The Basics of TDM Testing for Software Storage Applications

The idea behind TDM testing is to ensure a test meets the desired outcomes when it is run under real-life conditions. The testing team works with the user to create test cases and use the best available resources to address the user’s needs. This process is very important for data retention, as users do not want to be provided with invalid data, which could potentially lead to costly mistakes in the future. It can also help ensure that the test suite is as accurate as possible. Here are the four steps to TDM testing:

Define and design a data storage engine. TDM teams will first need to determine the performance of the application, as well as the user’s data requirements. The definition of a data storage engine can vary greatly between organizations. In order to create a functional architecture, TDM developers will need to consider both the hardware and software aspects of the system, along with the user’s data storage needs.

Define test cases. When a TDM team is designing a test case, they will first need to determine the test cases’ purpose. For instance, an enterprise may decide to perform a series of random tests that collect data over the course of an entire day or week. Other organizations may want to test a single product over a specific period of time.

Define test data. The test data will be used to map the application’s database to a consistent and standardized set of standards. In addition, the test data should also include a collection of performance counters. These counters can be used to compare the actual time a data check takes against the expected time, as well as to monitor the health of the storage engine. Once the data has been defined, the TDM testing team can begin to develop and execute the test procedures.

Create test cases. Once test data has been defined, the TDM testing team can begin to create test cases. A typical test case contains one or more collections of statements, which are executed in a specified order. All statements in a test case should be valid ones. A good TDM testing strategy makes use of predicates, which are conditions that entail a certain event to occur, and so they can be used to create test cases that depend on those conditions only.

Map the application’s database to the proper standard storage medium. Many companies make the mistake of mapping a user database to a memory image, which effectively nullifies the application’s capability to accurately represent time within the system. When the application is running, it typically contains application-specific information, and so this information will need to be translated into a data representation that can be stored and used for testing.

Back up the test data. This step is important because even though TDM testing guarantees a data loss prevention, it doesn’t ensure that the application will recover from a data loss. It can be extremely helpful if the TDM testing team is able to capture an accurate copy of all stored test data. Once the data is captured, the next step is to recreate the data in a common environment so that all of the application’s data will be consistent between the two testing environments. Creating test data duplicates the data that was previously lost, but introduces new issues that can be identified and resolved during both TDM and playback sessions.

Make a list of test cases that have been successful. Most organizations make the mistake of checking a piece of TDM testing and then not testing it thoroughly. Checking for a missing interface, missing collection methods, and invalid views is an easy way to identify areas that could greatly improve the functionality of the application’s storage management. When the test data has been verified, TDM testers should be able to generate an exhaustive test plan that will ensure the application’s test results are accurate and provide real-world insight into how well the software’s time and capacity planning are working.