Publications Repository - Helmholtz-Zentrum Dresden-Rossendorf

Data processing or analysis workflows are generally understood as processes running without any user intervention where usually only a small set of parameters being provided upon workflow submission, adjustment of which is also limited by low turnaround rates of workflow runs due to scheduling alone. Many types of experimental data analyses require manual experimentation with parameters to succeed, necessitating interactivity and fast iteration. In this talk we present examples of interactive workflow applications at HZDR, from data analysis and simulation; discuss challenges arising from differences to completely automated workflows and lay-out the related data-provenance and project-resource management features we envision for the HELIPORT workflow platform.