Skip to content

Feature Request: Configurable Working Directory, Safe Job Termination, and File Recovery #35

@poit0987-star

Description

@poit0987-star

Hello,

I am Park, who recently contacted you by email regarding OpenMP performance.

First of all, thank you for the recently added code_aster keyword suggestion feature. I find it very useful. I am also very satisfied with the mesh visualization feature.

I am currently performing nonlinear static analyses and nonlinear iterative analyses involving geometrical nonlinearity, material nonlinearity, and contact/friction conditions. In particular, my models include multiple contact regions, typically around 8 to 10 contact zones in a single model. For this reason, the mesh viewer and workflow integration in VS Code Aster are very useful for my work.

While using VS Code Aster in practical analyses, I noticed a few inconveniences and would like to share some suggestions that may be helpful for future improvements.

1.Configurable working / temporary directory

It would be very useful to have a GUI option to configure the working directory or temporary directory used by code_aster.

In particular, it would be helpful if the default temporary path inside the Docker image could be changed to a user-defined working drive or a bind-mounted host directory.

  1. Safe stop / terminate function for running calculations

It would be useful to have a simple and safe way to stop or terminate a running calculation in VS Code Aster.

Currently, when a calculation is running, the user may need to manually terminate the Python process or related processes in order to stop the analysis.

  1. Preservation of important files after abnormal termination

It would be very useful if important calculation files could be copied or preserved as safely as possible even when an analysis is interrupted or fails.

Currently, when a calculation is forcibly terminated or an exception occurs, the user may need to manually find the temporary working directory, and it can be difficult to recover the necessary files.

If possible, it would be very useful to add an exception-handling structure such as try / except / finally in the execution script, so that important files such as .base, .mess, and result files can be copied to the specified output directory even when the calculation does not terminate normally.

  1. Mesh group display behavior

When a specific mesh group is selected in the rainbow viewer or mesh viewer, the original full mesh automatically becomes transparent.

This behavior is useful in many cases, but it would be more convenient if there were an option to display or isolate only the selected mesh group without making the original mesh transparent.

  1. Improved keyword completion and diagnostics

It would be helpful if the auto-completion and diagnostic features for valid code_aster keywords could be further improved.

For example, valid keywords such as FORCE, DEFORMATION, and CRITERES in the CALC_CHAMP command do not appear in the suggestion list. Also, warnings for undefined keywords may temporarily appear while typing.

Since the code_aster calculation itself runs without any problem, this seems to be related to the extension’s auto-completion or diagnostic feature rather than code_aster itself.

  1. Handling of existing output files

When using VS Code Aster, if a calculation stops due to an error, related .base files may not be saved. Also, unlike ASTK, VS Code Aster seems not to overwrite files that already exist in the result directory, such as .base or .mess files.

It would be useful to have a clearer overwrite option or file-handling behavior for existing output files.

  1. Real-time residual graph monitoring

It would be very helpful to have a real-time residual graph monitoring feature, similar to AsterStudy.

For nonlinear analyses involving contact and friction, this is important because convergence behavior often needs to be monitored while the calculation is running.

In particular, items 1 to 3 and item 6 often cause significant inconvenience when performing complex analyses such as nonlinear contact/friction analyses, where the calculation time is long and the result files can be large. I would greatly appreciate it if these points could be considered in future improvements.

Thank you again for developing and maintaining this excellent extension.

Best regards,
Park

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions