[FLASH-USERS] hangup with laser2Din3D

[Mario Manuel]

Hi Klaus,

I tried the FLASH4.3 LaserSlab simulation with different IO settings. It ran with +noio, but ‘crashed’ with +hdf5typeio and +serialIO. My hdf5 library must be incompatible with the 3Din2D algorithm? I have been using hdf5-1.8.5/parallel with 3Din2D=false and had no problems. Is there a suggested hdf5 version for FLASH4.2.2? FLASH4.3?

Thanks for helping me narrow down the source of my ‘crash’.

Cheers,
Mario

______________________________________
                                                                                                                                
Dr. Mario Manuel
Einstein Postdoctoral Fellow, Smithsonian Astrophysical Observatory
mmanuel at umich.edu
617-899-8381

University of Michigan
Space Research Building
2455 Hayward Street
Ann Arbor, MI 48109
Office: 734-764-4585
______________________________________

On Apr 22, 2016, at 7:10 PM, Klaus Weide <klaus at flash.uchicago.edu> wrote:

> On Fri, 22 Apr 2016, Mario Manuel wrote:
> 
>> ... did notice that the top line of the output when it crashes is:
>> 
>> NOTE: Enabling curvilinear, cartesian_pm/cylindrical_pm/spherical_pm/polar_pm  so far was F F F F
>> 
>> Usually in my other simulations (that use 3Din2D=false) the first line looks like:
>> 
>> NOTE: Enabling curvilinear, cartesian_pm/cylindrical_pm/spherical_pm/polar_pm  so far was F T F F
>> 
>> I don't know if the lack of a T indicates something significant� 
> 
> Not really. This should change if you add +cylindrical or
> -geometry=cylindrical to your setup command, but FLASH should be 
> able to handle non-Cartesian geometries correctly even if they are not 
> specified on the setup command line.
> 
>> I guess I should also note that the first _chk_0000 file and the 
>> LaserEnergyProfile.dat file get created, but are empty (i.e. 0 bytes).
> 
> You seem to have serious filesystem trouble, an empty HDF5 file points to 
> rpblesm with the HDF5 library (especially if, as you seemed to say, FLASH
> did not die while having to the open for writing, but was just hanging).
> 
> Maybe try +serialIO, or even +noio.
> 
> Klaus

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20160422/f03622f2/attachment-0001.htm>