S. Mourad and D. Andrews, "On the Reliability of the IBM MVS/XA Operating System," IEEE Trans. on Software Engineering, October 1987.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....in timesharing systems and proposed a model for software reliability. Iyer [7] examined the effect of workload on the reliability of the IBM 3081 operating system. The study proposed and validated a load hazard model in order to measure the risk of system failure with increasing workload. Mourad [16] performed a reliability study on the IBM MVS XA operating system and found that the error distribution is heavily dependent on the type of system utilization. Meyer [13] presented an analysis of the influence of workload on the dependability of computer systems. Maxion [14] 15] presented useful ....

S. Mourad and D. Andrews, "On the Reliability of the IBM MVS/XA Operating System," IEEE Trans. on Software Engineering, October 1987.

....did not eliminate correlated faults. Both the studies evaluated only the error detection capabilities of N version programming. They did not evaluate whether N version programming can detect errors early enough to prevent fail stop violations. 50 A study done on the reliability of the IBM MVS XA [Mourad85] operating system looked at effectiveness of recovery routines. They found that recovery routines could recover the system from 88 92 of the failures. These recovery routines were specifically written for the operating system and the operating system software was also better tested than general ....

Samiha Mourad and Dorothy Andrews. The Reliability of the IBM MVS/XA Operating System. In Proceedings of the 1985 Symposium on Fault-Tolerant Computing (FTCS), June 1985.

....exercise defective code that the testing procedure has missed. Unfortunately, most of the published literature on field failures concentrates on reliability metrics and trends (for example, Levendel89] rather than error characteristics. A number of studies on field failure done at Stanford [Mourad87] Iyer86] and CMU [Castillo81] used data from error logs to track system failures. Error log records are generated by the system after programs fail with an abnormal end (ABEND) completion code. Although error logs accurately count failure frequency, they give little semantic information about ....

S. Mourad and D. Andrews. On the Reliability of the IBM MVS/XA Operating System. IEEE Trans. on Software Engineering, 13(10):1135-1139, Oct 1987

....run at the same time. This leads to an increased number of dynamic interactions between processes, increasing the potential for stimulating any latent parallelism related faults in the operating system code. The relationship between system workload and software error rate is well documented [CaS82, MoA87]. Put together, the increased rate of hardware errors, the increased use of parallel constructs in the operating system code, and the increased dynamic parallelism of the system suggest that the MTBF of large multiprocessors will be substantially lower than that of current multiprocessors if ....

....of faults that are found and fixed during operating system development. Another area of work has been techniques for avoiding the reboot that is required to recover from software errors in standard SMP kernels. Both the Nonstop UX kernel for Tandem Integrity S2 [Jew91] and the IBM MVS XA system [MoA87] provide data structure specific repair routines invoked when an inconsistency is detected. Another approach, implemented in MVS XA and Fault Tolerant Mach [RSS93] is to abort the operation in progress when an error is detected and 118 10.2 Multiprocessor operating systems retry it. HJK93] ....

[Article contains additional citation context not shown here]

S. Mourad and D. Andrews. "On the reliability of the IBM MVS/XA operating system." IEEE Transactions on Software Engineering, vol. SE-13, no. 10, pp. 1135--1139, October 1987.

....in a given piece of code. Managers can use this information to make decisions about release dates, but it is not the kind of information that can be used to evaluate potential error detection or recovery strategies. Several studies used data from error logs to track failures at customer sites [58][39] 15] Error log records are generated automatically by the system after a program fails. Because the log entries are generated automatically, they give extremely high level representations of the error. For example, the log entry might be a code indicating that the program tried to store into ....

....pointers, make up 25 to 30 percent of the APARs filed against IMS, MVS, and DB2. This is consistent with several other studies of software errors in operating systems summarized in Figure 2.5. The published studies in the figure are from DOS VS [24] and MVS (one from [79] and two from [58]) The Unpublished I study was a survey of errors reported in the 4.1 4.2 releases of BSD UNIX [72] The Unpublished II CHAPTER 2. A SURVEY OF SOFTWARE ERRORS IN SYSTEMS PROGRAMS 38 operating system error study was conducted internally at a company that would not allow the release of its name. ....

S. Mourad and D. Andrews. On the reliability of the ibm mvs/xa operating system. IEEE Transactions on Software Engineering, SE-13(10):1135--1139, October 1987.

No context found.

S. Mourad and D. Andrews. "On the reliability of the IBM MVS/XA operating system." IEEE Transactions on Software Engineering, vol. SE-13, no. 10, pp. 1135-- 1139, October1987.

No context found.

S. Mourad, D. Andrews, "On the Reliability of the IBM MVS/XA Operating System," IEEE Trans. on Software Engineering, SE13 (10):1135-1139, October 1987.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC