[Wien] Error in lapw2 running in the LDA+U calculation in parallel mode

[hassan jamnejad]

Dear wien users,

thank for your attention, based on your advise I checked “STDOUT“ and
“parallel” files . Below is the details .

*STDOUT*

.

.

.

STOP  LAPW1 END

real        0m5.669s

user       0m16.792s

sys          0m0.992s

STOP LAPW2 - FERMI; weighs written

STOP STOP  LAPW2 END

LAPW2 END

STOP  LAPW2 END

STOP  LAPW2 END

real        0m1.812s

user       0m2.188s

sys          0m0.432s

STOP  LAPW2 END

STOP  LAPW2 END

STOP  LAPW2 END

STOP  LAPW2 END

real        0m1.782s

user       0m2.164s

sys          0m0.448s

At line 332 of file sumpara.f (unit = 10, file = 'LaXX.dmatup_1')

Fortran runtime error: Bad integer for item 1 in list input

cp: cannot stat ‘.in.tmp’: No such file or directory

>;   stop error


*:parallel *

.

.

starting parallel lapw1 at Sat Jan  2 12:24:24 IRST 2016

vHost3 vHost3 vHost3 vHost3(4)      vHost4 vHost4 vHost4 vHost4(4)
Summary of lapw1para:


vHost3                k=0         user=0
wallclock=240


vHost4                k=0         user=0
wallclock=240

<;-  done at Sat Jan 2 12:24:36 IRST 2016

-----------------------------------------------------------------

starting parallel lapw1 at Sat Jan  2 12:24:38 IRST 2016

vHost3 vHost3 vHost3 vHost3(4)      vHost4 vHost4 vHost4 vHost4(4)
Summary of lapw1para:


vHost3                k=0         user=0
wallclock=240


vHost4                k=0         user=0
wallclock=240

<;-  done at Sat Jan 2 12:24:50 IRST 2016

-----------------------------------------------------------------

->;  starting Fermi on wien2kServer at Sat Jan 2 12:24:54 IRST 2016

->;  starting parallel lapw2 at Sat Jan 2 12:24:54 IRST 2016

      vHost3

      vHost4


Summary of lapw2para:


vHost3                user=0  wallclock=0


vHost4                user=0  wallclock=0

<;-  done at Sat Jan 2 12:24:58 IRST 2016

->;  starting sumpara 2 on wien2kServer at Sat Jan 2 12:24:59 IRST 2016

**  LAPW2 crashed at Sat Jan 2 12:25:00 IRST 2016

**  check ERROR FILES!



I will be glad to know your suggestion


thanks
H. Jamnejad


*Gavin Abo* gsabo at crimson.ua.edu
<wien%40zeus.theochem.tuwien.ac.at?Subject=Re%3A%20%5BWien%5D%20Error%20in%20lapw2%20running%20in%20the%20LDA%2BU%20calculation%20in%0A%20parallel%20mode&In-Reply-To=%3C568708A7.8010405%40crimson.ua.edu%3E>
*Sat Jan 2 00:15:51 CET 2016*

Since the crash happens in lapw2, in the lapw2para_lapw script file, it
looks like messages might be written to the .time2_*, .stdout2_*, or
:*parallel files depending on the settings in your parallel_options and
*.machines file.  So you could try checking for error messages in these
files with something like:

cat .time2_*
cat .stdout2_*
cat :parallel

You might also have to run the last command given in the :log file by
itself or look into the input/output files for the program (refer to
'Table 4.3: Input and output files of main programs in an SCF cycle' in
the WIEN2k usersguide) that crashed to determine the cause of the error
[ http://zeus.theochem.tuwien.ac.at/pipermail/wien/2008-December/011904.html
].

Based on reports of the "Error in SUMPARA" before [
http://zeus.theochem.tuwien.ac.at/pipermail/wien/2007-June/009291.html
, https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg05637.html
], it might happen if a case.scf2* file cannot be found.

---------- Forwarded message ----------
From: hassan jamnejad <hassan.jamnejad at gmail.com>
Date: Fri, Jan 1, 2016 at 11:44 PM
Subject: Error in lapw2 running in the LDA+U calculation in parallel mode
To: wien at zeus.theochem.tuwien.ac.at


Dear Prof. Peter Blaha,

I want to do a LDA+U calculation for a compound, First I ran SCF
calculation using 4 x 8 (= 32 processors) in my cluster . But the job
terminates with the following errors

in dayfile:

***  LAPW2 crashed!*

*1.6u 3.1s 0:07.72 61.7% 0+0k 0+1776io 0pf+0w*

*error: command   /home/pars/wien2k/13.1/lapw2para -up uplapw2.def   failed*

in Lapw2.error file:
***  testerror: Error in Parallel LAPW2*

in upsumpara.error file:



*Error in SUMPARA *
in the same time if I run LDA+U calculation in serial mode the  SCF
calculation was run successfully. Also when  I just run LDA calculation in
parallel mode the  SCF  calculation was run successfully too. This means
neither LDA+U calculation nor parallel calculation have problem. Meanwhile
I searched in the mail list. I found some similar errors, but I could not
find my answer. Please suggest me how to rectify above errors?

Thanks in advance

Regards
H. Jamnejad
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20160102/365afdea/attachment.html>