Evolution of distributed monitoring for scalable business process management in cloud-based databases

Embodying business process as a service (BPaaS) in cloud computing has been actively investigated by many research studies. There are many proposals to establish the scalable workflows, for instance, by adopting the multiple instances of workflow execute engine (WEE). However, it is difficult to track exactly the states of individual instances of the workflow running across multiple engines globally, by only applying the multiple WEE. It is necessary to implement an independent monitoring function from outside. This study considers making this monitoring function scalable in accordance with increasing the number of WEE. When designing the monitoring function and the maintenance protocol, the following important factors have been considered. First, to deliver the meta-data such as definitions of business processes among WEE for sharing. The second, to deliver the event instances asynchronously, by compensating inconsistent status for recoverable events. This study proposes architecture and a set of maintenance protocols to realise the scalable workflow management. It shows the critical requirements and selects a reasonable approach through qualitative evaluations.