Intermittently missing default instrument TXT outputs in Jython batch simulations (SPIS 6.1.0, Windows)

SPIS
1

Dear SPIS developers and users,

I am currently running a large number of SPIS simulations using SPIS 6.1.0 on Windows. I am using a Jython-based automation workflow to create projects, apply global parameters and material settings, run the simulation, perform data mining, and save each project.

The simulations themselves appear to complete successfully in many cases. In the log, I can see messages indicating that the simulation and data mining steps have reached the end, for example:

STEP data.mining OK
END SINGLE CASE: …
SINGLE CASE FINISHED

However, I have observed that some default instrument TXT output files under:

Simulations/Run1/NumKernel/Output

are intermittently missing.

Examples of the files that may be missing are:

Average_surface_potential_of_node_0_(V_,s)ElecNode0_Potential.txt
Total_current_on_node_0…txt
Total_current_on_spacecraft_surface…txt
Simulation_Control…time_steps…txt

The issue is not fully reproducible for every project. Some projects generated through the same Jython workflow contain the expected TXT outputs, while other projects are missing some of them.

At first, I suspected a Windows path-length issue, but some longer output filenames are generated correctly in other cases. Therefore, I am not sure that path length is the main cause.

I also noticed that some TXT files seem to appear slightly later than other output files after the simulation/data-mining step. Because of this, I am wondering whether the default instrument TXT outputs are written asynchronously, or whether an additional waiting/saving step is required after data mining.

My current workflow is roughly:

  1. Create or load a SPIS project
  2. Apply global parameters
  3. Apply material/group settings
  4. Convert UI to NUM
  5. Create default instruments
  6. Prepare and launch the simulation
  7. Wait for the simulation to finish
  8. Run data mining / extract time-dependent data
  9. Save the project

I would like to ask the following questions:

  1. Are all default instrument TXT outputs expected to be generated reliably when running SPIS through a Jython batch workflow on Windows?

  2. Is the creation of default instrument TXT files part of the numerical simulation itself, or is it handled by a UI/data-mining/export layer after the simulation?

  3. Is the default instrument creation step asynchronous? If so, is there a recommended way to confirm that all default instruments have been fully registered before launching the simulation?

  4. After the simulation has finished, is there a recommended command or API call to ensure that all time-dependent instrument data have been completely written to TXT files?

  5. Is it necessary or recommended to add a delay, save the project, or reload the project after the data-mining step before checking the TXT outputs.

I also found a forum discussion about VTK outputs in batch mode. My case appears to be different, because I am not requesting VTK export. The missing files are default instrument TXT outputs under NumKernel/Output. However, I would like to confirm whether similar batch-mode limitations could also affect TXT output writers, or whether the VTK limitation is specific only to VTK post-processing export.

As a temporary workaround, I am testing a safer workflow where each simulation project is run in a fresh SPIS/JVM process, and the script waits until the required average-potential TXT file appears before moving to the next case. This seems more stable, but I would like to understand the correct and recommended SPIS-side procedure.

Any advice on the proper Jython/batch workflow for reliable default instrument TXT generation would be greatly appreciated.

Thank you very much for your help.

Best regards,
Seohyun Park