- Setting up your system
- Cloning sources
- Using production-like development mode
- Using live-code-reload and debug mode
Understanding code base
Technology tips and tricks
Clone this wiki locally
Welcome to the REANA Developer Wiki!
This wiki is intended for software developers working on the REANA team or otherwise contributing to the REANA source code.
If you are a researcher interested in using REANA, please see the regular docs.reana.io documentation website.
If you are a cluster administrator interested in deploying REANA on your premises, please see the "Administration" section of docs.reana.io documentation website instead.
REANA is a reproducible analysis platform allowing researchers to run containerised data analyses on remote compute clouds.
The analysis is structured based on four questions: (i) where is the input data and parameters, (ii) what code is used to analyse the data, (iii) which computing environments are used to run the analysis code, and (iv) what are the computational steps taken to arrive at the results.
REANA supports several declarative workflow systems (CWL, Snakemake, Yadage), parses the workflow specification described by the researcher and dispatches its computational steps to several supported compute backends (Kubernetes, HTCondor, Slurm). The reproducibility of computations is assisted by means of using software containers (Docker, Singularity) that fully encapsulate the computational environment.
REANA allows researchers to launch multiple analysis runs with various input parameters, open interactive Jupyter notebooks on the workspace whilst the analysis is running, maintain and compare list of past runs, share results with colleagues over the web interface, and more.
Thanks to the declarative programming approach and the support for multiple workflow engines and multiple compute backends, REANA also allows using hybrid computational workflow paradigm where parts of the analysis is run on one compute backend whilst the other parts of the analysis are seamlessly dispatched to other compute backend based on high-throughput or high-performance computing needs of the step at hand.
For more information on REANA, please read:
- Open is not enough. Nature Physics 15 (2019) 113-119.
- Open science: A vision for collaborative, reproducible and reusable research. CERN Courier, 11 March 2019.
- REANA: A system for reusable research data analyses, EPJ Web Conf 214 (2019) 06034.
- Scalable declarative HEP analysis workflows for containerised compute clouds, Front. Big Data, 07 May 2021.
If you are new to the REANA development, please see Before you start.