[CP2K-user] [CP2K:17115] CPASSERT failed when trying to restart AIMD with periodic efield

Matt Watkins mattwatkinsuk at gmail.com
Fri Jun 10 08:22:57 UTC 2022
Previous message (by thread): [CP2K-user] [CP2K:17113] CPASSERT failed when trying to restart AIMD with periodic efield
Next message (by thread): [CP2K-user] [CP2K:17116] CPASSERT failed when trying to restart AIMD with periodic efield
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
It could be directly related to CP2K but looks suspiciously like a memory 
leak in the MPI stack. 
Just checking - it restarted OK and ran several steps before that error?
If there are no better ideas suggested, I'd check if there was a newer 
executable available and check reproducibility / report to your helpdesk.
Matt
On Friday, 10 June 2022 at 09:02:02 UTC+1 rajors... at gmail.com wrote:

> Hi,
>
> I face another problem while trying to restart calculations with periodic 
> electric fields. After running for some steps, I get the following error :
>
> [1654828609.897841] [jwc01n044:28373:0]          ib_md.c:329  UCX  ERROR 
> ibv_reg_mr(address=0x134e71de0, length=5664, access=0xf) failed: Cannot 
> allocate memory
> [1654828609.897897] [jwc01n044:28373:0]         ucp_mm.c:137  UCX  ERROR 
> failed to register address 0x134e71de0 mem_type bit 0x1 length 5664 on 
> md[5]=mlx5_0: Input/output error (md reg_mem_types 0x15)
> [1654828609.897902] [jwc01n044:28373:0]    ucp_request.c:280  UCX  ERROR 
> failed to register user buffer datatype 0x8 address 0x134e71de0 len 5664: 
> Input/output error
>
>
>
>  *******************************************************************************
>  *   ___                                                                   
>     *
>  *  /   \                                                                 
>      *
>  * [ABORT]                                                                 
>     *
>  *  \___/      MPI error 17 in mpi_waitall @ mp_waitall_1 : See the 
> MPI_ERROR  *
>  *    |                   field in MPI_Status for the error code           
>     *
>
>  *  O/|                                                                   
>      *
>  * /| |                                                                   
>      *
>  * / \                                                   
>  dbcsr_mpiwrap.F:1022 *
>
>  *******************************************************************************
>
> But when I restart the calculation again from this point, it runs well for 
> a few steps and again gives the same error. Also, I have not noticed this 
> error in simulations which do not involve periodic electric fields. I run 
> my calculations in JUWELS HPC with CP2K version 7.1. I have attached the 
> input file for your reference.
>
> Thank you in advance.
>
> On Fri, 27 May 2022 at 12:49, Rajorshi Chattopadhyay <rajors... at gmail.com> 
> wrote:
>
>> Hi,
>> Thanks a lot for your suggestion. Making &RESTART_RTP FALSE under 
>> &EXT_HISTORY solved the problem.
>>
>> On Thu, 26 May 2022 at 12:27, Jürg Hutter <hut... at chem.uzh.ch> wrote:
>>
>>> Hi
>>>
>>> this is the reason why your job stops:
>>>
>>>          ! periodic fields don't work with RTP
>>>          CPASSERT(.NOT. do_rtp)
>>>
>>> If you don't activate RTP, please send your input to further investigate.
>>> If you want to use RTP, you can have periodic efields with the current 
>>> version of
>>> the code.
>>>
>>> regards
>>>
>>> Juerg Hutter
>>>
>>> ________________________________________
>>> From: cp... at googlegroups.com <cp... at googlegroups.com> on behalf of 
>>> Rajorshi Chattopadhyay <rajors... at gmail.com>
>>> Sent: Wednesday, May 25, 2022 3:05 PM
>>> To: cp... at googlegroups.com
>>> Subject: Re: [CP2K:17034] CPASSERT failed when trying to restart AIMD 
>>> with periodic efield
>>>
>>> Hello,
>>> Sorry for missing out on this important information. Here is the full 
>>> error:
>>>
>>>
>>> *******************************************************************************
>>>  *   ___                                                                
>>>        *
>>>  *  /   \                                                                
>>>       *
>>>  * [ABORT]                                                              
>>>        *
>>>  *  \___/                             CPASSERT failed                    
>>>       *
>>>  *    |                                                                  
>>>       *
>>>  *  O/|                                                                  
>>>       *
>>>  * /| |                                                                  
>>>       *
>>>  * / \                                                  
>>> cp_control_utils.F:431 *
>>>
>>>  *******************************************************************************
>>>
>>>
>>>  ===== Routine Calling Stack =====
>>>
>>>             2 qs_init_subsys
>>>             1 CP2K
>>>
>>> On Wed, 25 May 2022 at 14:04, Jürg Hutter <hut... at chem.uzh.ch<mailto:
>>> hut... at chem.uzh.ch>> wrote:
>>> Hi
>>>
>>> there should be important information printed together with the CPASSERT 
>>> failure.
