Revolutionizing Software Testing: The Role of Cloud Computing

Abstract

With the advent of IT industry, software has become a primary element of almost all the organizations for different types of problem solving. Field of software is becoming advanced with up gradation of technology. Many software industries across the globe are striving towards the innovative concepts, technology and practices in order to satisfy growing customer needs. So it becomes mandatory to use the best practices and principles to full extent in each and every phase of software development process.

Although the testing phase is one among all the phases which becomes the deciding factor about the quality and reliability of the software.

Software testing has become costly in terms of both time and money.

This large resource requirement problem of software testing can be dealt by combining testing with new technologies. With advances of cloud computing and most of the IT industries are embracing it. Providing testing services in cloud environment has allowed testers to take advantage of vast resources of cloud by paying less amount of money.

The goal of this paper is to describe how cloud computing has evolved and how applications are tested in cloud environment, its necessity, type of testing suitable, and its benefits, challenges, needs and road ahead.

Introduction

Cloud computing has emerged as a new computing paradigm that facilitates the development and utilization of highly flexible, scalable, elastic services on-demand, and over broadband network access. More and more big ventures are using these attributes as a reason to move their businesses to cloud.

Virtualization, which is an empowering innovation of cloud computing, was first utilized for rapidly making virtual computing assets with various working frameworks to test programming applications on different stages.

Software testing on cloud has many pros, most of the organization require testing only after development of their product so these organizations should not bear cost of huge servers, network lines. These computing resources are either not used or underutilized after testing, thus incurring extra cost on budget.

If organizations invest heavily on testing of software, resources are either not used or underutilized after testing is done .Although if it’s known beforehand about the reach of application, if less company can invest in software testing. But for some systems like banking application millions of requests are maintained in less time these requires software testing on cloud. Testing is necessary in such scenario because people rush to their bank accounts regularly on every payday. Reproducing such a scenario would require the provider to set up a test harness (including the user databases) to emulate the actions of millions of users. Similarly, mobile application providers frequently have to deal with maintaining the quality of their services over various combinations of platforms .These computing platforms merge various browser technologies with backend support running on various mobile OS .For services to be reliable providers have to test on all these platforms.

Using cloud computing platforms for software testing can provide increase in both efficiency and effectiveness when compared to traditional testing methods because factors like reducing deployment, maintenance and licensing environment costs, flexibility for acquisition and installation of customized test environments on demand and automated scaling in timely and economical way.

Cloud Testing

The remainder of paper is organized as follows; in Section II, we give background on cloud computing, software engineering, and software testing; in Section III we give literature review of this type of technology which is already being used. Section IV will include benefits, challenges, and needs. And final section will deal with the future of STAAS (Software Testing As A Service) and what improvement can be made with a concluding statement.

Background

Software Testing

Software testing is a phase of SDLC which checks whether the software developed is defect free, and gives the desired output. Testing is of utmost importance to the Software Engineering industry as small software bugs can cost large monetary or human loss.

There are two types of testing methods:

1. Static Testing: Refers to verification in software testing, it’s a process to ensure our product fulfill requirements it was made for.
2. Dynamic Testing: Refers to validation, process to validate that the product so developed is defect free.

Three types of Testing Approaches:

1. White Box Testing: Testing the internal perspective of the system, designing test cases
2. Black Box Testing: Behavioral testing in which testers don’t look into internal perspective but evaluate the functionality of software
3. Grey Box Testing: It is the combination of both types of testing methods.

STLC-Software Testing Life Cycle

Software testing process consists of a series of steps carried out sequentially for best quality product.

