Sunday, November 23, 2008

What to analyze in test result and how to make recommendations?

It depends on what sort of testing we have conducted and what types of report we have for particular test result, and what types of statistical diagram resulting there of, and more.

Such as, the scalability index could show the average transaction per second (TPS) throughput. Transaction distribution chart could be useful to audit a test to validate the result by identifying usueally long or short response rates or trends.

So, in general load testing scenario, we have to analyze transaction per second (tps) results as per test operates; analyze the range of response times for the duration of each test use case; analyze the timing values for each transaction on each test node in a set of XML formatted files; analyze the resource utilization on CPU, network, and memory utilization for each test node for each test use case; analyze the results of multiple test scenarios.

How to design test/ test case/ test plan?

Before designing any test, test cases, or test plan; first of all we have to think about the core point - how could we maximize the quality of the software? Depending on the limitation of human resources, budget, time, risk, market, and the future requirement we have to be revolved around the 'quality'and follow the centralized theme of the 'quality'.

The testing strategy to be undertaken might be driven by different factors, basically resource/time/and cost oriented factors. Looking at the nature of the development and the requirement of the project, the testing strategy could vary from flexible strategy to strict one. We need to analyze the risk and need to know how to handle overall test management, error management, and how to execute the overall test cases.

Based on the size of the project, we have to choose the strategy to be taken by either going for manual testing or automated testing. We should not go for automated one unless it is really required. But, we can not avoid manual testing. If we choose automated testing, we need to think about its environment set up, technical resources requirement, updates requirements of the tool, and cost incurred from these activities. It all depends on how we formulate the strategy.

Thursday, November 20, 2008

Software Development Life Cycle

Software Development Life Cycle (SDLC) - The process of the development of the application or software product is called the software development life cycle.

Phases of SDLC -
  1. Project Initiation (Clients identify their business requirements and provide it to the Software Company)
  2. Requirement Analysis and Gathering (Business Analysts fetch the requirement from clients to form the requirement document and Use Case document)
  3. System Design (The application design and layout is developed and documented by design team)
  4. System Development (The application will be coded and developed by the team of developers)/(The developers also perform unit tests after completing the development)
  5. System Testing (The application will be tested for functionality, navigation, security, and performance by the test team)
  6. User Acceptance Testing (The application will be tested by business users to make sure system is acceptable and meets business requirements)
  7. System Implementation/ Production (The application will be deployed in the production environment)
  8. System Maintenance (Bug fixes, Change Requests or New Requirements)
(Reference: http://www.learningdom.com)