Techniques and applications of computation slicing

Abstract

Writing correct distributed programs is hard. In spite of extensive testing and debugging, software faults persist even in commercial grade software. Many distributed systems, especially those employed in safety-critical environments, should be able to operate properly even in the presence of software faults. Monitoring the execution of a distributed system, and, on detecting a fault, initiating the appropriate corrective action is an important way to tolerate such faults. This gives rise to the predicate detection problem which requires finding a consistent cut of a given computation exists that satisfies a given global predicate, if it exists.Detecting a predicate in a computation is, however, an NP-complete problem in general. In order to ameliorate the associated combinatorial explosion problem, we introduce the notion of computation slice. Formally, the slice of a computation with respect to a predicate is a (sub)computation with the least number of consistent cuts that contains all consistent cuts of the computation satisfying the predicate. Intuitively, slice is a concise representation of those consistent cuts of a computation that satisfy a certain condition. To detect a predicate, rather than searching the state-space of the computation, it is much more efficient to search the statespace of the slice.We prove that the slice exists and is uniquely defined for all predicates. We present efficient algorithms for computing the slice for several useful classes of predicates. We establish that the problem of computing the slice for an arbitrary predicate is NP-complete in general. We develop efficient heuristic algorithms for computing an approximate slice for such predicates for which computing the slice is otherwise provably intractable. Our experimental results demonstrate that slicing can lead to an exponential improvement over existing techniques for predicate detection in terms of time and space.

Monitoring Distributed Controllers: When an Efficient LTL Algorithm on Sequences Is Needed to Model-Check Traces

“…We must study how our method can improve the efficiency of existing MSC validation methods. Moreover, we would like to explore the possibility of integrating other techniques such as computation slicing [MG05], in order to gain in time an space during the validation.…”

A Distributed Abstraction Algorithm for Online Predicate Detection

“…Monotonic Channel Predicates [7]: Some examples are: all messages have been delivered (or all channels are empty), at least k messages have been sent/received, there are at most k messages in transit between two processes, the leader has sent all "prepare to commit" messages, etc. Centralized offline [8] and online [9] algorithms for slicing based predicate detection have been presented previously. In this paper, we present the first distributed online slicing algorithm for regular predicates in distributed systems.…”

Efficient online algorithm for identifying useless states in distributed systems

“…In the distributed system, a global predicate is comprised of variables from distinct processes. The global predicate detection problem finds a consistent global state of the distributed program execution on which is true [4,10,13]. Usually is used to formalize the undesired situation of the distributed system.…”