>>> Can you provide this, so we get a hint on the source of the problem?
>>>
>>> regards
>>>
>>> Juerg Hutter
>>>
>>> ________________________________________
>>> From: cp... at googlegroups.com<mailto:cp... at googlegroups.com> <
>>> cp... at googlegroups.com<mailto:cp... at googlegroups.com>> on behalf of 
>>> Rajorshi Chattopadhyay <rajors... at gmail.com<mailto:rajors... at gmail.com>>
>>> Sent: Wednesday, May 25, 2022 10:58 AM
>>> To: cp... at googlegroups.com<mailto:cp... at googlegroups.com>
>>> Subject: [CP2K:17027] CPASSERT failed when trying to restart AIMD with 
>>> periodic efield
>>>
>>> Hello all,
>>>
>>> I am trying to run an AIMD simulation to calculate wannier centers in 
>>> the presence of an external electric field (cp2k version 9.1.0). I 
>>> equilibrate the system (in the presence of the efield) and then use the 
>>> .restart and .wfn files to run a production simulation. However, during the 
>>> production run, I encounter the 'CPASSERT failed' error. Here are the 
>>> important sections of the input file for production simulation:
>>>
>>> &EXT_RESTART
>>>   RESTART_FILE_NAME Water_amb_Bulk_efield-1.restart
>>> &END EXT_RESTART
>>>
>>> &DFT
>>>       BASIS_SET_FILE_NAME ./BASIS_MOLOPT
>>>       POTENTIAL_FILE_NAME ./GTH_POTENTIALS
>>>   &MGRID
>>>       CUTOFF 280
>>>       REL_CUTOFF 40
>>>       NGRIDS 5
>>>   &END MGRID
>>>   &SCF
>>>      SCF_GUESS RESTART
>>>      MAX_SCF 1000
>>>      &OT
>>>         MINIMIZER DIIS
>>>         PRECONDITIONER FULL_SINGLE_INVERSE
>>>      &END OT
>>>   &END SCF
>>>
>>>   &PERIODIC_EFIELD
>>>      INTENSITY 5.0E-3
>>>      POLARISATION 1.0 0.0 0.0
>>>   &END PERIODIC_EFIELD
>>>   &LOCALIZE
>>>      METHOD CRAZY
>>>      EPS_LOCALIZATION 1.0E-8
>>>      MAX_ITER 2000
>>> &PRINT
>>>         &MOLECULAR_DIPOLES
>>>          PERIODIC TRUE
>>>          &EACH
>>>           MD 2
>>>          &END EACH
>>>         &END MOLECULAR_DIPOLES
>>>        &WANNIER_CENTERS
>>>             IONS+CENTERS
>>>             FILENAME =water_wannier.xyz<http://water_wannier.xyz><
>>> http://water_wannier.xyz>
>>>             &EACH
>>>                MD 2
>>>             &END EACH
>>>        &END WANNIER_CENTERS
>>>      &END PRINT
>>>   &END LOCALIZE
>>>   &XC
>>>      &XC_FUNCTIONAL BLYP
>>>      &END XC_FUNCTIONAL
>>>      &XC_GRID
>>>        XC_DERIV NN10_SMOOTH
>>>        XC_SMOOTH_RHO NN10
>>>      &END XC_GRID
>>>      &VDW_POTENTIAL
>>>          DISPERSION_FUNCTIONAL PAIR_POTENTIAL
>>>          &PAIR_POTENTIAL
>>>               TYPE DFTD3
>>>               PARAMETER_FILE_NAME dftd3.dat
>>>              REFERENCE_FUNCTIONAL BLYP
>>>          &END PAIR_POTENTIAL
>>>      &END VDW_POTENTIAL
>>>     &END XC
>>>     &PRINT
>>>      &MOMENTS
>>>       PERIODIC TRUE
>>>       FILENAME = moment.out
>>>       &EACH
>>>         MD 2
>>>       &END EACH
>>>      &END MOMENTS
>>>     &END PRINT
>>>    &END DFT
>>>
>>> When I use the same script but without the &PERIODIC EFIELD part, this 
>>> problem is solved. Can you give some suggestions to get over this issue ?
