A Survey of Software Fault Tolerance Techniques Authors: Zaipeng - PDF document

A Survey of Software Fault Tolerance Techniques Authors: Zaipeng Xie, Hongyu Sun and Kewal Saluja Dependable Software Design By: M. Abdollahi Azgomi Outline � Introduction � Types of recovery � Backward recovery � Forward recovery � Traditional Software Fault Tolerance � Design diversity � Recovery blocks (RcB) � N-version programming (NVP) � N-self checking programming (NSCP) � Data diversity � N-copy programming (NCP) � Retry blocks � Environmental diversity DSD - A Survey on Sowftare FT Techniques - By. M. A. Azgomi - IUST-CE 2 1

Introduction � What Is Software Fault Tolerance? � Provide service complying with the ) ﻦﻴﻣﺎﺗ ،ندروآﺮﺑ ( specification in spite of software faults DSD - A Survey on Sowftare FT Techniques - By. M. A. Azgomi - IUST-CE 3 Strategies of Software Fault Tolerance � Error Processing � Error Compensation ) ﺎﻄﺧ ناﺮﺒﺟ ( � Error Recovery ) ﺎﻄﺧ ﺖﻓﺎﻳزﺎﺑ ( � Forward Error Recovery (FER) � Backward Error Recovery (BER) � Fault Treatment ﺎﻄﺧ نﺎﻣرد ( ) DSD - A Survey on Sowftare FT Techniques - By. M. A. Azgomi - IUST-CE 4 2

Typical Software FT Techniques � Single version software FT techniques (mostly in self- checking software) � Multiple version software FT techniques (design diversity) � recovery blocks (RcB), � N -version programming (NVP) and � N self-checking programming (NSCP). � Multiple data representation techniques (data diversity) � retry blocks (RtB) and � N -copy programming (NCP) � Temporal diversity ) ﻲﻧﺎﻣز عﻮﻨﺗ ( � Environment diversity ) ﻲﻄﻴﺤﻣ عﻮﻨﺗ ( DSD - A Survey on Sowftare FT Techniques - By. M. A. Azgomi - IUST-CE 5 Types of Recovery � Error recovery is performed using � backward recovery or ) ﺐﻘﻋ ﻪﺑ ور ﺖﻓﺎﻳزﺎﺑ ( � forward recovery ) ﻮﻠﺟ ﻪﺑ ور ﺖﻓﺎﻳزﺎﺑ ( . DSD - A Survey on Sowftare FT Techniques - By. M. A. Azgomi - IUST-CE 6 3

Backward Recovery � Attempts to return the system to a correct or error- free state. � For transient faults � Example: recovery blocks (RcB) DSD - A Survey on Sowftare FT Techniques - By. M. A. Azgomi - IUST-CE 7 Forward Recovery � Attempts to find a new state from which the system can continue operation. � Utilize error compensation based on redundancy to select or derive the correct answer or an acceptable answer. � Examples: � N -version programming (NVP ), ) ﻲﺷرﺎﮕﻧﺪﻨﭼ ( � N -copy programming (NCP) and ) يا ﻪﺨﺴﻧﺪﻨﭼ ( � the distributed recovery block (DRB) DSD - A Survey on Sowftare FT Techniques - By. M. A. Azgomi - IUST-CE 8 4

Forward Recovery � Efficient � For Predictable Errors DSD - A Survey on Sowftare FT Techniques - By. M. A. Azgomi - IUST-CE 9 Traditional Software Fault Tolerance Software design and implementation errors cannot be � detected by simple replication of identical software units, assuming the same inputs are provided to each copy. Some form of diversity must accompany the � redundancy Software redundancy � Design diversity � Information or data redundancy � Data diversity � Temporal redundancy � Temporal diversity � Environment diversity � Hardware redundancy � DSD - A Survey on Sowftare FT Techniques - By. M. A. Azgomi - IUST-CE 10 5

Design Diversity � Higher cost DSD - A Survey on Sowftare FT Techniques - By. M. A. Azgomi - IUST-CE 11 Software Fault Tolerance Techniques Using Design Diversity Techniques Abbr. Error Processing Recovery Blocks RcB Error detection by AT and backward recovery N-Version Programming NVP Vote N Self-Checking NSCP Error detection by AT and Programming forward recovery DSD - A Survey on Sowftare FT Techniques - By. M. A. Azgomi - IUST-CE 12 6

RcB DSD - A Survey on Sowftare FT Techniques - By. M. A. Azgomi - IUST-CE 13 Recovery Blocks Method Recovery block Error Processing Error detection by AT and backward Technique recovery Criteria of Accepting Absolute, with respect to specification Result Execution Scheme Sequential DSD - A Survey on Sowftare FT Techniques - By. M. A. Azgomi - IUST-CE 14 7

NVP DM: Decision � Mechanism DSD - A Survey on Sowftare FT Techniques - By. M. A. Azgomi - IUST-CE 15 N-version Programming Method N-version programming Error Processing Vote Technique Criteria of Accepting Relative, on variant results Result Execution Scheme Parallel (on NMR hardware) DSD - A Survey on Sowftare FT Techniques - By. M. A. Azgomi - IUST-CE 16 8

NSCP DSD - A Survey on Sowftare FT Techniques - By. M. A. Azgomi - IUST-CE 17 N Self-Checking Programming Method N self-checking programming Error Processing Error detection and result switching Technique Then, Detection by comparison or by AT(s) Criteria of Accepting Relative, on variant results or Result Absolute with respect to specification Execution Scheme Parallel DSD - A Survey on Sowftare FT Techniques - By. M. A. Azgomi - IUST-CE 18 9

Data Diversity � To complement design diversity � Using data re-expression algorithms (DRA) to obtain logically equivalent variants of the input data Data re-expression via decomposition and recombination DSD - A Survey on Sowftare FT Techniques - By. M. A. Azgomi - IUST-CE 19 Software Fault Tolerance Techniques Using Data Diversity Software Fault Abbr. Error Processing Tolerance Techniques Retry Blocks RtB Acceptance test and Backward recovery N-Copy Programming NCP Run the same process concurrently or sequentially DSD - A Survey on Sowftare FT Techniques - By. M. A. Azgomi - IUST-CE 20 10

Retry Blocks Error Pass DSD - A Survey on Sowftare FT Techniques - By. M. A. Azgomi - IUST-CE 21 Retry Blocks Method Retry blocks Error Processing Error detection by AT and backward Technique recovery by DRA Criteria of Accepting Absolute, with respect to specification Result Execution Scheme Sequential DSD - A Survey on Sowftare FT Techniques - By. M. A. Azgomi - IUST-CE 22 11

NCP Inputs are � re-expressed DSD - A Survey on Sowftare FT Techniques - By. M. A. Azgomi - IUST-CE 23 N-copy Programming Method N-copy programming Error Processing Decision mechanism (DM) and Technique forward recovery Criteria of Accepting Relative, on variant results Result Execution Scheme Parallel DSD - A Survey on Sowftare FT Techniques - By. M. A. Azgomi - IUST-CE 24 12