<html>
<body>
<font color="#000080"> Dear Prof. Peter Blaha and Jorissen
Kevin,<br><br>
First of all I would like to thank for your help.<br><br>
Regarding to what Prof. Peter Blaha mentioned: <br><br>
</font><i>> Do you have an mpi and sclapack installed ?<br><br>
</i><font color="#000080">Yes, we have.<br><br>
</font>> If yes, check again your compile.msg in eg.
SRC_lapw0<br><br>
<font color="#000080">We checked and no error was reported from
compile.msg file. <br><br>
For Jorissen Kevin:<br><br>
</font>>0/ Please provide us with the following information
:<br>
>* your MPI software<br><br>
<font color="#000080">we use mpich 125<br><br>
</font>>* your fortran compiler <br><br>
<font color="#000080">compiler = pathf90 (Fortran); pathcc (C)<br><br>
</font>> and compilation settings<br><br>
<font color="#000080">- 0 -freeform<br><br>
</font>>* the libraries you are using<br><br>
<font color="#000080">ATLAS: libatlas, libblas<br><br>
libscalapack (from netlib.org)<br><br>
</font> <br>
>2/ First you should check that the MPI commands are
passed on correctly by looking at the parallel scripts lapw1para etc (The
reason I >say this : using your second machines file, the program
should not be looking for executables lapw1, but for executables
lapw1_mpi !! >Adding the <font color="#000080">-x switch</font>
on the first line of lapw1para gives you extra information). <br>
<br><br>
<font color="#000080">We are trying to fix this. Could you explain more
detail on this? Why can it pass lapw0_mpi but not lapw1_mpi?
Thanks.<br><br>
</font>>granularity:1<br>
>18:opto024<br>
>18:opto025<br>
>36:opto030<br><br>
<font color="#000080">cannot run: lapw1 crashed in the first cycle.
<br><br>
We did notice to what Prof. Blaha mentioned on the efficiency of
mpi on dual node. The problem is not it will speed up the calculation or
saves memory but when we run one wien2k job it will occupy all most >
99% CPU on one node. If in the same time someone run also wien2k, this
will slow down the whole system. No one else can use host anymore :(
because, in this case, 2 or 3 nodes will be occupied with more than 99%
CPU. Due to this we would like to run wien2k in parallel mode.
<br><br>
We notice that we can run wien2k in parallel mode under the localhost,
i.e., <br><br>
</font>>granularity:1<br>
>18:localhost<br>
>18:localhost<br>
>36:localhost<br><br>
<font color="#000080">Everything is perfect with this .machines file
(very fast, save a lot of time ) <b>but</b> only one person can use the
system and no one else can use it :).<br><br>
We are still looking for help on this. Many thanks in advance.<br><br>
Have a nice weekend.<br><br>
Regards,<br>
Khuong<br>
</font></body>
</html>