Page 2 of 5

Re: yambo install

Posted: Fri Apr 22, 2016 8:54 am
by andrea.ferretti
you are welcome!
thanks for reporting
Andrea

Re: README first

Posted: Thu Sep 08, 2016 7:19 am
by joydev de
Dear forum,

I have installed yambo-4.0.4-rev.107. I have done chi calculation for small molecules in serial mode. But when i want to do the same calculation in parallel mode the plot able file not create.
What is the regain i don't understand. Please help me any of you. If you want, I can share my cluster computer path and user ID.

regards
joydev de
Ph.D student
NISER
INDAI

Re: yambo install

Posted: Thu Sep 08, 2016 9:08 am
by Daniele Varsano
Dear Joydev,

It is impossible to understand what is the problem with the information you provided. Have you compiled the code using parallel compiler libraries (MPI)?

Please post the error message you receive if any, and the config.log file you get when configuring the compilation.
Best,

Daniele

Re: yambo install

Posted: Thu Sep 08, 2016 9:55 am
by joydev de
Dear sir,

I attach the config.log file. My actual problem is that when i run the yambo in parallel mode using job script, the output file is not created.
There is no error message. But when i run it in serial mode the out put file created.

Regards
joydev de
Ph.D student
NISER,BBSR
INDIA

Re: yambo install

Posted: Thu Sep 08, 2016 10:15 am
by andrea.ferretti
Dear Joydev,

you config.log file seems to be ok (MPI env correctly detected etc).
Most likely the problem is either in your run script, or in the interaction with the local MPI environment.
Where you bale to run other programs in parallel (eg. quantum espresso) ?
Could you also provide the batch script you are using ?
Does yambo write anything in the report file (r-*) or in the logs file
(in parallel logs are written in LOG/l-* files)

Andrea

Re: yambo install

Posted: Thu Sep 08, 2016 10:46 am
by joydev de
Dear Andear,

Thank you.

Here i give my job script.
Using this job script, i have done my quantum espresso calculation in parallel mode but here only the path and input file are different.
job.sh:-
#!/bin/bash
#$ -N sih4-yambo
#$ -S /bin/bash
#$ -cwd
#$ -q all.q
#$ -e $JOB_ID.$JOB_NAME.error
#$ -o $JOB_ID.$JOB_NAME.output
#$ -pe orte 16
#/opt/intel/openmpi-intel/bin/mpirun -np $NSLOTS --mca mtl psm /root/sanjiv/espresso-5.1_cluster/bin/pw.x -inp < relax.in > relax.out
#/opt/intel/openmpi-intel/bin/mpirun -np $NSLOTS --mca mtl psm /root/sanjiv/espresso-5.1_cluster/bin/pw.x -inp < sih4_scf.in > sih4_scf.out
/opt/intel/openmpi-intel/bin/mpirun -np 16 /home/joydev/yambo-4.0.4-rev.107/bin/yambo < yambo.in > yambo.out
#/opt/intel/openmpi-intel/bin/mpirun -np $NSLOTS --mca mtl psm /root/sanjiv/espresso-5.1_cluster/bin/dos.x -inp < dos.in > dos.out
#/opt/intel/openmpi-intel/bin/mpirun -np $NSLOTS --mca mtl psm /root/sanjiv/espresso-5.1_cluster/bin/projwfc.x -inp < pdos.in > pdos.out

When i am running it in parallel mode, the LOG file is created.
But in the out file ,there is no data.
My yambo.in file is given by-
yambo.in:
# __ __ ______ __ __ _______ ______
# | \ / \ / \ | \ / \| \ / \
# \$$\ / $$| $$$$$$\| $$\ / $$| $$$$$$$\| $$$$$$\
# \$$\/ $$ | $$__| $$| $$$\ / $$$| $$__/ $$| $$ | $$
# \$$ $$ | $$ $$| $$$$\ $$$$| $$ $$| $$ | $$
# \$$$$ | $$$$$$$$| $$\$$ $$ $$| $$$$$$$\| $$ | $$
# | $$ | $$ | $$| $$ \$$$| $$| $$__/ $$| $$__/ $$
# | $$ | $$ | $$| $$ \$ | $$| $$ $$ \$$ $$
# \$$ \$$ \$$ \$$ \$$ \$$$$$$$ \$$$$$$
#
# GPL Version 4.0.4 Revision 107
# MPI Build
# http://www.yambo-code.org
#
optics # [R OPT] Optics
chi # [R CHI] Dyson equation for Chi.
tddft # [R K] Use TDDFT kernel
Chimod= "ALDA" # [X] IP/Hartree/ALDA/LRC/BSfxc
X_all_q_CPU= "2 2 2 2" # [PARALLEL] CPUs for each role
X_all_q_ROLEs= "q k c v" # [PARALLEL] CPUs roles (q,k,c,v)
FxcGRLc= 1 RL # [TDDFT] XC-kernel RL size
NGsBlkXd= 1 RL # [Xd] Response block size
% QpntsRXd
1 | 1 | # [Xd] Transferred momenta
%
% BndsRnXd
1 | 40 | # [Xd] Polarization function bands
%
% EnRngeXd
5.19315 | 17.82734 | eV # [Xd] Energy range
%
% DmRngeXd
0.10000 | 0.10000 | eV # [Xd] Damping range
%
ETStpsXd= 100 # [Xd] Total Energy steps
% LongDrXd
0.1000E-4 | 0.000 | 0.000 | # [Xd] [cc] Electric Field
%
~

Regards
joydev

Re: yambo install

Posted: Thu Sep 08, 2016 10:52 am
by andrea.ferretti
please note that yambo needs to be run, e.g., as

mpirun -np 16 /path-to-yambo/bin/yambo -F yambo.in

and this will automatically generate a report file (r-*) and the log files

andrea

Re: yambo install

Posted: Thu Sep 08, 2016 10:54 am
by Daniele Varsano
Dear Joydev,
can you also post your report and one of the LOG files?

Daniele

Re: yambo install

Posted: Thu Sep 08, 2016 11:40 am
by joydev de
Dear Daniele,

Using /opt/intel/openmpi-intel/bin/mpirun -np 16 /home/joydev/yambo-4.0.4-rev.107/bin/yambo -F yambo.in, the log file and r_optics_chi_tddft is created. But
o.eps_q1_inv_alda_dyson and o.eel_q1_inv_alda_dyson files are not created. But these last two file are created when i run it in serial mode.
This two file give me the electron energy loss and electron absorption. But in parallel mod this file are not created. I attach one of my log file below.

Thank you

Regards
joydev

Re: yambo install

Posted: Thu Sep 08, 2016 11:54 am
by andrea.ferretti
Hi Joydev,

have you inspected the report and log files, searching for errors ? (if the code stops for some reason it should be reported there).
Have you set the variables related to parallelism (e.g. X_q_0_CPU and X_q_0_ROLEs) ?
the code tries to guess reasonable values for them but it is not always able to do so...

in order go get them in the automatically generated input file, I think you need to specify -V parallel or -V all

Andrea

BTW: the previous message didn't have attached files