Skip to content

Explore bridges between CWL and Dask #908

@fmigneault

Description

@fmigneault

Description

There is a great overlap potential between (Dask Clusters + Argo Workflows) and (Docker/Kubernetes + CWL), given EOEPCA, and considering the tools that can leverage different HPC/Clouds. Given how Dask ties well with multiple data representations for scientific work (xarray, numpy, pandas, dataframe, scikit, etc.) it is often a good (or more easily accessible?) for users working within the same Python environment where they manipulate data. However, Dask lacks the workflow annotation layer that CWL provides.

If we could define some helpers/converters between them (can they be natively represented, or simply embedding the Python script is enough?), we could lower the adoption/introduction bar of CWL concepts (notably for "simpler" use cases that do not employ advanced Dask features).

To Do

References

Extra Docs:

Issues;

Metadata

Metadata

Assignees

Labels

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions