DARE Platform: a Developer-Friendly and Self-Optimising Workflows-as-a-Service Framework for e-Science on the Cloud

Iraklis A. Klampanos, Sissy Themeli, Alessandro Spinuso, Rosa Filgueira, Malcolm Atkinson, André Gemünd, Vangelis Karkaletsis
October 2020
PDF Cite Project DOI

Abstract

The DARE platform is designed to live in-between user applications and the underlying com- puting resources. It is built on top of containerisation as well as parallelisation technologies, e.g., Kubernetes and MPI. Interfacing with client systems and end-users is achieved via REST- ful APIs. The execution of scientific workflows is achieved via a Workflows-as-a-service layer, which can handle workflows described in either the dispel4py Python library (Filgueira, Krause, Atkinson, Klampanos, & Moreno, 2017), or in the Common Workflow Language (CWL) (Am- stutz et al., 2016).

Type
Conference paper
Publication
In The Journal of Open Source Software
hatespeech nlp ngramgraphs neuralnets
Sissy Themeli
Sissy Themeli
Senior Software Engineer

Experienced Software Engineer, currently working as Senior Software Engineer at Unitary AI.

Related