An unknwn error in an Orca computation

saeed_E · 2025-02-19 22:33:13

Dear Tian,
If possible, please let me state that I have frequently faced problems since upgrading my Orca 5.0.3 to 6.0.1; I have never encountered such problems with the older version.

Please let me ask a question regarding a new problem. The below Orca task was ran:
***********************************************************************************************************************
! ExtrapolateEP2(3/4,def2,MP2) tightSCF noautostart miniprint
%pal nprocs 8 end
%maxcore 6000
* xyz 0 1
C -3.00886523 -0.60110479 0.00000305
H -3.83275726 0.09558074 0.00003101
H -3.13625859 -1.67101162 -0.00011453
N -1.81764115 -0.14353739 0.00008421
H -0.98978097 -0.78273714 -0.00002296
O -1.48821106 1.09766974 0.00016013
C 1.46382272 -0.12319892 -0.00021950
O 0.71189603 -1.08884815 -0.00013887
C 2.95709267 -0.26671500 0.00015158
H 3.36553743 0.23708253 -0.87460156
H 3.36370189 0.22895946 0.88045529
H 3.23148130 -1.31574290 -0.00408829
O 1.05435101 1.11577850 -0.00024573
H 0.02497543 1.16194217 -0.00006346
*

$new_job
! ExtrapolateEP2(3/4,def2,MP2) tightSCF noautostart miniprint Pmodel
%pal nprocs 8 end
%maxcore 6000
* xyz 0 1
C -3.00886523 -0.60110479 0.00000305
H -3.83275726 0.09558074 0.00003101
H -3.13625859 -1.67101162 -0.00011453
N -1.81764115 -0.14353739 0.00008421
H -0.98978097 -0.78273714 -0.00002296
O -1.48821106 1.09766974 0.00016013
C : 1.46382272 -0.12319892 -0.00021950
O : 0.71189603 -1.08884815 -0.00013887
C : 2.95709267 -0.26671500 0.00015158
H : 3.36553743 0.23708253 -0.87460156
H : 3.36370189 0.22895946 0.88045529
H : 3.23148130 -1.31574290 -0.00408829
O : 1.05435101 1.11577850 -0.00024573
H : 0.02497543 1.16194217 -0.00006346
*

$new_job
! ExtrapolateEP2(3/4,def2,MP2) tightSCF noautostart miniprint Pmodel
%pal nprocs 8 end
%maxcore 6000
* xyz 0 1
C : -3.00886523 -0.60110479 0.00000305
H : -3.83275726 0.09558074 0.00003101
H : -3.13625859 -1.67101162 -0.00011453
N : -1.81764115 -0.14353739 0.00008421
H : -0.98978097 -0.78273714 -0.00002296
O : -1.48821106 1.09766974 0.00016013
C 1.46382272 -0.12319892 -0.00021950
O 0.71189603 -1.08884815 -0.00013887
C 2.95709267 -0.26671500 0.00015158
H 3.36553743 0.23708253 -0.87460156
H 3.36370189 0.22895946 0.88045529
H 3.23148130 -1.31574290 -0.00408829
O 1.05435101 1.11577850 -0.00024573
H 0.02497543 1.16194217 -0.00006346
*

$new_job
! ExtrapolateEP2(3/4,def2,MP2) tightSCF noautostart miniprint Pmodel
%pal nprocs 8 end
%maxcore 6000
* xyz 0 1
C -3.00886523 -0.60110479 0.00000305
H -3.83275726 0.09558074 0.00003101
H -3.13625859 -1.67101162 -0.00011453
N -1.81764115 -0.14353739 0.00008421
H -0.98978097 -0.78273714 -0.00002296
O -1.48821106 1.09766974 0.00016013
*

$new_job
! ExtrapolateEP2(3/4,def2,MP2) tightSCF noautostart miniprint Pmodel
%pal nprocs 8 end
%maxcore 6000
* xyz 0 1
C 1.46382272 -0.12319892 -0.00021950
O 0.71189603 -1.08884815 -0.00013887
C 2.95709267 -0.26671500 0.00015158
H 3.36553743 0.23708253 -0.87460156
H 3.36370189 0.22895946 0.88045529
H 3.23148130 -1.31574290 -0.00408829
O 1.05435101 1.11577850 -0.00024573
H 0.02497543 1.16194217 -0.00006346
*
******************************************************************************************************************************************************************
But after a long time, the task was terminated and the below message was displayed on the screen:

