Towards Business Process Observability

Abstract

The recent move towards migrating business processes in the cloud often requires organisations to have some parts of their business processes execute on cloud-native systems and others on the organizations' own infrastructure. In many scenarios, there could be multi-cloud configurations with parts of the process run by different cloud providers. Hence, current-day business process executions rely on complex software ecosystems with high levels of cross-dependencies, heterogeneity, and redundancies. This growing complexity amplifies the need to monitor and maintain business process executions near real time to ensure the key performance indicators are met. Instances of system failure require early detection and diagnosis of the problem demanding minimal time to understand its impact on the process execution. We propose the notion of business process observability where the intention is to leverage parameters amenable to external monitoring, to adequately represent the state of a business process. The objective of the proposed approach is to offer an end-to-end observability of business processes provisioned in distributed environments that allows for (1) detection, (2) diagnosis, and (3) actions to address business process execution failures. The paper proposes a solution approach that uses observability data to build a cross-layer topology linking a business process and its underlying software ecosystem and provide detailed diagnostics correlating the business process and software execution failures.