intro story D-Flow FM

 

D-Flow Flexible Mesh

D-Flow Flexible Mesh (D-Flow FM) is the new software engine for hydrodynamical simulations on unstructured grids in 1D-2D-3D. Together with the familiar curvilinear meshes from Delft3D 4, the unstructured grid can consist of triangles, pentagons (etc.) and 1D channel networks, all in one single mesh. It combines proven technology from the hydrodynamic engines of Delft3D 4 and SOBEK 2 and adds flexible administration, resulting in:

  • Easier 1D-2D-3D model coupling, intuitive setup of boundary conditions and meteorological forcings (amongst others).
  • More flexible 2D gridding in delta regions, river junctions, harbours, intertidal flats and more.
  • High performance by smart use of multicore architectures, and grid computing clusters.
An overview of the current developments can be found here.
 
The D-Flow FM - team would be delighted if you would participate in discussions on the generation of meshes, the specification of boundary conditions, the running of computations, and all kinds of other relevant topics. Feel free to share your smart questions and/or brilliant solutions! 

 

=======================================================
We have launched a new website (still under construction so expect continuous improvements) and a new forum dedicated to Delft3D Flexible Mesh.

Please follow this link to the new forum: 
/web/delft3dfm/forum

Post your questions, issues, suggestions, difficulties related to our Delft3D Flexible Mesh Suite on the new forum.

=======================================================

** PLEASE TAG YOUR POST! **

 

 

Sub groups
D-Flow Flexible Mesh
DELWAQ
Cohesive sediments & muddy systems

 


Message Boards

Running in parallel FLOW (FLOW2D3D Version 6.02. - 01_standard example)

JP
Juan Pineda, modified 2 Years ago.

Running in parallel FLOW (FLOW2D3D Version 6.02. - 01_standard example)

Youngling Posts: 3 Join Date: 3/18/15 Recent Posts

Hi everybody,

I just compiled Delft3D (8799) and everything looks good. I could ran the first example (01_standard) in parallel using MPICH2-3.2.1 but I found an issue when I tried to run with more than 5 cores this simulation, I mean, I can run this example with 1 or 2 or 3 or 4 or 5 cores but when I try to run using 6 or more cores it fails...

I got the following error:

Part IV   - Reading complete MD-file...                     
 Flow exited abnormally
 Check diagnosis file
 Flow exited abnormally
 Check diagnosis file
 Flow exited abnormally
 Check diagnosis file
 Flow exited abnormally
 Check diagnosis file
 Flow exited abnormally
 Check diagnosis file
 Flow exited abnormally
 Check diagnosis file
terminate called after throwing an instance of 'char const*'
terminate called after throwing an instance of 'char const*'
terminate called after throwing an instance of 'char const*'
terminate called after throwing an instance of 'char const*'
terminate called after throwing an instance of 'char const*'
terminate called after throwing an instance of 'char const*'

 

Is there any way to run this example with more cores? What file have I to pay attention to increase the number of cores?

 

Thanks in advanced!

 

 

Adri Mourits, modified 2 Years ago.

RE: Running in parallel FLOW (FLOW2D3D Version 6.02. - 01_standard example)

Yoda Posts: 1224 Join Date: 1/3/11 Recent Posts

Hi Juan,

example "01_standard" is too small to run with more than 5 partitions. When "max(mmax,nmax)/npart <= 4", the computation will not run.

Regards,

Adri

JP
Juan Pineda, modified 2 Years ago.

RE: Running in parallel FLOW (FLOW2D3D Version 6.02. - 01_standard example)

Youngling Posts: 3 Join Date: 3/18/15 Recent Posts

Hi Adri, thanks for your reply. Where can I check "max(mmax,nmax)/npart <= 4" part? Which file?

Regards

Adri Mourits, modified 2 Years ago.

