What is the biggest pitfall to avoid when executing cloud database testing? That is, when the database is hosted within the cloud, what sort of challenges should software testers expect?
Make sure the testing environment has an exact replica of the production database. By "exact," I mean exact. Not close, not similar, not "basically the same." It has to be exact. Otherwise, you're not testing your application as it will perform for your users, including not testing the performance, accessibility or availability of your production database.
If there is a cloud database for the application, you can also host a test database in the cloud that mimics production. It's the only way to ensure your team is testing the application as it'll behave for a user. You will need to use test data rather than live production data because of privacy and confidentiality regulations. Ask your database testing team to provide masked test data to replace production data so you do not break any privacy or confidentiality rules applied to your type of software.
If there is a cloud database for the application, you can also host a test database in the cloud that mimics production.
By "masked," I am referring to data masking. Data masking provides protection from breaching sensitive information when copying production data to test environments. Masking removes the original data with fictitious data so it can be used for testing. Unlike data encryption -- or a virtual private database where data is hidden but can be retrieved with access -- data masking removes the ability to access the original data at all.
Taking the time to verify a data breach is an extremely expensive and reputation-damaging mistake for a company of any size. Use the rule of thumb: Trust but verify. Your test team is part of the larger development team. Team members need to trust each other, but it is more critical to the company that confidential data is not breached, so always verify that the data has been masked.
Data masking allows you to neutralize sensitive data records and still have usable data for cloud database testing -- data that resembles the original in appearance and integrity. This ensures applications are fully tested while removing the risk of unauthorized access to the original data. For more information on data masking for Oracle systems, go to section 15 of its user guide.
This was first published in June 2013