Self-adapting recovery nets for policy-driven exception handling in business processes

Abstract

In this paper, we propose Self-Adapting Recovery Net (SARN), an extended Petri net model, for specifying exceptional behavior in business processes. SARN adapts the structure of the underlying Petri net at run time to handle exceptions while keeping the Petri net design easy. The proposed framework caters for the specification of high-level recovery policies that are incorporated either with a single task or a set of tasks, called a Recovery Region. These recovery policies are generic directives that model exceptions at design time together with a set of primitive operations used at run time to handle the occurrence of exceptions. We identified a set of recovery policies that are useful and commonly needed in many practical situations. A tool has been developed to illustrate the viability of the proposed exception handling technique.

Dependability Assessment of Web Service Orchestrations

“…Recently, with the emergence of self-adaptive architectures, several approaches have been proposed in literature to deal with self-healing business processes [6], [7]. Although our approach is mainly focused on design-time evaluation, its extension to support runtime self adaptation is currently being investigated.…”

Log-based transactional workflow mining

“…Basically, there are two main approaches that deal with failures in existing BPMSs [7]: Ad Hoc and Run Time. The former specifies the exception handling logic within the normal behavior of the workflow.…”

“…Traditional transaction processing models [15] are inadequate for processing specialized transactions including workflow applications, which have long-duration transactions, and a set of activities that must be properly coordinated to ensure correct behavior. To address this shortcoming, researchers have proposed various advanced transaction processing models [7,[16][17][18][19] that are suitable for executing such specialized applications. These advanced transaction processing models coordinate their activities using different kinds of dependencies.…”

“…The basic ideas of: (i) distinguishing between inter and intra activity transactional dependencies, (ii) attaching compensation activities, and cancellation dependencies to the activities of the workflow, (iii) declaring some of the activities to be critical, and (iv) defining coherent points in the process up to which a rollback occurs in case of failure, are common to many of the transactional models proposed earlier, (e.g. SARN [7], WAMO [28], WIDE [29], COO [30] and CREW [31]). These models typically differ in how much flexibility the business process designer has in specifying the backward and forward execution process.…”

“…A significant amount of work applies the concepts introduced in transaction management to business process environments. Hamadi et al [7] offer an exhaustive survey. Besides, they describe a framework for the specification of high-level recovery policies that are incorporated either with a single task or a set of tasks, called Recovery Regions.…”