Thursday, November 25, 2010

Software Testing Life Cycle(STLC)


Software Testing Life Cycle
Software Testing Life Cycle consists of six phases:
1. Test Planning ( Product Definition Phase):

The test plan phase mainly signifies preparation of a test plan. A test plan is a high level planning document derived from the project plan (if one exists) and details the future course of testing. Sometimes, a quality assurance plan - which is more broader in scope than a test plan is also made.
Contents of a test plan are as follows :
Scope of testing
Entry Criteria (When testing will begin?)
Exit Criteria (When testing will stop?)
Testing Strategies (Black Box, White Box, etc.)
Testing Levels (Integration testing, Regression testing, etc.)
Limitation (if any)
Planned Reviews and Code Walkthroughs
Testing Techniques (Boundary Value Analysis, Equivalence Partitioning, etc.)
Testing Tools and Databases (Automatic Testing Tools, Performance testing tools)
Reporting (How would bugs be reported)
Milestones
Resources and Training

Contents of a SQA Plan, more broader than a test plan, are as follows :
The IEEE standard for SQA Plan Preparation contains the following outline :
Purpose
Reference Documents
Management
Documentation
Standards, Practices and Conventions
Reviews and Audits
Software Configuration Management
Problem Reporting and Corrective Action (Software Metrics to be used can be identified at this stage)
Tools, Techniques and Methodologies
Code Control
Media Control
Supplier Control
Records, Collection, maintenance and Retention
2.Test Analysis (Documentation Phase)
The Analysis Phase is more an extension of the planning phase. Whereas the planning phase pertains to high level plans - the Analysis phase is where detailed plans are documented. This is when actual test cases and scripts are planned and documented. This phase can be further broken down into the following steps :
Review Inputs : The requirement specification document, feature specification document and other project planning documents are considered as inputs and the test plan is further disintegrated into smaller level test cases.
Formats : Generally at this phase a functional validation matrix based on Business Requirements is created. Then the test case format is finalized. Click here to see a format for a test case. Also Software Metrics are designed in this stage. Using some kind of software like Microsoft project - the testing timeline along with milestones are created.
Test Cases : Based on the functional validation matrix and other input documents, test cases are written. Also some mapping is done between the features and test cases. Click here to download a database that maps requirements to features to test cases. In case whit box testing is to be done, then the
Plan Automation : While creating test cases, those cases that should be automated are identified. Ideally those test cases that are relevant for Regression Testing are identified for automation. Also areas for performance, load and stress testing are identified.
Plan Regression and Correction Verification Testing : The testing cycles, i.e. number of times that testing will be redone to verify that bugs fixed have not introduced newer errors is planned.
3. Test Design (Architecture Document and Review Phase)
One has to realize that the testing life cycle runs parallel to the software development life cycle. So by the time, one reaches this phase - the development team would have created some code or at least some prototype or minimum a design document would be have been created.
Hence in the Test Design (Architecture Document Phase) - all the plans, test cases, etc. from the Analysis phase are revised and finalized. In other words, looking at the work product or design - the test cases, test cycles and other plans are finalized. Newer test cases are added. Also some kind of Risk Assessment Criteria is developed. Also writing of automated testing scripts begin. Finally - the testing reports (especially unit testing reports) are finalized. Quality checkpoints, if any, are included in the test cases based on the SQA Plan.
4. Test Execution (Unit / Functional Testing Phase)
By this time. the development team would have been completed creation of the work products. Of Course - the work product would still contain bugs. So, in the execution phase - developers would carry out unit testing with testers help, if required. Testers would execute the test plans. Automatic testing Scripts would be completed. Stress and performance Testing would be executed. White box testing, code reviews, etc. would be conducted. As and when bugs are found - reporting would be done.
5. Test Cycle (Re-Testing Phase)
By this time, minimum one test cycle (one round of test execution) would have been completed and bugs would have been reported. Once the development team fixes the bugs, then a second round of testing begins. This testing could be mere correction verification testing -- that is checking only that part of the code that has been corrected. It could also be Regression Testing - where the entire work product is tested to verify that correction to the code has not affected other parts of the code. Hence this process of :
Testing --> Bug reporting --> Bug fixing (and enhancements) --> Retesting
is carried out as planned. Here is where automation tests are extremely useful to repeat the same test cases again and again.
During this phase - review of test cases and test plan could also be carried out.

6. Final Testing and Implementation (Code Freeze Phase)
When the exit criteria is achieved or planned test cycles are completed - then final testing is done. Ideally - this is System or Integration testing. Also any remaining Stress and Performance testing is carried out. Inputs for process improvements in terms of software metrics is given. Test reports are prepared. if required, a test release note - releasing the product for roll out could be prepared. Other remaining documentation is completed.
7. Post Implementation (Process Improvement Phase)
This phase, that looks good on paper, is seldom carried out. In this phase, the testing is evaluated and lessons learnt are documented. Software Metrics (Bug Analysis Metrics) are analyzed statistically and conclusions are drawn. Strategies to prevent similar problems in future projects is identified. Process Improvement Suggestions are implemented. Cleaning up of testing environment and Archival of test cases, records and reports are done

No comments:

Post a Comment