--------------------------------------------------------------------------
Primary job terminated normally, but 1 process returned
a non-zero exit code. Per user-direction, the job has been aborted.
--------------------------------------------------------------------------
--------------------------------------------------------------------------
mpirun noticed that process rank 2 with PID 0 on node saeed exited on signal 9 (Killed).
--------------------------------------------------------------------------
[file orca_tools/qcmsg.cpp, line 394]:
.... aborting the run
***************************************************************************************************************************************************************
Moreover, at the bottom of the output file, one can see:

Maximum memory used throughout the entire LEANSCF-calculation: 57.2 MB

************************************************************
* Program running with 8 parallel MPI-processes *
* working on a common directory *
************************************************************

------------------------------------------------------------------------------
ORCA MP2
------------------------------------------------------------------------------

Freezing NCore=14 chemical core electrons

----------
MP2 ENERGY (disk based algorithm)
----------

Dimension of the basis ... 609
Memory devoted to MP2 ... 6000 MB
Data format for buffers ... DOUBLE
Compression type for matrix containers ... UNCOMPRESSED

-------------------------
SHARK HALF TRANSFORMATION (Exchange order)
-------------------------

Number of basis functions ... 609
Number of operators ... 1
Operator 0: 7- 27 (p-index in (p*|q*))
Operator 0: 7- 27 (q-index in (p*|q*)
Integral generator used ... SHARK
Contraction scheme used ... SEGMENTED CONTRACTION
MaxCore in resort ... 6000 MB

Half transformed integrals for op= 0 ... R1_S.SHARK_MNPQ0.tmp
Resorted half transformed integrals ... R1_S.KAO_aa.tmp

ORCA finished by error termination in MP2
Calling Command: mpirun -np 8 /home/saeed/orca-6.0.1/orca_mp2_mpi R1_S.mp2inp.tmp R1_S
[file orca_tools/qcmsg.cpp, line 394]:
.... aborting the run

Please, if possible and like always, make me benefit of your highest kindness and let me know how I can resolve this problem.

In advance, many thanks for your valuable time and energy and, please excuse me for bothering you.

Sincerely,
Saeed

sobereva · 2025-02-20 00:01:35

If older version works normally, just use older version. I think some evident issues will be fixed in the future releases. By the way, I've heard from a developer that Extrapolate keyword will no longer be maintained, they prefer to use the compound script feature in ORCA to perform the extrapolation.

saeed_E · 2025-02-20 01:48:44

Dear Tian,

Thank you very much for your prompt reply with informative comments.

I also have seen such scripts in the Orca 6 tutorials (https://www.faccts.de/docs/orca/6.0/tut … rapol.html) but, at least for me, they are not very clear. And, I think "Extrapolate" keyword is really so nice and practically useful. If possible, please let me share with you two issues: 1) Can I install both older and newer versions together on the same system? Indeed, while the older version worked best for some keywords, the newer version provides some very nice features such as "conformed search".

2) Regarding the problem mentioned, I performed exactly the same calculation but on a system including fewer atoms, and, the task was terminated normally. Given that I used 8 processors any of which with 6000MB of ram, it seems that 8*6000=48GB exceeded the total AVAILABLE ram and, the task has been aborted in the first calculation. Do you agree with me?

Sincerely,
Saeed

Last edited by saeed_E (2025-02-20 01:51:49)

sobereva · 2025-02-20 05:15:56

1 You can. PS: However, I strongly suggest using my code Molclus (http://www.keinsci.com/research/molclus.html) to perform conformation search, usually it is the best and most flexible choice.

2 Perhaps the reason is too many memory was assigned.

saeed_E · 2025-02-20 06:51:38

Dear Tian,
Thank you very very much.
Interestingly, I completely understood your very valuable code, Molclus, and now I can work with this nice code perfectly.
It is also very surprising that the task is aborted if memory to be much more than is needed. Commonly, the low amount of memory encounters problem not the high amount!

Sincerely yours,
Saeed

Multiwfn forum

#1 2025-02-19 22:33:13

An unknwn error in an Orca computation

#2 2025-02-20 00:01:35

Re: An unknwn error in an Orca computation

#3 2025-02-20 01:48:44

Re: An unknwn error in an Orca computation

#4 2025-02-20 05:15:56

Re: An unknwn error in an Orca computation

#5 2025-02-20 06:51:38

Re: An unknwn error in an Orca computation

Board footer