>>>
>>> Thanking you in advance
>>>
>>>
>>>
>>>
>>> --
>>> With Regards,
>>> Rajorshi Chattopadhyay,
>>> PhD Researcher in Mineralogy/Crystallography,
>>> Institut für Geologie und Mineralogie,
>>> Universität zu Köln, Germany
>>>
>>>
>>> --
>>> You received this message because you are subscribed to the Google 
>>> Groups "cp2k" group.
>>> To unsubscribe from this group and stop receiving emails from it, send 
>>> an email to cp2k+uns... at googlegroups.com<mailto:
>>> cp2k%2Bunsu... at googlegroups.com><mailto:cp2k+uns... at googlegroups.com
>>> <mailto:cp2k%2Bunsu... at googlegroups.com>>.
>>> To view this discussion on the web visit 
>>> https://groups.google.com/d/msgid/cp2k/CACTJSRec8FNnOXTY1kT8-uC89d7rn7zagyz3RS1T0hDZLJya7A%40mail.gmail.com
>>> <
>>> https://groups.google.com/d/msgid/cp2k/CACTJSRec8FNnOXTY1kT8-uC89d7rn7zagyz3RS1T0hDZLJya7A%40mail.gmail.com?utm_medium=email&utm_source=footer
>>> >.
>>>
>>> --
>>> You received this message because you are subscribed to the Google 
>>> Groups "cp2k" group.
>>> To unsubscribe from this group and stop receiving emails from it, send 
>>> an email to cp2k+uns... at googlegroups.com<mailto:
>>> cp2k%2Bunsu... at googlegroups.com>.
>>> To view this discussion on the web visit 
>>> https://groups.google.com/d/msgid/cp2k/ZR0P278MB07595B068CC804C61A0E919A9FD69%40ZR0P278MB0759.CHEP278.PROD.OUTLOOK.COM
>>> .
>>>
>>>
>>> --
>>> With Regards,
>>> Rajorshi Chattopadhyay,
>>> PhD Researcher in Mineralogy/Crystallography,
>>> Institut für Geologie und Mineralogie,
>>> Universität zu Köln, Germany
>>>
>>> 
>>>
>>> --
>>> You received this message because you are subscribed to the Google 
>>> Groups "cp2k" group.
>>> To unsubscribe from this group and stop receiving emails from it, send 
>>> an email to cp2k+uns... at googlegroups.com<mailto:
>>> cp2k+uns... at googlegroups.com>.
>>> To view this discussion on the web visit 
>>> https://groups.google.com/d/msgid/cp2k/CACTJSRfvXFGcQS3_yMvkF97gQOV1D7o1wjjvU9MMpT9OE%3DpuFg%40mail.gmail.com
>>> <
>>> https://groups.google.com/d/msgid/cp2k/CACTJSRfvXFGcQS3_yMvkF97gQOV1D7o1wjjvU9MMpT9OE%3DpuFg%40mail.gmail.com?utm_medium=email&utm_source=footer
>>> >.
>>>
>>> -- 
>>> You received this message because you are subscribed to the Google 
>>> Groups "cp2k" group.
>>> To unsubscribe from this group and stop receiving emails from it, send 
>>> an email to cp2k+uns... at googlegroups.com.
>>> To view this discussion on the web visit 
>>> https://groups.google.com/d/msgid/cp2k/ZR0P278MB07594A6DAD79431CA39E0C599FD99%40ZR0P278MB0759.CHEP278.PROD.OUTLOOK.COM
>>> .
>>>
>>
>>
>> -- 
>> With Regards,
>> Rajorshi Chattopadhyay,
>> PhD Researcher in Mineralogy/Crystallography,
>> Institut für Geologie und Mineralogie,
>> Universität zu Köln, Germany
>>
>> 
>>
>
>
> -- 
> With Regards,
> Rajorshi Chattopadhyay,
> PhD Researcher in Mineralogy/Crystallography,
> Institut für Geologie und Mineralogie,
> Universität zu Köln, Germany
>
> 
>

-- 
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+unsubscribe at googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/25975f33-b8c6-4c61-87b5-a1f64fe9bef3n%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20220610/084a3a55/attachment-0001.htm>
Previous message (by thread): [CP2K-user] [CP2K:17113] CPASSERT failed when trying to restart AIMD with periodic efield
Next message (by thread): [CP2K-user] [CP2K:17116] CPASSERT failed when trying to restart AIMD with periodic efield
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the CP2K-user mailing list