Phases of STLC:

1. Requirement Analysis: Involves identification types of tests to be performed and details about testing priorities and where to focus more is extracted, specified area where test is to be carried out. Requirement traceability matrix and Automation feasibility report are generated from this phase.
2. Test Planning: Preparation of test strategy document for various types of testing, selecting of tools and estimating efforts for testing, so this phase basically involves with management of resources and roles. Output of this phase is Test plan /strategy document and effort estimation document.
3. Test case development: This phase involves creation of test cases and automation scripts also create test data. Test cases/scripts and Test data are generated from this phase.
4. Test Environment setup: Understand the required architecture, environment set-up and prepare hardware and software requirement list for the Test Environment. Setup test Environment and test data Perform smoke test on the build Environment ready with test data set up and Smoke Test Results.
5. Test Execution: Involves execution of tests as per plan and document these test results, and prepare a log of defects for failed cases, map defects to test cases in Requirement traceability matrix .Retest the Defect fixes .Track the defects to closure
6. Outputs of this phase are:
7. Completed Requirement traceability matrix with the execution status
8. Test cases updated with results
9. Defect reports
10. Test Cycle closure: Evaluate cycle completion criteria based on Time, Test coverage, Cost, Software, Critical Business Objectives, Quality Prepare test metrics based on the above parameters. Document the learning out of the project.

Outputs:

Test Closure report and Test metrics

Software Testing on Cloud

Software testing when performed on web application which uses cloud environments to simulate user traffic for scaling purposes and for load and stress testing of software is known as cloud testing .Software test like Security testing, performance testing, load testing can be performed easily using cloud infrastructure. With cloud-testing you have unlimited resources at your disposal, paying only for what you consume, only when and if you consume it.

Why use Cloud Computing for Testing?

• Cost Savings

Every enterprise wants to maximize its profits and minimize the cost. By performing testing on cloud they don’t need to buy testing software as they can now be rented for example if purchase of automated test software cost Rs.200000, the test phase only requires 2 months, but if rented it cost 15000rupee/per month, so enterprise has to pay only 30000 instead of 200000.Also the versions of software keep on upgrading so their testing software also are upgraded which incur more cost, which can easily be dealt if testing done on cloud.

• More Realistic Performance Testing

By testing on cloud platforms we can open many users and create a virtual scenario to test the software in worst case .So that the software developed work perfectly in peak conditions.

• Improve testing Efficiency

Time for building test environment is reduced significantly there is no need of machine and network preparation and operating system installation .So we are now able to test our software faster and with paying much less.

• Easy Simulation

These cloud testing environments can be easily documented, be easily understood as there is no need to maintain our own hardware, only thing needed is internet access.

Cloud testing services can be provided in following two ways:

Using Web Access to Cloud Testing Services：

Access to the page provided by the cloud testing services, the enterprise does the functional testing and performance testing. It is easy to achieve for the WEB-based application software products of B/S structure.

Using Virtual Machine Technology:

Cloud testing service providers provide the IP of the virtual machine; clients connect the virtual machine through remote desktop and other ways. But the way of non-WEB application software, still needs to be installed on the virtual machine and then it can be tested.

IaaS platform architecture — Virtualization technology

The main process of cloud testing as follows:

1. Users login cloud testing provider's website and register user information;
2. Users apply for the test platform resources, applications need to describe the configuration requirements of virtual machine environment, Such as operating system version, browser version, memory size, hard disk size, hard disk speed, network bandwidth, firewall, etc.;
3. The service provider of cloud testing review the application and configure the test platform.
4. Under the rental agreement, the user (online) pre-paid service fee.
5. Users login the cloud testing platform, until the end of the testing.
6. According to the flow or time to pay the actual costs, the lease ends.

Cloud Testing

Cloud testing is a type of software testing used by organizations. With the help of cloud's infrastructure the organizations can perform software test like Security testing, performance testing, Load testing etc.

Objectives of Cloud Testing:

The objectives of Cloud testing are as follows:

• To test cloud's inter-operation capability and computability.
• To assure the quality of cloud based applications.
• To check whether the application is accessed by authorized users or not.

Types of Cloud Testing

1. Performance Testing: Performance testing is used to find out the threshold, bottlenecks and limitations of the cloud based application
2. Load Testing: Load testing is a type of performance testing where the application is compared under normal load conditions and peak load conditions.
3. Compatibility testing: It will test the capacity of the application to deal with various working frameworks like Windows 10.
4. Browser Performance testing: It is like similarity testing yet by utilizing this sort of testing we can check whether our application will deal with various kinds of internet browsers like Google Chrome, Microsoft Edge and so forth.
5. Stress Testing: It is a sort of act testing where we will test the capacity of the cloud based application to keep up its adequacy even past limit.
6. Security Testing: It is a sort of testing where we check whether the information in the cloud had been gotten to or altered by an (information honesty) any unapproved client or program.

