The Software Reliability Dataset
The Software Reliability Dataset was compiled by John Musa of
Bell Telephone Laboratories. His objective was to collect failure
interval data to assist software managers in monitoring test status
and predicting schedules and to assist software researchers in
validating software reliability models. These models are applied in
the discipline of
Software Reliability Engineering.
The dataset consists of software failure data on 16
projects. Careful controls were employed during data collection to
ensure that the data would be of high quality. The data was collected
throughout the mid 1970s. It represents projects from a variety of
applications including real time command and control, word processing,
commercial, and military applications.
The DACS distributes this
dataset on a floppy disk. For each software failure in the dataset,
the following items are recorded:
- Project Identification (System Code). An
internally assigned identification number.
- Failure Number. A number identifying a particular
failure. Failures are consecutively numbered from the first failure
recorded.
- Failure Interval. The time elapsed from the previous
failure to the current failure. For project 6, this time is given in
CPU seconds; for the remaining projects, the time is given in
wall-clock seconds. A negative sign for the last failure interval in
a data set indicates that this interval represents the time from the
previous failure to the end of execution of the software.
- Day of Failure. Represents the day on which the
failure occurred in terms of the number of working days from the start
of the current phase or data collection period.
The following table displays the size of each sample of failure data for each project in the dataset, as well as other useful information concerning the individual projects. More detailed information on the specific characteristics of each project is available in
John D. Musa, Software Reliability Data,
Data & Analysis Center for Software, January 1980.
Data is available by system code. Select the system code:
Software Reliability Data Project Information
| System Code | Application | Size (Delivered Object Code Instructions) | Number of Failures | Phases Represented By Sample |
| 1 | Real Time Command & Control | 21,700 | 136 | System Test Operations |
| 2 | Real Time Command & Control | 27,700 | 54 | System Test Operations |
| 3 | Real Time Command & Control | 23,400 | 38 | System Test Operations |
| 4 | Real Time Command & Control | 33,500 | 53 | System Test Operations |
| 5 | Real Time Commercial | 2,445,000 | 831 | System Test * |
| 6 | Commercial Subsystem | 5,700 | 73 | Subsystem Test |
| 14C | Real Time | (Hundreds of Thousands) | 36 | Operations |
| 17 | Military | 61,900 | 38 | System Test |
| 27 | Military | 126,100 | 41 | System Test |
| 40 | Military | 180,000 | 101 | System Test |
| SS1A | Operating System | (Hundreds of Thousands) | 112 | Operations * |
| SS1B | Operating System | (Hundreds of Thousands) | 375 | Operations * |
| SS1C | Operating System | (Hundreds of Thousands) | 277 | Operations * |
| SS2 | Time Sharing System | (Hundreds of Thousands) | 192 | Operations * |
| SS3 | Word Processing System | (Hundreds of Thousands) | 278 | Operations * |
| SS4 | Operating System | (Hundreds of Thousands) | 196 | Operations * |
Note: Starred phases (*) indicate the failure sample is not complete for that phase.
|
DACS Gold Practice Initiative
