Software Development Life Cycle (SDLC)

SDLC is a problem-solving procedure to examine an information system and identify appropriate processes to improve it.

Four common SDLC models: Waterfall Model Incremental Waterfall Model Prototyping Model Spiral Model

Accenture Development Method (ADM) is a standard approach or framework for delivering high quality solutions and services to the client.

SDLC in Accenture includes: Plan Analyze Design Build Test Deploy

The ADM V-model is a systems development model designed to simplify the understanding of the complexity associated with developing systems.

Each phase in the ADM V-model has a complementary Verification/Validation phase. Four major phases of a standard project are:

Project Initiation Project Planning Project Execution Project Closure

The phases of Product Life Cycle: Idea or innovation Development Introduction or adaptation Growth Maturity Decline

Introduction to Testing

The goal of testing is to ensure that the business system meets the established functional and technical requirements.

The different types of software testing are (according to the Accenture Testing Framework):

Unit testing - Tests individual units of the application. Assembly testing - sometimes called String test. Ensures that the

interactions between the components function correctly. Product testing - sometimes called System test, as it tests that all business

requirements are met by the system. • The Application Product test validates that the product requirements

are met by each individual application.• The Integration Product test is an end-to-end test of the product

requirements across all applications and platforms. After successfully completing the application product test, integration product test can occur.

Performance testing - Is carried out to ensure that a release is capable of operating at the load levels specified in the business performance requirements and any agreed-on Service Level Agreements (SLAs) such as (load, stress, endurance/ volume, and throughput)

Acceptance testing - Ensures that the users and stakeholders are satisfied with the solution.

• User Acceptance Testing – Testing with the client.• Operational Acceptance Test - focuses on the operations/support

users to perform a set of tests to verify that the system operates as intended. It verifies that the correct functionality, architecture, and procedures are defined and established to allow production support teams to run, maintain, and support the system in production in accordance with common practices, Standard Operating Procedures (SOP) or defined SLAs, where applicable.

Deployment testing - Tests the production environment’s readiness to handle the new system.

The Testing Framework helps the software development project teams to plan and execute testing tasks.

Quality Management (QM) is a method for ensuring that all the activities necessary to design, develop, and implement a product or service are effective and efficient with respect to the system and its performance.

There are four QM programs at Accenture: Quality and Process Improvement (QPI) - Provides projects with standard

processes, tools, coaching, and training on systems/software engineering and project management disciplines.

Capability Maturity Model Integration (CMMI) - Provides best practices covering a product’s life cycle that are organized into process areas. It is used for continuous improvement of the development practices.

Delivery Excellence (DEX) - Establish process standards and maintain process metrics and quality criteria. Provide key process education.

Software Quality Assurance (SQA) - Verifies that processes and deliverables/work products comply with project standards and procedures. Reviews support the delivery of high-quality products and services.

Accenture Delivery Suite (ADS) Overview

Accenture Delivery Suite (ADS) is an integrated set of components that increases the reliability, quality, speed, and predictability of the end-to-end delivery work.

There are four components of ADS: Accenture Delivery Methods (ADM) and ADM Estimators - provides

projects with common starting points such as roles and activities. Projects can then customize to specifications without breaking the tight integration among ADS components. ADM Estimators enable the process of determining the probable cost or effort of something, using historical information, experience, judgement, practice, analysis, and other techniques.

Accenture Delivery Tools (ADT) - a set of standard industry-leading tools to support application development and work maintenance.

Accenture Delivery Architectures (ADA) - Accenture Delivery Metrics - metric is the measurement of attributes that

allows comparison or prediction of a process or a product. Use of Accenture Delivery Metrics enables engagements to better understand when execution is within normal operating tolerance.

ADM is a collection of Methods that cover management, strategy and planning, development, and operations.

The Testing Framework helps the software development project teams to plan and execute testing tasks.

Accenture Delivery Tools (ADT) are standard tools that automate and simplify processes, tasks, and activities.

ADS Practices are common management and development processes used so projects can perform consistently across all workforces for multi-service, multi-location deals.

Accenture Delivery Metrics is a set of attributes that help engagements improve delivery performance and support the improvement of key delivery processes.