Types of Cloud Testing

Steps in Cloud Testing

In general whenever testing is to be performed the first thing that comes to our mind is the test cases.

1. In the first step we will be developing the user scenarios. A scenario in general has a goal and the goal is functional related.
2. In the second step we will have to design test cases. In Third step we will select one of the cloud service vendors who will provide the tools for cloud testing.
3. Then the next few steps will deal about how to create web traffic for testing with the help of cloud service provider
4. The next step is to start the cloud test with help of test cases.
5. After the test is complete we check whether the goals for the particular testing are met or not.
6. At the last step we will deliver the result back to the organization who had requested the testing.

Testing As A Service

As mentioned earlier there are three basic models of cloud computing namely SaaS, PaaS and IaaS. But whenever we want to perform cloud testing there is an another model known as TaaS (Testing as a Service). TaaS is used by companies like Oracle, Wipro etc for cloud testing purpose.

Workflow of TaaS: The main goal of TaaS is to reduce the budget of IT companies. It is due to the fact that the companies can concentrate on their own business and can outsource the software testing to third parties.

Workflow of TAAS

The workflow of TaaS has the following service capabilities:

1. TaaS process management: It offers test project management and process control.
2. QoS requirement management: It is mainly about QoS requirements and modeling of the software testing.
3. Test environment service: It provides an on-demand test environment services to establish virtual computing resources.
4. Testing solution service: It offers test modelling and test methods which are together known as testing solutions.
5. Testing simulation service: It establishes test simulation environments.
6. On-demand test service: It provides on demand test execution services based on selected schedules and rest wares
7. Tracking and monitor service: It allows test engineers to track and monitor diverse program behaviours.

Benefits, Challenges and Needs

As TaaS is a new area of computing, a significant amount of research effort is expended on exploring the potential benefits and studying the feasibility of migration. Benefits, challenges, and needs are described in the following sections.

Benefits

One of the real advantages of TaaS is small expenses. Representative compensations and advantages, resources overhauls and support, programming licenses and apparatuses, and physical structure space and utilities are reliably expanding the expense of software testing. TaaS can help decrease these expenses by offering a pay per-use testing model that handles the situations and software required for testing and just charging at the time testing is really occurring. Moreover, the utilization of testing administrations can enormously lessen work costs by limiting the number of testers required for a given application and the measure of time spent on running tests. Distributed computing offers clients’ access to immense and flexible assets, giving servers and virtualized frameworks at a value point that can't be accomplished with conventional computing test labs.

Past the monetary advantages, these assets give a productive and less tedious technique for testing, enabling larger suites of tests to be kept running in more conditions and designs at the same time. Environments, applications, and different tools are provisioned by the TaaS supplier, enabling customers to get to the product they require without paying the forthright costs for licenses. The capacity to begin and quit testing at whatever point important, pay just for the time spent testing, and abstain from authorizing charges all make an unmistakably progressively adaptable condition.

Traditional testing has many cons. With only so much time thus numerous PCs accessible for testing, potential client conditions are missed. The capacity to test in more conditions offers progressively reasonable test outcomes. Applications under test have a superior indicator for in general quality and unwavering quality.

Another advantage of TaaS is consumability. A significant part of the work in TaaS is done off camera. The rest of the means required by the purchaser ought to be of negligible exertion. In the Regression TaaS contextual investigation, the clients just required a couple of hours to ace the fundamental strides of the apparatus.

Concerning security, numerous administrations enable clients to submit parallels or executables for testing. This alternative places less classified or restrictive data in the cloud condition, securing intellectual property.

Challenges

Despite the many benefits, new technologies are not without issues and challenges. Several of the provider challenges have the potential to translate into consumer issues. One example is the demand for constant global availability. TaaS providers must have their services and technical support available at all time, especially when working in international markets. While a difficult problem to solve on the provider side, the inability to test or the loss of test data due to an outage can cripple a consumer. Furthermore, providers must develop and maintain testing services and technical and support skills that match or exceed consumer needs.

