Welcome, Guest
Username: Password: Remember me
  • Page:
  • 1
  • 2

TOPIC: mpi problem when running telemac in parallel

mpi problem when running telemac in parallel 11 years 4 days ago #11066

  • yugi
  • yugi's Avatar
  • OFFLINE
  • openTELEMAC Guru
  • Posts: 851
  • Thank you received: 244
Try launching runcode.py with the option --ncsize=number of precessor.
Could you post the whole listing with the command ?
There are 10 types of people in the world: those who understand binary, and those who don't.
The administrator has disabled public write access.

mpi problem when running telemac in parallel 11 years 4 days ago #11067

  • Willvan
  • Willvan's Avatar
tried it, no succes
Attachments:
The administrator has disabled public write access.

mpi problem when running telemac in parallel 10 years 10 months ago #11504

  • Willvan
  • Willvan's Avatar
hi,
the problem with running parallel seems fixed, had some double libraries in het systel config file :
#libs_parallel: /home/vsantewi/metis-5.1.0/build/Linux-x86_64/libmetis/libmetis.a
libs_partel: /home/vsantewi/metis-5.1.0/build/Linux-x86_64/libmetis/libmetis.a

i put the libs_parallel in comment and that seemed to do the trick. Still 1 problem remaining :

LOGICAL UNIT 3 ERROR LINE 305
LOGICAL UNIT 3 ERROR LINE 305
LOGICAL UNIT 3 ERROR LINE 306
LOGICAL UNIT 3 ERROR LINE 306
LOGICAL UNIT 3 ERROR LINE 307
LECDON_TELEMAC2D: ERROR FOR FILE NUMBER
I= 1 NAME=
THIS FILE SHOULD HAVE A STRING SUBMIT
IN DICTIONARY

Somebody has an idea?
thx.
The administrator has disabled public write access.

mpi problem when running telemac in parallel 10 years 10 months ago #11505

  • Willvan
  • Willvan's Avatar
additional info:

runcode::main:
/home/vsantewi/telemac/v6p3r1/examples/telemac2d/seiche:
|runCode: Fail to run
|/opt/openmpi/gcc/64/1.6.3/bin/mpiexec -n 4 /home/vsantewi/telemac/v6p3r1/examples/telemac2d/seiche/t2d_seiche.cas_
2014-01-09-14h35min07s/out_t2d_seiche
|~~~~~~~~~~~~~~~~~~
|STOP 2
|STOP 2
|STOP 2
|
|mpiexec has exited due to process rank 0 with PID 30067 on
|node eul05 exiting improperly. There are two reasons this could occur:
|
|1. this process did not call "init" before exiting, but others in
|the job did. This can cause a job to hang indefinitely while it waits
|for all processes to call "init". By rule, if one process calls "init",
|then ALL processes must call "init" prior to termination.
|
|2. this process called "init", but exited without calling "finalize".
|By rule, all processes that call "init" MUST call "finalize" prior to
|exiting or it will be considered an "abnormal termination"
|
|This may have caused other processes in the application to be
|terminated by signals sent by mpiexec (as reported here).
|
The administrator has disabled public write access.

mpi problem when running telemac in parallel 10 years 10 months ago #11511

  • yugi
  • yugi's Avatar
  • OFFLINE
  • openTELEMAC Guru
  • Posts: 851
  • Thank you received: 244
In your configuration file it there is somthing wrong with the value for option.
It should be "parallel mpi" and not "parallel openmpi".

That could be the source of the error change it and try to rerun.
There are 10 types of people in the world: those who understand binary, and those who don't.
The administrator has disabled public write access.

mpi problem when running telemac in parallel 10 years 10 months ago #11513

  • Willvan
  • Willvan's Avatar

File Attachment:

File Name: h41f1243.txt
File Size: 4 KB

tried the mpi openmpi switch to no avail.

but on advice of a collegea i changed the gfortran to mpif90 and recompiled the complete installation but in the compile info i still see messages like :

COLLECT_GCC=/cm/shared/apps/gcc/4.7.0/bin/gfortran

allthough i have no reference to gfortran in the config file.

i will attach the config file
The administrator has disabled public write access.

mpi problem when running telemac in parallel 10 years 10 months ago #11534

  • yugi
  • yugi's Avatar
  • OFFLINE
  • openTELEMAC Guru
  • Posts: 851
  • Thank you received: 244
You do need to replace openmpi by mpi in your systel.cfg because that is an error.

I suggest to recompile with the option --clean.

You should also chanage the value of options in general because here you have "options: parallel openmpi" for every configurations even the serials one.
You should set it to nothing by default as you are changing it in every parallel configuration.

As for the fact that gfortran appears that is normal mpif90 is a wrapper on gfortran.
FYI if you type mpif90 -show (Or something like that) you will see what is actually called.

Hope it helps.
There are 10 types of people in the world: those who understand binary, and those who don't.
The administrator has disabled public write access.
The following user(s) said Thank You: iknponza
  • Page:
  • 1
  • 2
Moderators: borisb

The open TELEMAC-MASCARET template for Joomla!2.5, the HTML 4 version.