Abstract: One of the most appealing factors about any distributed task execution system is its processing speed to process the large amounts of data presented by incoming inputs. With increasing scope and popularity of Cloud services in Industrial applications it becomes important to achieve speeds over the Cloud which should be equal to or, if possible, faster than the existing traditional computing structure. Today, most of the existing frameworks utilizing Cloud services are based on Master/Slave architecture. The limitations associated with these Master/Slave architectures include lack of scalability and a single point of failure, the point of vulnerability being the Master. Our proposed system, LoBUCS, will provide better scalability and fast processing of HPC [4] and MTC [2] workloads using a fully parallel Client-Worker based structure eliminating the single point of failure. By utilization of Amazon Web Services (Amazon EC2 [5], DynamoDB [7] and SQS [6]) we will be substituting the complexity of existing distributed job management systems while maintaining good utilization.

Keywords: Cloud Services; parallel processing; distributed scheduling; LoBUCS.