Rare conditions and their effect on software failures

Data gathered by the authors as well as those published by others show that rarely executed code has a much higher failure rate (expressed in execution time) than frequently executed code during the early operational period. More detailed analysis of the data shows that the inability to handle multiple rare conditions, such as response to hardware failures or exception conditions caused by the computer state, is a prominent cause of program failure in well-tested systems. Approaches to coping with this difficulty are discussed.<<ETX>>

[1]  Daniel P. Siewiorek,et al.  WORKLOAD, PERFORMANCE, AND RELlABlLlTY OF DIGITAL COMPUTlNG SYSTEMS , 1995, Twenty-Fifth International Symposium on Fault-Tolerant Computing, 1995, ' Highlights from Twenty-Five Years'..

[2]  Paola Velardi,et al.  A Study of Software Failures and Recovery in the MVS Operating System , 1984, IEEE Transactions on Computers.

[3]  F. J. Redmill Dependability of Critical Computer Systems 1 , 1988 .

[4]  David F. McAllister,et al.  An Experimental Evaluation of Software Redundancy as a Strategy For Improving Reliability , 1991, IEEE Trans. Software Eng..

[5]  Daniel P. Siewiorek,et al.  Workload, Performance, and Reliability of Digital Computing Systems. , 1980 .

[6]  Ravishankar K. Iyer,et al.  A STATISTICAL LOAD DEPENDENCY MODEL FOR CPU ERRORS AT SLAC , 1995, Twenty-Fifth International Symposium on Fault-Tolerant Computing, 1995, ' Highlights from Twenty-Five Years'..