We found two that looked interesting, we’ll discuss them here. We started by looking at the systems that were available to anonymize databases. This is where the anonymized version comes in. Sometimes however, this empty database isn’t enough we need some actual data to test certain features or bugfixes. For development and testing purposes, we usually use the stripped database. The best way to avoid data leakage is to avoid using the data in the first place, which is why we decided to only ever use the production data when it is absolutely necessary. When working with customer data, we feel it is our duty to handle these as carefully as possible to avoid data leakage (not to mention GDPR).
MASKERADE NL FULL
We store three versions of the database there the full backup including all customer and order data, a stripped version without any customer or order data and recently, also an anonymized version where all personally identifiable information has been anonymized, that is replaced with fake data. When we set up a review environment for our clients, we download the database from our backup location on Amazon S3. We now have tackled this issue and in this blogpost I’ll explain the why and the how. Anonymize database so we have near-production data available.I ended that blog with a few bullet points for future implementation, of which one was: That blogpost details how we leverage Gitlab CI with Docker & Kubernetes to create disposable Magento testing environments. A bit less than a year ago, we wrote about our deployment pipeline with Kubernetes.
0 kommentar(er)
