FRCL based Runtime Verification of Inner Cores for networks
نویسندگان
چکیده
High throughput, low latency and high system availability are the most important criteria for an internet router. Internet Router‟s complexity has been on constant rise to serve the exploding internet backbone. With rise in design complexity comes the increased possibility of bugs escaping into silicon and affecting the user at runtime. In this project we provide a runtime verification scheme for a network router. We implement a verilogbased learning router and describe FRCL (field repairable control logic) based state matching system to guarantee the correct functionality of our router system in case of permanent and transient hardware faults as well as design errors. In [1] it has been shown that FRCL technique has been used to develop a fault-tolerant processor system wherein the core switches to less featured but fully functional inner core on occurrence of an error. The aim of our project is to extend this technique to a network router design. We have been able to demonstrate that the addition of the state matcher in the router design introduces very low area overhead in case of fault-free execution while guaranteeing lossless delivery of packets in case of fault occurrence.
منابع مشابه
FRCLe based Runtime Verification of Inner Cores for networks
High throughput, low latency and high system availability are the most important criteria for an internet router. Internet Router’s complexity has been on constant rise to serve the exploding internet backbone. With rise in design complexity comes the increased possibility of bugs escaping into silicon and affecting the user at runtime. In this project we implement a verilog-based learning rout...
متن کاملRethinking Runtime Verification on Hundreds of Cores: Challenges and Opportunities
We propose a novel approach for runtime monitoring and verification on computers with a large number of computation cores. The goal of the approach is to minimize the impact of runtime verification on the performance of the application being monitored. We distinguish between two kinds of computational overhead: (i) overhead caused by instrumentation and/or logging, and (ii) parallelizable overh...
متن کاملKUDA: GPU Accelerated Split Race Checker
We propose a novel approach for runtime verification on computers with a large number of computation cores, without any hardware extension to mainstream PC environment. The goal of the approach is making use of all hardware resources to decouple the computational overhead of traditional race checkers via parallelizing the runtime verification. We distinguish between two kinds of computational o...
متن کاملOn the Runtime Enforcement of Evolving Privacy Policies in Online Social Networks
Online Social Networks have increased the need to understand well and extend the expressiveness of privacy policies. In particular, the need to be able to define and enforce dynamic (and recurrent) policies that are activated or deactivated by context (events) or timeouts. We propose an automaton-based approach to define and enforce such policies using runtime verification techniques. In this p...
متن کامل