The tpv33 calculation result is invalid and the tpv5 calculation program terminates

[dxycugb]

Hi, I am trying to run the benchmark of seissol on my personal computer but I have encountered problems.

1. About tpv33:

I changed Format = 6 in the parameter file to enable hdf5 output, but the tpv33 output file I got when paraview visualized didn't seem to have anything available.

Also, how should I use viewrec to visualize dat files?

The parameter file of tpv33 is officially provided and only Format = 6 has been modified to enable 3D wave field output.I'm not sure if the log I provided is what you need. Parameter files, log files, and output xdmf files are attached.

2.About tpv5:

I managed to run tpv33 with order 3 but the results did not seem correct, however the program was aborted when I tried to run tpv5 with order 2 and 3. I'm not sure if it's a lack of memory issue, and I haven't tried running it on a computer with more memory.
The parameter file is also given by the official, only modify Format = 6.
Output log:
dxy@pc:~/Examples/tpv5$ ./SeisSol_Release_dhsw_2_elastic parameters.par
Wed Feb 21 22:13:07, Info: Welcome to SeisSol
Wed Feb 21 22:13:07, Info: Copyright (c) 2012 - 2024 SeisSol Group
Wed Feb 21 22:13:07, Info: Version: v1.1.2-163-gf21db452
Wed Feb 21 22:13:07, Info: Built on: Feb 21 2024 20:23:04
Wed Feb 21 22:13:07, Info: Last commit: f21db45 at 2024-01-26T14:13:16+00:00
Wed Feb 21 22:13:07, Info: Running on: "pc"
Wed Feb 21 22:13:07, Info: Using MPI with #ranks: 1
Wed Feb 21 22:13:07, Info: Using asynchronous MPI routines.
Wed Feb 21 22:13:07, Info: Using OMP with #threads/rank: 16
Wed Feb 21 22:13:07, Info: OpenMP worker affinity (this process): "0123456789|012345"
Wed Feb 21 22:13:07, Info: OpenMP worker affinity (this node) : "0123456789|012345"
Wed Feb 21 22:13:07, Info: Using polling for advancing MPI communication, due to having only one MPI rank running.
Wed Feb 21 22:13:07, Info: The stack size ulimit is 8192 [kiB] ( = 8.0000 MiB ).
Wed Feb 21 22:13:07, Warn: Stack size of 8192 [kiB] ( = 8.0000 MiB ) is lower than recommended minimum of 2097152 [kiB] ( = 2.0000 GiB ). You can increase the stack size by running the command: ulimit -Ss unlimited.
Wed Feb 21 22:13:07, Info: Reading SeisSol parameter file...
Wed Feb 21 22:13:07, Info: In your parameter file, you have specified a non-positive interval for "checkpointinterval" . This mechanism is deprecated and may be removed in a future version of SeisSol. Consider disabling the whole module by setting "checkpoint" to 0 instead by adding "checkpoint = 0" to the "output" section of your parameter file instead.
Wed Feb 21 22:13:07, Info: The field "faultoutputflag" is no longer in use. You may safely remove it from your parameters file.
Wed Feb 21 22:13:07, Info: SeisSol parameter file read successfully.
Wed Feb 21 22:13:07, Info: Model information:
Wed Feb 21 22:13:07, Info: Elastic model: yes
Wed Feb 21 22:13:07, Info: Viscoelastic model: no
Wed Feb 21 22:13:07, Info: Anelastic model: no
Wed Feb 21 22:13:07, Info: Poroelastic model: no
Wed Feb 21 22:13:07, Info: Anisotropic model: no
Wed Feb 21 22:13:07, Info: Plasticity: no
Wed Feb 21 22:13:07, Info: Begin init mesh.
Wed Feb 21 22:13:07, Info: Mesh file: "tpv5_f200m.puml.h5"
Wed Feb 21 22:13:07, Info: Skipping mini SeisSol (SeisSol is used with a single rank only).
Wed Feb 21 22:13:07, Info: Reading PUML mesh
Wed Feb 21 22:13:07, Info: Found 1724102 cells
Wed Feb 21 22:13:07, Info: Found 283077 vertices
Wed Feb 21 22:13:15, Info: Computing LTS weights.
Wed Feb 21 22:13:18, Info: Limiting number of clusters to 2147483646
Wed Feb 21 22:13:18, Info: Computing LTS weights. Done. (6840 reductions.)
Wed Feb 21 22:13:18, Info: Using the Parmetis partition library and strategy.
Wed Feb 21 22:13:30, Info: PUML mesh read in: 23.8525 s (min: 23.8525 s, max: 23.8525 s)
Wed Feb 21 22:13:30, Info: The mesh has been read. Starting post processing.
Wed Feb 21 22:13:30, Info: Extracting fault information.
Wed Feb 21 22:13:31, Info: Exchanging ghostlayer metadata.
Wed Feb 21 22:13:31, Info: Mesh initialized in: 24.4142 s (min: 24.4142 s, max: 24.4142 s)
Wed Feb 21 22:13:31, Info: End init mesh.
Wed Feb 21 22:13:31, Info: Begin init model.
Wed Feb 21 22:13:31, Info: Initialize LTS.
Wed Feb 21 22:13:31, Info: Deriving clusters ids for min. time step width / multiRate: 0.000968002 / 2
Wed Feb 21 22:13:31, Info: Due to wiggle factor of 1 the minimum timestep size is reduced to 0.000968002
Wed Feb 21 22:13:31, Info: Number of elements in time clusters:
Wed Feb 21 22:13:31, Info: 0: 27379
Wed Feb 21 22:13:31, Info: 1: 585166
Wed Feb 21 22:13:31, Info: 2: 561365
Wed Feb 21 22:13:31, Info: 3: 299070
Wed Feb 21 22:13:31, Info: 4: 175181
Wed Feb 21 22:13:31, Info: 5: 75064
Wed Feb 21 22:13:31, Info: 6: 877
Wed Feb 21 22:13:31, Info: maximum theoretical speedup (compared to GTS): 4.89026 per cell LTS, 3.37401 with the used clustering.
Wed Feb 21 22:13:31, Info: Number of elements in dynamic rupture time clusters:
Wed Feb 21 22:13:31, Info: 0 (dr): 773
Wed Feb 21 22:13:31, Info: 1 (dr): 25707
Wed Feb 21 22:13:31, Info: 2 (dr): 0
Wed Feb 21 22:13:31, Info: 3 (dr): 0
Wed Feb 21 22:13:31, Info: 4 (dr): 0
Wed Feb 21 22:13:31, Info: 5 (dr): 0
Wed Feb 21 22:13:31, Info: 6 (dr): 0
Wed Feb 21 22:13:31, Info: Initialize Friction Model
已杀死(Killed)

attachments:
tpv33+tpv5.zip

[sebwolf-de]

1. It is to be expected to see blue box in Paraview initially. Choose another quantity and another time step. I can't visualize the xdmf files without the *.bin files.
2. I am not surprised that a mesh with 1.7M elements does not work on a workstation. Try to go for <200k elements.

Since this is not a problem of SeisSol, but a question about the proper usage of SeisSol, I am goint to convert this issue to a discussion.