[Wien] A problem in a parallel execution of WIEN2k_8.1

[Peter Blaha]

The relevant output is in the following lines (with remotemachine):

...
if ( 0 > 1 ) echo running parallel lapw2
touch .lock_xps011
echo -n xps01
set ttt= ( `echo $mpirun | sed -e "s^_NP_^$number_per_job2[$loop]^" -e "s^_EXEC_^$WIENROOT/${exe}_mpi ${def}_$loop.def $loop^" -e "s^_HOSTS_^.machine$mach[$loop]^"` )
echo mpirun -np _NP_ -machinefile _HOSTS_ _EXEC_
sed -e s^_NP_^4^ -e s^_EXEC_^/usr/local/WIEN2k_08.1/lapw2_mpi lapw2_1.def 1^ -e s^_HOSTS_^.machine1^
if ( 1 == 1 ) then
set remote = rsh
remotemachine: Undefined variable.
else
...

If I compare this with the source of lapw2para_lapw, at first I cannot locate this in the
present version of lapw2para_lapw.

However, the problem is during installation due to siteconfig_lapw. siteconfig_lapw
changes the line:

	         set remotemachine = `head -1 .machine$mach[$loop]`
to
                 set remote = rsh

leading to your problems.
--------------------------------------------------
Fix:

copy the attached  siteconfig_lapw file into $WIENROOT
cd $WIENROOT
cp SRC/lapw2para_lapw $WIENROOT

either:
  edit lapw2para_lapw  and change   set remote = ssh     to  rsh   (only if you are using rsh)
or:
./siteconfig
   configure Parallel execution
       and select again the necessary options (shared memory y/n;  rsh or ssh;  mpi  y/n)
-----------

Since this is quite serious, I've updated the sources on the web (still keeping Version-Nr. WIEN2k_08.1)




> In addition, I did the -xf test. A result, I could get a message in
> STDOUT (attached in this mail).   
> But I don't know what I can do for this message.
> 
>>>>> I have a problem in a parallel execution of lapw2. It looks like lapw0
>>>>> and lapw1 were completed in parallel environment without any problem,
>>>>> but when it proceeds to lapw2, it aborts with an error message:
>>>>> --
>>>>> LAPW2 - FERMI; weighs written
>>>>> remotemachine: Undefined variable.
>>>>> remotemachine: Undefined variable.
>>>>> remotemachine: Undefined variable.
>>>>> remotemachine: Undefined variable.
>>>>> cp: cannot stat `.in.tmp': No such file or directory
>>>>> rm: cannot remove `.in.tmp': No such file or directory
>>>>> rm: cannot remove `.in.tmp1': No such file or directory
>>>>> --
-- 

                                      P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-15671             FAX: +43-1-58801-15698
Email: blaha at theochem.tuwien.ac.at    WWW: http://info.tuwien.ac.at/theochem/
--------------------------------------------------------------------------
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: siteconfig_lapw
Url: http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20080125/3a4617cb/siteconfig_lapw.ksh