Detection method and device for verifying distributed protocol active attribute fault-tolerant mechanism

Abstract

The invention relates to the field of model detection, in particular to a detection method and device for verifying a distributed protocol active attribute fault-tolerant mechanism, and the method comprises the following steps: initializing a to-be-verified system; obtaining a migration set T_Set which can be executed by the to-be-verified system in the current state; defining peer nodes; definingthe same operation; reducing the migration set T_Set based on a peer-to-peer reduction strategy; obtaining the state of the to-be-verified system after migration; checking the security attribute of the state of the to-be-verified system after migration, and obtaining a security state; checking the active attribute of the security state, obtaining the active state, injecting a fault, and re-checking the security attribute and the active attribute of the state of the to-be-verified system; and outputting a counter-example set C. The method and the device are better in verification effect, and can equivalently traverse the whole space under the condition of exploring less execution sequences.

Description

technical field [0001] The invention relates to the field of model detection, in particular to a detection method and device for verifying a fault-tolerant mechanism of active attributes of a distributed protocol. Background technique [0002] Model detection is used in software and hardware systems that require high reliability and security. Model the system as a whole to formally describe all possible behaviors of the system during runtime. By traversing these behaviors, find out whether there are violations. Behavior of system design properties, outputting violations as counterexamples if present. By traversing the system space, the model checking technology can find out all the bugs that violate the attributes in the system to be verified, which ensures the high security and high reliability of the system. [0003] Existing model checking technologies are divided into implementation-level model checking (implementation-level) and abstract model checking. Implementation-...

AI Technical Summary

Problems solved by technology

However, due to the need to traverse all the state spaces of the system, model checking technology has been plagued by the state space explosion problem of the system to be verified.

[0004] At the same time, model checking using an abstract model can only verify the design of the system and its abstract model, but cannot find the defects introduced by the code implementation; at the same time, abstract modeling of the system to be verified may introduce the problem of mismatch between the model and the system , model checking on mismatched models is pointless

However, the existing implementation-level model checking tools have less consideration for the verification of liveness properties and their fault tolerance.

[0005] That is to say, the existing model checking tools focus on the verification of the security attributes of distributed systems, ignoring the detection of liveness attributes and fault-tolerant mechanisms, especially liveness attribute fault-tolerant mechanisms. Injecting faults to verify fault tolerance, but the faults injected by these model checking tools are often random injections or exhaustive all possible situations of faults (faults are injected at any time when the system to be verified is running), which often lacks purpose for fault injection However, it brings many problems and challenges. For example, using random injection cannot find out where all fault-tolerant mechanisms do not work properly, and exhaustive injection will aggravate the problem of space explosion, making the model detector not scalable.

Images