Requirements and Test Strategy

A requirement is a condition or capability to which a system must perform. Good requirements are clear, concise, have a tangible benefit, and fully capture

the customers’ expectation. Bad requirements may be suggestive, relative, ambiguous, or unrealistic. The customer or their Domain Expert is the source of requirements. The Business Analyst manages and controls the requirements. Requirements Traceability captures all functional and technical requirements and

can easily track whether requirements are covered by specific work products. The Test Strategy documents the overall strategy for testing the application,

technical architecture, and training and performance support. Accenture Requirements Engineering Suite (AcRES) - Created by Accenture

Technology Labs, in conjunction with Accenture Delivery Tools (ADT) and the Requirements and Analysis Capability. Composed of three tools:

Accenture Requirements Analysis Tool (RAT): Increases the quality of requirements and reduces project rework by analyzing requirements text to help you bring them into alignment with Accenture best practices for clear requirements. It highlights potential trouble spots and provides suggestions on how risks can be addressed. The newest version also provides enhanced visualizations – Visio diagrams generated automatically to help you see the interactions depicted in you requirements text.

Requirements Glossary Assistant (RGA): Automates much of the work of populating entity glossaries by scanning requirements documents and suggesting potential entity glossary entries automatically.

Requirements Documentation Assistant (RDA): Helps reduce the effort to write requirements by making it easy to create well-structured requirements the first time/every time. It provides a dynamic wizard allowing users to focus on the requirements content; it then automatically generates a well-formed requirements sentence from that content.

Testing Technique

In the ADM for Testing perspective, both Discrete Testing and Embedded Testing (or projects for testing in development projects), share the common disciplines Analyze Test, Design Test, Prepare Test, and Execute Test.

In the Analyze stage, the Test Team will Identify Application Test Requirements, plan for Functional and Non-functional Testing, Define the Test Environment and Data, and Confirm Test Analysis Deliverables.

Requirements guide the design, building, and testing of the product. And are initially developed during the Solution Planning/Plan stage.

Requirements and Traceability Work Product is used to maintain bi-directional traceability from the application requirement items back to the high-level business requirements and to the downstream analysis, design, build, and test components.

Product Testing is the first point where the solution’s functional and business requirements are tested together.

Assembly Test, ensures that related application components function properly when assembled.

When a Production system is updated with new functionality, existing code must be tested to ensure the new functionality will not adversely impact it. This is called a Regression Test.

UAT proves to the responsible business units that the application performs as expected and meets all the customer and product requirements included in the current release.

Non-Functional Testing is a testing of an application for its non-functional requirements.

Testing Work Products are created during Test Planning, Design and Preparation. These work products are:

Test Approach (Analyze Test) - Test Scenarios (Design Test) Test Conditions and Expected Results (Design Test) Test Cycle Control Sheet (Analyze Test) Test Scripts (Prepare Test)

Regression Testing Techniques Retest-All Technique - All test cases are executed. Regression Test Selection Technique - Only a number of selected test cases

are to be re-executed. Dataflow Technique - Only test cases where the data flow (or data

interactions) of the affected modifications are to be re-executed. Test Case Prioritization Technique - The re-execution of some test cases

that depends on a number of factors such as criticality of the module, impact to the business, and modules where there are more user interactions.

Test Planning Overview

The purpose of the Design Test discipline is to create the test conditions and expected results, test scenarios, and test cycle control sheets for all of the major test streams.

Some of the most important testing techniques are: Static Dynamic Positive Negative Black Box White Box Grey Box

Static Testing is also called as Dry Run Testing. Dynamic Testing requires the code to be executed. Positive Testing is performed in the system by providing valid data input. Negative Testing implies “test to fail.”

Black Box testing does not require the knowledge of the code or internal structure.

Performance Engineering is the systematic process by which performance of an application is ensured.

White Box Testing tests the structure of the code. Grey Box Testing is a combination of Black Box and White Box testing. Litmus is a test design artifact generation tool that semi-automates the test

design phase.

Applied Statistics

Applied Statistics testing is a statistical approach to Test Design that provides maximum coverage with a minimum number of test scripts.

Applied Statistics provides three key benefits during the Test Design phase: Addresses common challenges in the Test Design process. Provides a structured approach to Test Design. Provides actionable information to determine the appropriate amount of

testing. The Applied Statistics process involves the following key steps:

Define parameters and values. Prepare and generate tests. Analyze and adjust coverage. Confirm expected results and scripts. Review and sign-off. Update test execution plan.

Two tools on the market that use Applied Statistics for test design are: Hexawise, a part of Accenture Delivery Tools (ADT):

• Is a software Test Design tool.• Ensures all important combinations are tested by maximizing test

coverage with minimum number of test conditions..

AllPairs Testing or Pair-wise Testing:• Is a combination of sets to be tested.• Is a method that finds all double-mode faults that are two parameters

conflicting with each other.

The following are the features of the Hexawise Test Design tool: Creates test conditions quickly and easily. Flexible to use on any kind of test project. Generates the smallest possible number of tests to meet the coverage

requirements. Automatically maximizes the variation between tests and minimizes the

repetition in tests.

