REANA-Client is a component of the REANA reusable and reproducible research data analysis platform. It provides a command-line tool that allows researchers to submit, run, and manage their computational workflows.
- seed workspace with input code and data
- run computational workflows on remote compute clouds
- list submitted workflows and enquire about their statuses
- download results of finished workflows
$ # create new virtual environment
$ virtualenv ~/.virtualenvs/reana
$ source ~/.virtualenvs/reana/bin/activate
$ # install reana-client
$ pip install reana-clientThe detailed information on how to install and use REANA can be found in docs.reana.io.
The create and validate commands upload a scoped specification bundle for
server-side workflow loading. Declare every imported source explicitly under
workflow.files or workflow.directories so it is available to the loader.
For example, given this Snakemake project:
analysis/
├── reana.yaml
├── Snakefile
└── rules/
└── common.smk
where Snakefile contains include: "rules/common.smk", declare the included
source in reana.yaml:
version: 0.9.0
workflow:
type: snakemake
file: Snakefile
directories:
- rulesPaths are relative to the directory containing the selected specification.
Absolute paths, paths that escape through .., and symbolic links are rejected.
Use workflow.files only for workflow definitions and configuration needed
while loading the workflow; input datasets belong under inputs.files or
inputs.directories.
Validation snapshots accept at most 1,000 files, 2,000 directories, 100 MiB of file content, and 64 relative path components. Symbolic links are not followed.
reana-client validate --environments performs offline image-reference checks
and reports effective runtime identities. Add --pull to verify availability
and inspect those images with your local container runtime and registry
credentials; the REANA server does not contact image registries.