Software-based Fault Tolerance Mission (Im)possible? Peter Ulbrich - PowerPoint PPT Presentation

�� Software-based Fault Tolerance –   Mission (Im)possible? Peter Ulbrich The 29th CREST Open Workshop on Software Redundancy November 18, 2013 System Software Group http://www4.cs.fau.de

� �� Soft Errors – A Growing Problem � � � � � � � � � � � � � ■ Soft-Errors (Transient hardware faults) ! ■ Induced by e.g., radiation, glitches, insufficient signal integrity ■ Affecting microcontroller logic ! ! Peter Ulbrich – ulbrich@cs.fau.de 2

�� Soft Errors – A Growing Problem � � � � � � � � � � � � � � � � � � � � � � � � � � ■ Soft-Errors (Transient hardware faults) ! ■ Induced by e.g., radiation, glitches, insufficient signal integrity ■ Affecting microcontroller logic ! ! Peter Ulbrich – ulbrich@cs.fau.de 2

�� Soft Errors – A Growing Problem � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � [3] ■ Soft-Errors (Transient hardware faults) ! ■ Induced by e.g., radiation, glitches, insufficient signal integrity ■ Affecting microcontroller logic ■ Future hardware designs: more performance performance and parallelism parallelism ! → On the price of being less and less r On the price of being less and less reliable eliable ! Peter Ulbrich – ulbrich@cs.fau.de 2

�� Soft Errors – A Growing Problem Toyota Acceleration Case � � � ■ Electronic throttle control system (2005 Camry) � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � “Toyota claimed the 2005 Camry's main CPU had error detecting and correcting RAM. It didn't. ” 2 ■ Unintended acceleration potentially involving 261 deaths 1 [3] ■ Experts identified soft errors as possible cause 1 ■ Soft-Errors (Transient hardware faults) ! ■ 1 US News, Mar 17, 2010 2 Investigation Report, EDN Network, Oct 28, 2013 Induced by e.g., radiation, glitches, insufficient signal integrity ■ Affecting microcontroller logic ■ Future hardware designs: more performance performance and parallelism parallelism ! → On the price of being less and less r On the price of being less and less reliable eliable ! Peter Ulbrich – ulbrich@cs.fau.de 2

Software-Based Fault Tolerance Safety-Critical System ! ↯ ✗ Replica(1( Replica(1( Majority( Sensors( Interface( Replica(2( Actuators( Voter( Replica(3( ( Isola&on(domain( Sphere(of(redundancy((SOR)( ( ■ Software-based redundancy ! ■ Triple Modular Redundancy riple Modular Redundancy (e.g., recommended by ISO 26262) ! Selective Selective and adaptive adaptive ! Resour Resource efficient ce efficient ! Peter Ulbrich – ulbrich@cs.fau.de 3

Software-Based Fault Tolerance Safety-Critical System ! ↯ ↯ Replica(1( Majority( Majority( Sensors( Interface( Interface( Replica(2( Actuators( Voter( Voter( Replica(3( ( Isola&on(domain( Sphere(of(redundancy((SOR)( ( ■ Software-based redundancy ! ■ Triple Modular Redundancy riple Modular Redundancy (e.g., recommended by ISO 26262) ! Selective Selective and adaptive adaptive ! Resour Resource efficient ce efficient ■ Single points of failur Single points of failure ! ■ Interface and Majority V Interface Majority Voter oter ■ Allowing for Silent Data Corruptions Silent Data Corruptions (SDC) (SDC) → Replication is impossible! Replication is impossible! Peter Ulbrich – ulbrich@cs.fau.de 3

�� Threats to Applicability – Mission failed? ■ Triple modular redundancy reliability ! R TMR = R Voter ⋅ R 2 − of − 3 ! ! Peter Ulbrich – ulbrich@cs.fau.de 4

� �� Threats to Applicability – Mission failed? � �� ■ Triple modular redundancy reliability ! R TMR = R Voter ⋅ R 2 − of − 3 ■ Voting on unreliable hardware? ! ■ Very small � residual err esidual error pr or probability? obability? ■ Risk analysis � inherently complex complex (no random error distribution! [4]) ! Peter Ulbrich – ulbrich@cs.fau.de 4

Software-based Fault Tolerance Mission (Im)possible? Peter Ulbrich - PowerPoint PPT Presentation

Software-based Fault Tolerance Mission (Im)possible? Peter Ulbrich The 29th CREST Open Workshop on Software Redundancy November 18, 2013 System Software Group

Roadmap for Section 10.1 The Notion of Fault-Tolerance Fault-Tolerance Support in NTFS Volume

Hypervisor-Based Fault-Tolerance Thomas C. Bressoud, Isis

Mission-critical 101 Resiliency Performance Fault Scalability Tolerance Disaster

Fault Tolerance in Message Passing Fault Tolerance in Message Passing and in Action and in

General Principles of Fault- Tolerance Daniel Gottesman Perimeter Institute Whats Left For

Rigorous fault-tolerance thresholds Ben Reichardt UC Berkeley N gate circuit 0/1 N gate

Lecture 10: Fault Tolerance Fault Tolerant Concurrent Computing The main principles of fault

On the Design of Fault-Tolerance in a Decentralized Software Platform for Power Systems Purboday

Algorithm-Based Fault Tolerance for Linear Algebra Thomas Herault University of Tennessee

CH NG 8: FAULT TOLERANCE TS. Tr n H i Anh Content 2 1. Introduction to fault

Distributed Systems 5. Fault Tolerant Systems Fault-Tolerance - 1 Lszl Bszrmnyi

CSci 5105 Introduction to Distributed Systems Fault Tolerance Last Time Replication and

Challenging Malicious Inputs with Fault Tolerance Techniques Bruno Luiz Agenda Threats

Fault tolerance based on the Publish- subscribe Paradigm for the BonjourGrid Middleware Heithem

Distributed Systems (ICE 601) Fault Tolerance Dongman Lee ICU Class Overview Introduction

A Survey of Software Fault Tolerance Techniques Authors: Zaipeng Xie, Hongyu Sun and Kewal Saluja

Welcome! Todays Agenda: Recap Flow Control AVX, Larrabee, GPGPU Further

Deep Learning for Natural Language processing Jindich Libovick March 1, 2017 Introduction

COSC 5351 Advanced Computer Architecture Slides modified from Hennessy CS252 course slides

openvswitch.ko minus Open vSwitch Joe Stringer, VMware

Commissioning of the ATLAS Tile Hadronic Calorimeter with cosmic muons, single beams and first

SoC SoC Design Design g Lecture L Lecture 3: Introduction to ASICs 3 I : Introduction to

AMMI Introduction to Deep Learning 6.3. Dropout Fran cois Fleuret

Transformers Pre-trained Language Models LING572 Advanced Statistical Methods for NLP March