 |
|
List Archives >
Maple User Group List Archive >
Archive by date >
This Month By Date >
This Month By Topic
[MUG] maple and bproc
| [MUG] maple and bproc |
|
Author:
Posted: Mon, 21 Oct 2002 15:48:25 -0400
|
>> From: "jdwright"
Dear MUG,
I am trying to distribute a maple calculation across a beowulf cluster of
50 processors running Linux (www.linux.org) and the Scyld distribution kernel
2.2.17 (www.scyld.com).
The Scyld distribution is built for running a beowulf cluster. It is
based on a process migration package called bproc. bproc is a process
migration system. It can start processes on the master and then
migrate them to a node. To do so, it starts a SUBSHELL on a node,
meaning that all properties of the shell are taken from the current
shell (current directory, path, environment variables, ...). After
migration the process is represented by a ghost on the master. This
differs radically from a system using rsh or telnet to log into a node
and then run a command. For our beowulf cluster bproc is the only way
to run programs.
The most important command is bpsh which runs commands on a node in
the cluster. For example, the following command runs the program ls
on node 5.
bpsh 5 ls
A similar type of command should run maple on node 5 of the cluster.
bpsh 5 /usr/maple8/bin/maple
however I receive the following error.
/usr/maple8/bin/maple: /usr/maple8/bin/maple: No such file or directory
The system administrator for the cluster says that maple cannot run
because maple calls the program cmaple. Then we try to bypass the
maple script with the following command.
bpsh 5 /usr/maple8/bin.IBM_INTEL_LINUX/cmaple
We make progress and get a new error.
mserver: error in loading shared libraries: libmaple.so: cannot open
shared object file: No such file or directory
I need help to figure out how to make this work. Any relevant
information would be helpful. Ideally, someone out there has already
dealt with this problem using similar tools to our implementation.
Short of this ideal case, I have some more specific questions. Is
there a way to bypass the maple shell script "maple" and send commands
directly to the maple kernel? What is the directory and filename for
the maple kernel? Is the maple kernel called cmaple? Is the maple
kernel called mserver? Does cmaple need to run mserver? Does mserver
need cmaple in order to run? Can maple run multiple instances of
itself on a single machine? Does maple support dual processor machines
let alone a 50 processor beowulf?
Any helpful information would be greatly appreciated.
Thanks,
David Wright
Physics Department
Wesleyan University
Middletown, CT 06459
"jdwright"
|
| [MUG] Re: maple and bproc |
|
Author: Douglas B Meade
Posted: Wed, 23 Oct 2002 17:39:33 -0400
|
>> From: "Douglas B. Meade" "meade"
MUG and David,
I am in the process of putting together a collection of Maple tools to
run Maple jobs in parallel. At present I am using the Sockets package
in Maple to open and manage multiple copies of Maple on distributed
processors. I have a prototype that operates but have not had time to
work on this lately.
If you are interested, I have placed the PPT file for my presentation
at last summer's Maple Summer Workshop on my webpage. Please follow
the link to Conferences, Meetings, and Seminars.
Let me know if you have any questions or further interest.
Doug
-----------------------------------------------------------------------
Prof. Douglas B. Meade Phone: (803) 777-6183 FAX: (803)
777-6527
Department of Mathematics URL: http://www.math.sc.edu/~meade/
USC, Columbia, SC 29208 E-mail: "mailto:meade"
|
Previous by date: [MUG] Re: Crashing in Maple 7 for Linux when removing output, Maple User Group
Next by date: [MUG] Re: Bug VectorCalculus[Jacobian], Maple User Group
Previous thread: [MUG] bug in Map ?, Jean Brillet
Next thread: [MUG] Bug VectorCalculus[Jacobian], Jarausch
|
|
|
(NAG)