Test Prep Overview

Test Preparation for the ADM Perspective involves three key steps: Prepare Test Scripts and Data:

• Expand the Test Deliverables by defining Input Data• Refine the Expected Results of the Test Conditions• Create Test Scripts Deliverable from the refined TCERs• Prepare and Finalize Test Cycle Schedule (TCCS)

Prepare Test Environment:• Establish the Test Environment• Prove that the Test Environment is working before the Test

Execution Manage Test:

• Periodically measure the test progress• Periodically report the test progress• Document progress

Test scripts detail the exact steps a Tester must follow while executing a test. Test data is a data that has been specifically identified for the purpose of testing a

particular system or application being developed. Application Product test (ADM task 4720) is designed to validate the functional

requirements of the system’s internal operations.

ADM task 4720 comprises four steps: Prepare Application Product Test scripts. Prepare Application Product Test script transaction data. Prepare Application Product Test script expected results.

Perform peer review. Integration Product Test is an end-to-end test of the product requirements for all

applications and platforms. ADM task 4725 comprises four steps:

Prepare integration product test scripts. Prepare integration product test script transaction data. Prepare integration product test script expected results. Perform peer review.

Performance test (ADM task 4730) fixes performance- related issues before the system goes live.

ADM task 4730 comprises four steps: Prepare performance test scripts. Prepare performance test script transaction data. Prepare performance test script expected results. Perform peer review.

User Acceptance test (UAT) is performed by the client. UAT is generally taken up as the last phase of testing. UAT (ADM task 4750) is devised to answer the important question ‘does the

system enable the users to do their jobs?’ ADM task 4750 has four steps:

Prepare user acceptance test scripts. Prepare user acceptance test script transaction data. Prepare user acceptance test script expected results. Perform peer review.

Operational Acceptance Test (ADM task 4755) is meant to validate the usability and operability of the systems involved in the test.

ADM task 4755 has four steps: Prepare Operational Acceptance Test scripts. Prepare Operational Acceptance test script transaction data. Prepare operational acceptance test script expected results. Perform peer review.

Automation (ADM task 4760) involves the designing of automation frameworks and components based on the automation test scripts.

ADM task 4760 has four steps: Create test automation scripts.

Update transaction data and common test data. Perform peer review.

ADM Task 4791: Review Test Script Deliverables involves review the deliverables with the various stakeholder groups.

ADM Task 4793: Start Assembly Test to Verify Readiness involves loading the test scripts and test data into the test execution tools/environments.

ADM Task 4795: Finalize Changes and Revise Assembly Test Estimates involves reviewing the work plan for the Assembly Test.

The purpose of ADM Task T4799: Transition Test Script Deliverables is to transition the test preparation deliverables to the Test Execution team.

Rational Functional Tester (RFT) is an object-oriented testing tool that tests Java, HTML, VB.net, and Windows applications.

Quick Test Professional (QTP) is an advanced solution for functional and regression test automation.

Peer Review

Verification and Validation are two important review methods. Verification ensures that ‘you build the product right’ Validation ensures that ‘you build the right product’.

Peer reviews: Promote early detection and removal of defects from a deliverable Facilitate knowledge transfer Provide an understanding of the content Gain inputs from the reviewer Encourage continuous improvement Identify areas of improvement in overall processes Support overall quality and solution delivery excellence Reduce time and costs resulting from rework.

The roles involved in a peer review: Project Manager Deliverable Owner (Author) Reviewer Moderator (Optional).

The phases in the Peer Review process are:

Planning for the Peer Review Scheduling the Peer Review Preparing for the Peer Review Conducting the Peer Review Performing Rework Reviewing Changes Reviewing and Submitting Team Results Analyzing Project Results.

The Peer Review process is followed on all system development lifecycle deliverables:

Requirements, Design, Build, Test, and Deploy. Metrics are processed values that help to quantitatively evaluate the efficiency of

the Peer Review process. Metrics are interpreted by comparing them with a set of baseline targets.

Peer Review

The Test Environment is a setup of the software and hardware on which the Testing Team is going to test the newly built software product or application.

Common test data provides control over the quality of testing, which significantly improves the productivity of the testing effort.

Test Data Management helps define standards, tasks, ownership, roles, and responsibilities to ensure the accuracy, completeness, and integrity of test data.

The five areas covered in an effective Test Data Management Approach are: Data Identification - the data items which the project will require in order

to test the application and state sources of this data and any dependencies which exist.

Data Acquisition - the methods used to acquire data, such as data stubs, data mining, data extracts and data conversion.

Data Conditioning - Processes which need to be followed on the acquired data such as scrubbing, modification, or transformation.