Security concerns also exist. Especially when development code is examined rather than binary packages, testing software on the cloud creates a potential risk of intellectual property being compromised. The usage of binaries mitigates this risk, but does not eliminate it. Additionally, testing may be dependent on specific test data that necessitates actual client information. Nondisclosure agreements may prohibit the transfer of this information, limiting or preventing testing by a third party.

Several other challenges may hinder the creation of a consumable TaaS service. TaaS systems must be capable of integrating properly with the applications under test and no interoperability issues can exist. They must also be highly scalable for use with large systems. These and other non-functional requirements must be satisfied to motivate potential consumers.

Another concern exists when creating services that can test applications for end users, such as TaaS. While most end users have benevolent intentions, others could use these tools to location and exploit weaknesses in a software product. The potential of malicious usage, though, may also result in a push for better software quality.

Needs

In addition to resolutions for the aforementioned challenges and issues, various needs must be satisfied for TaaS to become an attractive alternative to traditional testing. Reasonable pricing models must be developed to entice consumers. Services need to be offered with transparent pricing, service descriptors, and metrics that detail their value to the consumer.

Standards should also generate. Consumers and providers require well-defined testing and quality of service standards for both the technology at hand and to assist in the establishment of fair Service Level Agreements (SLAs). Standards must not be overly restrictive, though, as too many standards could potentially create more predictable tests.

Finally, in order to increase the demand for TaaS and to maintain its relevance over time, innovative test methods and solutions and dynamic test platforms and tools must be developed.

Future Works

For academics, TaaS provides much room for future research. Some suggestions follow:

• Exploring types of applications best suited for TaaS.
• Determining the effects of TaaS on a customer’s business.
• Developing pricing models and service descriptions for TaaS.
• Identifying how to handle test data, especially when related to confidential customer information.
• Defining quality of service standards for TaaS.
• Evaluating which test models and metrics are important for new types of test services in the cloud

Conclusion

Software testing as a service is a relatively new aspect of software testing and cloud computing that offers many advantages over traditional methods of testing. As cloud computing continues to increase in popularity and applications grow even larger, the need for new and innovative testing techniques will push testing into the cloud. TaaS promises significant savings in cost, effort, and time by offering services that handle the testing needs of an organization or group. Testing services are easy to consume and highly scalable, allowing any individual with access to the internet to test software on their own schedule, paying only for the time spent in test.

These benefits are not offered without challenges. TaaS systems must maintain constant global availability and offer services and skills that match the client needs. Test methods and solutions, as well as test platforms and tools, must constantly evolve, providing innovative services for consumers while offering reliable results. Security risks must be mitigated and integration and interoperability challenges must be overcome. Quality of service standards have yet to be written and reasonable prices must be set.

The goal of this paper was to inform readers on the current state of academic research within testing as a service; however, this state is constantly moving forward with new advancements. In order to make TaaS a truly viable method of testing, the issues described need to be addressed. Once that is done, the vast and elastic resources of cloud computing can be fully harnessed to truly revolutionize testing.

Acknowledgment

I would like to thank Dr. MadhuriBhavsar, for guiding and encouraging me to make term paper in this domain I am grateful for being given the opportunity for writing a term paper based on cloud computing under the university syllabus as it helped me learn many new things and expand my knowledge. . I would also like to thank my family and friends for continuous motivation and support.

References

1. Wang Jun1,2 Fanpeng Meng2, Software Testing Based on Cloud Computing,IEEE
2. A Survey of Software Testing in the CloudKoray˙Inc¸ki∗†, ˙Ismail Arı †, HasanSozer¨ †
3. A Framework for Automated Software Testing on the CloudGustavo Sávio deOliveira, AlexandreDuarte
4. Software Testing and its Scope in “CLOUD: A Detailed Survey” NarasimhaMurthy Ms, Suma V
5. Software Testing as a Service An Academic Research Perspective Brianna Floss, Scott Tilley
6. A Survey of Testing as a Service in Cloud computing, Mr.PillutlaHarikrishna, Dr. A. Am uthan
7. STLC :https://www.guru99.com/software-testing-life-cycle.html