RE: Running in parallel FLOW (FLOW2D3D Version 6.02. - 01_standard example)

Yoda Posts: 1224 Join Date: 1/3/11 Recent Posts

Hi Juan,

 

Flow does that for you and echoes an error message in the "tri-diag"-file.

If you want to check it manually:
mmax and nmax are the dimensions of your grid (mdf-file, line "MNKmax", for example "MNKmax= 15 22 5"), npart is the number of partitions that you have chosen (for example 5 or 6), so in this example "max(15,22)/5 = 4.4" is bigger than 4 will run fine and "max(15,22)/6 = 3.666" is smaller than 4 and will stop with an error message.

Regards,

Adri

JP
Juan Pineda, modified 2 Years ago.

RE: Running in parallel FLOW (FLOW2D3D Version 6.02. - 01_standard example)

Youngling Posts: 3 Join Date: 3/18/15 Recent Posts

Hi Adri! Thanks again for your explanation!

In this case my MNKmax variable is 583 335 14. If I apply your explanation, I got the following -> 583/14 = 41.642. That means that I can use 14 cores to run this simulation.

Unfortunately the only way that I could ran this simulation was using just 1 core,  so when I used more cores it failed and I got the following errors on the tri-diag files:

*** WARNING Ratio in layer thickness of the two bottom layers is larger than 1.5.
            May lead to inaccurate representation of the bottom boundary layer.
*** MESSAGE Upwind advection scheme only near momentum discharges
*** WARNING Discharge (m,n,k)=(42,311,0) is disabled: inlet and/or outfall not in this partition
*** WARNING Discharge (m,n,k)=(70,252,0) is disabled: inlet and/or outfall not in this partition
*** WARNING Discharge (m,n,k)=(79,243,0) is disabled: inlet and/or outfall not in this partition
*** WARNING Discharge (m,n,k)=(231,163,0) is disabled: inlet and/or outfall not in this partition
*** WARNING Discharge (m,n,k)=(254,248,0) is disabled: inlet and/or outfall not in this partition
*** WARNING Discharge (m,n,k)=(317,64,0) is disabled: inlet and/or outfall not in this partition
*** WARNING Discharge (m,n,k)=(481,180,0) is disabled: inlet and/or outfall not in this partition
*** WARNING Using Smoothing time and initial condition file
*** MESSAGE Ocean heat model: Stanton number specified to be    0.10000E-02
*** MESSAGE Ocean heat model: Dalton number specified to be    0.10000E-02
*** MESSAGE Using UNESCO density formulation by default
*** MESSAGE Evaporation not taken into account in continuity equation
*** MESSAGE Number of pivot points to convert wind speed to wind drag coef.: 2
*** MESSAGE Updating tidal node factors every   0.6000E+02 [min]
*** MESSAGE Momentum solver cyclic method is specified
*** MESSAGE DRYFLP and DPSOPT both specified in MD-file. Using DPSOPT: MEAN
*** MESSAGE Transport solver cyclic-method method is specified
          remap n=  19 mnit1 =    1 334 mint2 =   14 334 - node number 005
          remap n=  20 mnit1 =   14 334 mint2 =   29 334 - node number 005
*** MESSAGE His, map, drogue, and fourier files written in single precision (except for time and horizontal coordinates).
*** MESSAGE Uniform wind and pressure specified
*** MESSAGE Heat model: Secchi depth specified to be   30.0000000 m
*** ERROR Drogues/walking monitor points are not available in parallel computations
*** ERROR Flow exited abnormally

 

I attached the mdf file and one of the tri-diag files to explain better this case.

Regards,

Juan.

 

Adri Mourits, modified 2 Years ago.

RE: Running in parallel FLOW (FLOW2D3D Version 6.02. - 01_standard example)

Yoda Posts: 1224 Join Date: 1/3/11 Recent Posts

Hi Juan,

Unfortunately the combination of drogues (line "Filpar= #117.par#" in your mdf-file) with parallel computation is not implemented yet. If you remove the drogues, I expect the parallel computation to run.

Regards,

Adri