-
Notifications
You must be signed in to change notification settings - Fork 1
Home
SciCumulus/C² (SCC) is a parallel Scientific Workflow Management System (SWfMS), which allows scientific workflow modeling, execution and analysis. The SCC was developed for supporting the parallel execution on High Performance Computing (HPC) environments. Specifically, we developed SCC to execute computer simulations on cluster and cloud environments, as an integration of Chiron and SciCumulus.
SCC components support parallel data execution according to a workflow specification. SCC is able to gather and query provenance data during workflow execution. Input data (represented by a relation) is fragmented in order to allow parameter sweep in HPC environments. The parallel processing of SCC follows the MapReduce (e.g. Hadoop) style. Although, the SCC implements his own algebraic approach that can perform optimization, dynamic scheduling and workflow steering.
Provenance database of SCC follows the data model PROV-Wf, one extension of the W3C PROV standard. Through this base, users are capable of analyze the domain-specific data, performance data and data transformations involved during workflow execution. SCC uses PostgreSQL RDBMS (Relational Database Management System) to manage the provenance database.
SCC was created in collaboration with teams of Computing Institute from the Fluminense Federal University (UFF) and COPPE Institute from the Federal University of Rio de Janeiro in March 2015. Recent contributions have been made by students from the Polytechnic School of the Federal University of Rio de Janeiro.
In other pages, it is possible to understand the SCC components, to execute SCC with a few examples and to read some publications of our research team.