Data Population - the processes to define how the data will be populated to test environments (for example, will it be deployed with a build).

Consideration should be given to versions of data if multiple versions need to be active and deployable at one time.

Data Maintenance - Detail how this data model will be maintained and updated. The baseline recommendation is the project’s standard defect management process.

A poorly managed test environment can impact timelines, quality, and budget. When dealing with Test Data and Environments, the Test Strategy determines:

What environments are available? Environment configurations and test data management. What environments are dedicated or shared? Approach to managing test data and environments.

Poor test data management compromises the efficiency testing process of an enterprise in many ways.

Benefits of Test Data Management Tools: Test Data Management tools enable standardized and easy creation of

test data. Users can easily view and edit data in a familiar environment. Test data created can be checked to make sure all table links are valid. Production data can be reused and duplicated many times. Test cases can be created to test invalid data.

Test Execution Overview

To execute assembly test, follow the steps below: Execute Assembly Test Manage Defects Re-test Fixes Create Test Closure Memo

Ensure that entry criteria have been met to start any test execution. Test Closure Memo summarizes the

testing activities. A defect is any deviation from the end client requirement that manifests during a

project.

Defect Management is the process of capturing, reviewing, assigning, fixing, deferring, rejecting and closing defects throughout the system’s life cycle.

SIR is the Key form of communication between the Tester and Fix-it team. The four most typically-used SIR Priority Levels are: Critical, High, Medium, Low. Defect meetings should include:

Agenda Roles and Responsibilities Meeting Minutes Representation Schedule Defect Reports

Other Testing Methodologies

The types of testing used by agile development teams include:

Automated testing - using an automation tool to execute test case suite for an application. In this technique, the automation tool enters test data into the system under test, compares expected and actual results, and generates detailed test reports.

Regression testing - is a full or partial selection of already executed test cases which are re-executed to ensure existing functionalities work fine. Regression testing is done to make sure that new code changes should not have side effects on the existing functionalities. It ensures that old code still works once the new code changes are done.

Acceptance testing - The focus of Acceptance testing is not to find defects but to check whether the software application meets the project specifications or customer requirements, since this is the first time, when the client sees their requirements which was in plain text into an actual working system.

The Agile testing life cycle has four phases: Initiation - project’s foundation is established. The Initiation phase is

usually of a relatively short duration and involves establishing a project’s foundation. It is essential to use the Initiation phase for projects that have

fixed deadlines to determine when a project needs to enter the End Game phase.

Construction - The system is developed in an evolutionary or iterative and incremental manner. Most of the testing conducted during an agile software development project occurs during the Construction phase.

End Game - The system is transitioned into production. During the End Game phase, some critical software systems may often require Final or Full System tests and Acceptance tests. For example, life-critical systems, such as medical software.

Production – involves operating systems and providing support for its users. The goal of the Production phase is to have the products useful and productive post deploy to the users. If the support for the release is ended or if the release of the newer version is getting ready, then the Production phase will be ended.

Most tests occur during the Construction phase. During the End Game phase, a team may need to perform full System and

Acceptance tests. Risk-based testing is a testing technique where tests are designed on the concept

of refutability. In Risk-based testing, Testers design tests to challenge their basic or critical

assumptions. Model-based testing derives tests from models. Models are created and used to

represent the desired behavior of SUT. Scenario-based testing is based on a scenario that revolves around how the

application will be used in a real-world situation. Domain-based testing is used to ensure that the application only accepts valid

input.

More Tools of the Trade

SQL is a standard language for accessing and manipulating databases. The commonly used SQL commands are:

SELECT: Used for selecting data from a database WHERE: Used for extracting only specific records

DELETE: Used for deleting a row in a table UPDATE: Used for updating the records in a table

INSERT: Used for inserting a new row in a table Depending on the functionality, SQL commands are classified into four different

categories, such as: Data Definition Language (DDL) - A language used by a database

management system, which allows users to define the database and its objects, specifying data types, structures, and constraints on the data.

Data Manipulation Language (DML) - a language that enables users to access or manipulate data as organized by the appropriate data model. DML commands are used to query and manipulate existing objects like tables.

Data Control Language (DCL) - a language that provides users with privilege commands. The segment of SQL used or controlling access to data in a database. DCL allows protecting the tables and other objects created by a user from accidental manipulation by another user. DCL allows to control privileges assigned to users.

Transaction Control Language (TCL) - Transactions are a collection of operations that form a single logical unit of work. Transactions ensure that all changes made are either committed to the database or if there is a failure at any point of time then none of the changes are committed in the database. Transaction changes can be made permanent to a database only if they are committed.