Skip to content

[Documentation] Extend documentation for debugging submission commands #969

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
jan-janssen wants to merge 1 commit into from
Closed

[Documentation] Extend documentation for debugging submission commands #969

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
23 changes: 23 additions & 0 deletions docs/trouble_shooting.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -83,6 +83,29 @@ of the [Python simple queuing system adatper (pysqa)](https://pysqa.readthedocs.
All parameters in the resource dictionary `resource_dict` are optional. When `pydantic` is installed as optional
dependency the `resource_dict` is validated using `pydantic`.

## Debugging Submission Command
When using the [HPC Job Executor](https://executorlib.readthedocs.io/en/latest/3-hpc-job.html) the Python functions
are submitted to the queuing system as individual jobs. If the submission of a job fails, for example due to an
incorrect `resource_dict` or a configuration issue in the queuing system, `executorlib` captures this error.

In this case, the `future.result()` call will raise the underlying error (e.g., `subprocess.CalledProcessError` or
`ValueError`). Note that in some environments like Jupyter notebooks, the submission error might occur in a background
thread, and `future.result()` might hang if the error is not correctly propagated.

To debug a failed submission, you can inspect the cache directory (default `executorlib_cache`). `executorlib` stores the
input and output for each task as HDF5 files. If a submission fails, you can find the corresponding `_i.h5` file in the
cache directory and manually try to submit the command to get more detailed error messages.
Comment on lines +96 to +97

If you specify an `error_log_file` in the `resource_dict`, `executorlib` will also append any exceptions raised during
the execution to this file:

```python
with SlurmJobExecutor(max_workers=2) as p:
future = p.submit(calc, 2, j=4, resource_dict={"error_log_file": "error.log"})
```

The error log will include the function name, arguments, and the full stack trace of the exception.

## SSH Connection
While the [Python simple queuing system adatper (pysqa)](https://pysqa.readthedocs.io) provides the option to connect to
high performance computing (HPC) clusters via SSH, this functionality is not supported for executorlib. The background
Expand Down
Loading