Challenges of Cloud Testing Cloud testing techniques are still in its initial stage, there are several challenges associated as below (Jain & Malhotra, 2013; Gopalakrishnan , Prakash & Ramadoss, May 2012): Cloud Service Model : With a variety of service models and having customers from multiple domains, it becomes very challenging for the testing team to generate an optimal test matrix. For each service model, the subscriber has access to different sets of system components and therefore has different use cases for testing availability, security and performance. For example, in case of SaaS, CSP is responsible for the performance experienced by the end user while in PaaS, CSP is responsible only for the performance of the environment and infrastructure. Furthermore, different business areas present different risks and therefore different acceptance matrices, for example for a medical company data security is a major concern, for a travel agency service availability is a major concern. Security Challenges: Information security takes on a whole new dimension when it moves out of the reach of the customer's internal location and becomes part of the external domain space shared with multiple customers. Since the cloud supports multi-tenancy, the testing team must perform a comprehensive set of security tests to ensure that data is protected not only from Internet attacks but also from other customers sharing the same cloud. Data security must address aspects such as data access, storage, retention and integrity policies. This can be done by running data access and multi privilege tests, the challenge is how to simulate a test environment where any change in the load of other clients can change the result. Running the scenario with multiple tenants is helpful to ensure there are no loopholes. Along with multi-tenancy testing, other tests also require… half the paper… It is a challenge for a tester to determine the impact of an updated interface on other interfaces and embedded systems. This test can be handled by performing backward compatibility testing of the interface. Again, automated testing of regression cases goes a long way in successfully testing changes. Frequently releasing feature-rich applications: The SaaS vendor frequently releases feature-rich applications to compete with market requirements. To ensure that new features are released, the application communicates correctly with other interfaces, and that all features work properly, the tester must perform end-to-end testing. Running end-to-end tests is a tedious and challenging task. This challenge can be overcome by creating an automated test library to test SaaS application updates. Tests should measure the impact on functional and non-functional aspects of the application.