna <jayesh at mcs.anl.gov>


Hi,
 The error that the ticket refers to is related to limitations (number of
network shares) of using a client version of windows.
 Your problem may not be related to the ticket. Are you using network
shares ? What is the OS/ver that you use ? Please provide us as much
details as possible.
 
Regards,
Jayesh

  _____  

From: mpich-discuss-bounces at mcs.anl.gov
[mailto:mpich-discuss-bounces at mcs.anl.gov] On Behalf Of Luiz Carlos da
Costa Junior
Sent: Tuesday, September 22, 2009 9:56 AM
To: MPICH Discuss
Subject: Re: [mpich-discuss] MPICH2 1.1a2 - problems with more than
4computers


Dear all,

I got an error with MPICH2-1.1.1p1 and I found the following ticket that
describes exactly the error that I got:

http://trac.mcs.anl.gov/projects/mpich2/ticket/355

But I couldn't figure out which is the conclusion (or solution, if any)
for this error...
Could you please give me a clue?

Thanks,
LC



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/mpich-discuss/attachments/20090922/a89bbfa2/attachment.htm>