Tools

Software Reliability Tools
Reliametrics tools are designed to help development organizations design, build, test, and implement software applications. We know that many developers have failed to realize their full benefits because of time-to-market pressures, scarce resources, lack of training, or inexperience.

We overcome these challenges by providing leading-edge tools and services that are easy to use and help achieve higher quality and more reliable software. We also provide the instruction, guidance and support needed to take the greatest advantage of our products.

Here are some of the tools and processes we can provide to help you build better software.

Tools

EMERALD (Enhanced Measurement for Early Risk Assessment of Latent Defects) EMERALD is a client based server application with an Oracle database that contains leading-edge risk models based on factors including static characteristics of source code, dynamic problem data, and customer usage. EMERALD analyzes code and assigns risk levels indicating likelihood of fault occurrence in color--green (no risk), yellow (moderate risk) and red (high risk). EMERALD enhances software development, testing, and maintenance by integrating measurements, models, and risk analysis in desktop reports.

EMERALD has been developed and used extensively by Nortel Networks. If code is not yet written, EMERALD will create estimates based on the code's complexity, size, and other structured analysis factors.

GURU
GURU is a suite of software development tools to answer the need for a scalable, high-performance, and integrated development environment. This tool provides cross-reference tracking, hierarchy browsing, string searching, source viewing, and image file querying. GURU tools facilitate management, search, view, edit, and debug functions. GURU supports object oriented and procedural source code and includes an information model database created by analyzing your source code.

Methods and Models

FMECA (Failure Modes and Effect Criticality Analysis) Techniques
The Failure Modes and Effects Criticality Analysis (FMECA) methodology for software is being developed in co-operation with the University of Tennessee Software Quality Research Lab and NSPaN (Nortel Subsystems and Performance Networks). Reliametrics has defined a method for capturing the functional failure modes for software and incorporating them into an operational profile or usage model of the software functions. This approach enables statistical testing of the software and proves the reliability of the software operating in a network node prior to delivery

JUMBL Model
JUMBL is a toolkit to support statistical testing of software using Markovian models. The JUMBL Model will produce the system (hardware and software) reliability assessment and provide input to the network availability assessment. This tool will also provide automatically generated test cases. It has been prepared by the University of Tennessee with support from Nortel Networks and has been used by IBM and other companies.

NSPaN Hardware Availability Model
This modeling tool was developed by NSPaN to validate the reliability, performance, and interoperability of applications residing on a network provided by Nortel Networks.

Kullback Discriminant
The Kullback is a theoretic measure between usage models and testing records showing how closely testing experience matches expected field use. There are no absolute values for acceptable "closeness," but it is possible to observe the gap closing and to simulate the amount of testing required to close the gap to a specific degree.

Sayre's Probability of Approximate Equality
This is a variation on the Kullback measure, which results in a probability value. Given an acceptable gap between testing experience and expected field use, this tool measures the probability that testing experience is within tolerance.

Tools

NSPaN Hardware Availability Model This modeling tool was developed by NSPaN to validate the reliability, performance, and interoperability of applications residing on a network provided by Nortel Networks.

Sayre's Probability of Approximate Equality This is a variation on the Kullback measure, which results in a probability value. Given an acceptable gap between testing experience and expected field use, this tool measures the probability that testing experience is within tolerance.

NSPaN Hardware Availability Model
This modeling tool was developed by NSPaN to validate the reliability, performance, and interoperability of applications residing on a network provided by Nortel Networks.

Sayre's Probability of Approximate Equality
This is a variation on the Kullback measure, which results in a probability value. Given an acceptable gap between testing experience and expected field use, this tool measures the probability that testing experience is within tolerance.