Choreography-Based Analysis of Distributed Message Passing Programs

Abstract

Abstract-Distributed programs are vulnerable to subtle communication faults (e.g. deadlocks, lost messages). Although the behaviour of individual processes can be relatively straightforward, faults can still be hard to detect because the macroscopic behaviour emerging from interactions can easily become very complex. This paper reports the analysis of gen_server, an Erlang popular library to build client-server applications. Our analysis uses a tool based on choreographic models of communicating finite state machines (CFSMs). We discuss how, once the library has been modelled in terms of CFSMs, an automated analysis can be used to detect potential communication errors. The results of our analysis suggest how to properly use gen_server in order to guarantee the absence of communication errors.

Verifying Asynchronous Interactions via Communicating Session Automata

“…These protocols are used to generate various endpoint APIs implementing a Scribble specification [31,32,48] and to produce runtime monitoring tools [47,49,50]. Taylor et al [68] use multiparty compatibility and choreography synthesis [40] to automate the analysis of the gen_server library of Erlang/OTP. We believe that we can transparently widen the set of safe programs captured by these tools by using k-mc instead of synchronous multiparty compatibility.…”

“…Communicating automata and asynchronous multiparty session types [29] are closely related: the latter can be seen as a syntactical representation of the former [17] where a sending state corresponds to an internal choice and a receiving state to an external choice. This correspondence between communicating automata and multiparty session types has become the foundation of many tools centred on session types, e.g., for generating communication API from multiparty session (global) types [31,32,48,62], for detecting deadlocks in message-passing programs [51,68], and for monitoring session-enabled programs [5,16,47,49,50]. These tools rely on a property called multiparty compatibility [6,18,40], which guarantees that communicating automata representing session types interact correctly, hence enabling the identification of correct protocols or the detection of errors in endpoint programs.…”

“…for detecting deadlocks in message-passing programs [51,68], and for monitoring session-enabled programs [5,16,47,49,50]. These tools rely on a property called multiparty compatibility [6,18,40], which guarantees that communicating automata representing session types interact correctly, hence enabling the identification of correct protocols or the detection of errors in endpoint programs.…”

Interconnectability of Session-Based Logical Processes

“…Recent work develops the connections of multiparty session types and communicating automata [6,22,37,38], denotational semantics [21], Petri Nets [24], applications to, e.g., secure information flow analysis [10,16], dynamic monitoring [5], and reversible computing [15]. Multiparty session types have also been integrated into mainstream programming languages such as MPI [41,47], Java [32,33,42,56], Python [20,43,46], C [49], Go [48], Erlang [25,45,57], Scala [54], and F# [44].…”