Decentralised LTL monitoring

Abstract

Users wanting to monitor distributed or component-based systems often perceive them as monolithic systems which, seen from the outside, exhibit a uniform behaviour as opposed to many components displaying many local behaviours that together constitute the system's global behaviour. This level of abstraction is often reasonable, hiding implementation details from users who may want to specify the system's global behaviour in terms of an LTL formula. However, the problem that arises then is how such a specification can actually be monitored in a distributed system that has no central data collection point, where all the components' local behaviours are observable. In this case, the LTL specification needs to be decomposed into sub-formulae which, in turn, need to be distributed amongst the components' locally attached monitors, each of which sees only a distinct part of the global behaviour. The main contribution of this paper is an algorithm for distributing and monitoring LTL formulae, such that satisfaction or violation of specifications can be detected by local monitors alone. We present an implementation and show that our algorithm introduces only a minimum delay in detecting satisfaction/violation of a specification. Moreover, our practical results show that the communication overhead introduced by the local monitors is considerably lower than the number of messages that would need to be sent to a central data collection point.

Trace Partitioning and Local Monitoring for Asynchronous Components

“…The closest work to ours is [6,12] where global LTL formulae are monitored locally over partitioned traces. The authors propose and evaluate a decentralised approach that decomposes a given global LTL specification into smaller subproperties that analyse separate trace partitions and communicate amongst themselves to handle subformula dependencies accordingly, as depicted in Fig.…”

“…components F and N from E.g. 1 can both exhibit event e), (iv) the evaluation in [6,12] focusses on decentralised communicating monitors that still regard correctness from a global perspective (analogous to [o]ff ∨ [s]ff from E.g. 6); our evaluation rather concentrates on properties that can be cleanly decomposed into local ones that fully capitalise on trace partitioning, (v) their tool assumes a fixed number of components that remains constant throughout execution as opposed to ours, which can handle dynamic partitioning as well.…”

“…When a connection is established 2 , the acceptor exchanges its newly acquired client socket information with the connections supervisor pid_r 3 . Consequently, a protocol handler is spawned by the connections supervisor so that ownership of the client socket is transferred to the handler 4 , permitting it to engage in direct communication with the client from that point onwards 6 . Upon successful creation of the protocol handler, the connections supervisor acknowledges back to the acceptor 5 .…”

Decentralised Evaluation of Temporal Patterns over Component-Based Systems at Runtime

“…Inspired by definitions in [4], our notions of progression and urgency are adapted to the FTPL semantics: they take into account external and internal events as well as scopes of linear temporal patterns. For decentralised evaluation of the FTPL formulae, instead of the set B 4 as in [3], let us consider the set …”

“…Inspired by the LTL decentralised evaluation [4] for closed systems, this paper provides a progressive FTPL semantics allowing a decentralised evaluation of FTPL formulae over open component-based systems -the first contribution. The second contribution consists of an algorithm to answer the temporal pattern decentralised evaluation problem in B 4 and of the correctness and uniqueness results saying that whenever an FTPL property is evaluated in the decentralised manner, it matches the FTPL evaluation using the basic semantics in [3].…”