Open mr-c opened 2 years ago
Thanks for creating the ticket for me personally only the metadata (rdf files, workflows files) are needed. As the input and output files are preserved on a cloud store.
Areas to investigate, (add flag to skip the copying, but still calculate and store the checksums)
https://github.com/common-workflow-language/cwltool/blob/a1e3449560b964d90818b2f1bfeb9b411415a786/cwltool/provenance.py#L790 https://github.com/common-workflow-language/cwltool/blob/a1e3449560b964d90818b2f1bfeb9b411415a786/cwltool/provenance.py#L929
https://github.com/common-workflow-language/cwltool/blob/a1e3449560b964d90818b2f1bfeb9b411415a786/cwltool/provenance.py#L741 called from https://github.com/common-workflow-language/cwltool/blob/a1e3449560b964d90818b2f1bfeb9b411415a786/cwltool/main.py#L1413
To update this, when providing Directory or Files as input it will copy the entire content to /tmp. Solution for now is to use Strings instead of Directory when possible.
Solution for now is to use Strings instead of Directory when possible.
FYI, while that may work for now, that will break mulit-node execution of the workflow
https://matrix.to/#/!RQMxrGNGkeDmWHOaEs:gitter.im/$AJGFCdt6jVAn3aR5lQ0PK3_0SGgvFrubf5SMClsOgGA (a.k.a https://gitter.im/common-workflow-language/common-workflow-language?at=61d6a7bfbfe2f54b2e04661d )
--prov-exclude-inputs
Skips copying the input files into the CWLProv ResearchObject--prov-exclude-intermediates
Skips copying the intermediate files into the CWLProv ResearchObject--prov-exclude-outputs
Skips copying the output files into the CWLProv ResearchObject