[FLASH-USERS] Error writing initial HDF5 checkpoint and plot files

Deepali ddeepali at hs.uni-hamburg.de
Wed Dec 21 11:35:58 EST 2022


Hi Adam & Aaron,


Thank you for your time and help!

I am using serial HDF5 for AMR grid. I cleaned my Makefile further and 
tried running FLASH 4.6.2 again with HDF5 1.10.5 and 1.12.0 (both 
usingintel/19.0.5 and impi/2019.5.281). I am getting the following error:

Gravity initialized

  Initial dt verified

  Particles_sinkAccelSinksOnGasAndSinksOnGas:: sink_softening_type_gas = 
linear

HDF5-DIAG: Error detected in HDF5 (1.10.5) thread 46912496411200:

   #000: H5Ddeprec.c line 126 in H5Dcreate1(): not a location ID

major: Invalid arguments to routine

minor: Inappropriate type

   #001: H5Gloc.c line 246 in H5G_loc(): invalid object ID

major: Invalid arguments to routine

minor: Bad value

Driver_abortC called

Error: H5Dcreate particle names, io_h5write_particles


Calling MPI_Abort for immediate shutdown

Abort(1) on node 0 (rank 0 in comm 0): application called 
MPI_Abort(MPI_COMM_WORLD, 1) - process 0


Please find attached the log (flash*.log) and output files (op*.log) for 
each hdf5 version for more details and the Makefile used. I checked that 
this query has been addressed before 
https://flash.rochester.edu/pipermail/flash-users/2019-September/003039.html 
but I do not quite understand how to proceed further.


Best regards,

Deepali


On 19.12.22 19:05, Adam Reyes wrote:
> Hi Deepali,
>
> FLASH should work with hdf5 1.12.0. Which IO implementation did you 
> set up with (+hdftTypeIO, +parallelIO)? You could try the other one. 
> If you could share the log file from the run it might be helpful as well.
>
> *************************************************************************
> Adam Reyes
>
> Code Group Leader, Flash Center for Computational Science
> Research Scientist, Dept. of Physics and Astronomy
> University of Rochester
> River Campus: Bausch and Lomb Hall, 369
> 500 Wilson Blvd. PO Box 270171, Rochester, NY 14627
> Emailadam.reyes at rochester.edu <mailto:adam.reyes at rochester.edu>
> Webhttps://flash.rochester.edu <https://flash.rochester.edu/>
>  (he / him / his)
>
> *************************************************************************
>
> On Mon, Dec 19, 2022 at 12:00 PM Deepali <ddeepali at hs.uni-hamburg.de 
> <mailto:ddeepali at hs.uni-hamburg.de>> wrote:
>
>     Hi Ryan,
>
>
>     I am using FLASH4.6.2. I tried other available HDF5 versions -
>     hdf5/intel/1.12.0, hdf5/intel/1.12.1 , but received different
>     errors with 10.5 and 12.x (tried with -DH5_USE_16_API,
>     -DH5_USE_110_API, -DH5_USE_112_API).
>
>     I also went through the mailing lists where a patch was made
>     available for FLASH4.5, but it didn't work for me.
>
>     Could someone tell me which library versions are compatible with
>     FLASH 4.6.2? Is there a documentation available for the compatible
>     HDF5 library and FLASH versions?
>
>
>     Best,
>
>     Deepali
>
>
>     On 15.12.22 18:40, Ryan Farber wrote:
>>     Hi Deepali,
>>
>>     I think if you search the flash-users mailing list there were
>>     issues with hdf5 1.10.x a few years ago. I think those were
>>     patched but definitely worth trying different hdf5 versions. For
>>     reference I'm using
>>
>>     hdf5-mpi/1.8.21
>>
>>
>>     but I'm still on FLASH4.2.2...which version of FLASH are you using?
>>
>>     Best,
>>     --------
>>     Ryan
>>
>>
>>     On Thu, Dec 15, 2022 at 12:33 PM Deepali
>>     <ddeepali at hs.uni-hamburg.de <mailto:ddeepali at hs.uni-hamburg.de>>
>>     wrote:
>>
>>         Hi Ryan,
>>
>>
>>         Thank you for your reply!
>>
>>         I checked the space on hard drive and majority of it is
>>         available (~ 3 TB). So no problem there. Also I have the
>>         permission to write. Could it be a problem between flash and
>>         the HLRN cluster HDF5 library version?
>>
>>
>>         Best,
>>
>>         Deepali
>>
>>
>>
>>         On 12.12.22 18:54, Ryan Farber wrote:
>>>         Hi Deepali,
>>>
>>>         I haven't seen your error before:
>>>         flash4: io_h5create_dataset.c:29: io_h5create_dataset:
>>>         Assertion `diskSize[i] > 0' failed.
>>>
>>>         But from the sound of it, that file doesn't think you have
>>>         space on the hard drive it's trying to write to. Maybe a
>>>         permissions problem(?)
>>>
>>>         Best,
>>>         --------
>>>         Ryan
>>>
>>>
>>>         On Mon, Dec 12, 2022 at 5:44 PM Deepali
>>>         <ddeepali at hs.uni-hamburg.de
>>>         <mailto:ddeepali at hs.uni-hamburg.de>> wrote:
>>>
>>>             Dear Flash users & developers,
>>>
>>>
>>>             I am trying to simulate a 3D Bonnor-Ebert sphere type
>>>             setup using AMR grid on HLRN machine
>>>             (https://www.hlrn.de/doc/display/PUB/Usage+Guide). I use
>>>             the following compiler and hdf5 modules:
>>>
>>>             intel/18.0.6
>>>
>>>             impi/2018.5(default)
>>>
>>>             hdf5/intel/1.10.5
>>>
>>>             But I am facing a problem with the writing of initial
>>>             HDF5 checkpoint and plot files. FLASH initializes all
>>>             the units needed and then produces the error mentioned
>>>             in the simulation log file (flash.log) attached with
>>>             this email.
>>>
>>>             I would appreciate any help or suggestions in order to
>>>             resolve this error. Please let me know if I should share
>>>             any other information.
>>>
>>>             Thanks and regards,
>>>
>>>             Deepali
>>>
>>>             _______________________________________________
>>>             flash-users mailing list
>>>             flash-users at flash.rochester.edu
>>>             <mailto:flash-users at flash.rochester.edu>
>>>
>>>             For list info, including unsubscribe:
>>>             https://flash.rochester.edu/mailman/listinfo/flash-users
>>>
>     _______________________________________________
>     flash-users mailing list
>     flash-users at flash.rochester.edu
>     <mailto:flash-users at flash.rochester.edu>
>
>     For list info, including unsubscribe:
>     https://flash.rochester.edu/mailman/listinfo/flash-users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20221221/97b07df4/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: op_hdf1_12.log
Type: text/x-log
Size: 2214 bytes
Desc: not available
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20221221/97b07df4/attachment.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: op_hdf1_10.log
Type: text/x-log
Size: 2123 bytes
Desc: not available
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20221221/97b07df4/attachment-0001.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Makefile.h
Type: text/x-chdr
Size: 7443 bytes
Desc: not available
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20221221/97b07df4/attachment-0002.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: flash_hdf1_12.log
Type: text/x-log
Size: 33843 bytes
Desc: not available
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20221221/97b07df4/attachment-0003.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: flash_hdf1_10.log
Type: text/x-log
Size: 33843 bytes
Desc: not available
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20221221/97b07df4/attachment-0004.bin>


More information about the flash-users mailing list