The problem is that your application may not be designed for nondisruptive updates, especially if it was not originally developed as a SaaS application. There are key aspects to consider a updating the application code b changing the data structure in the storage Special strategies can be used to roll out applications, in which a new release is deployed in a new software environment stack , then tested and launched if the deployment was successful. Old stack resources can be decommissioned and reused later.
One approach to seamless updating is to make the data model n compatible . This means that if a deployed release has data model version n, it is compatible with the previous Whatsapp Mobile Number List model version n , thus ensuring that an update does not break it. Where are the guarantees? Throughout the development cycle, a lot of discipline is required and following certain guidelines, for example, you can not delete columns, providingnecessary upgrade scripts to handle any data migration needs, etc.
In case the update does not happen, it is possible to roll it back. It is now clear that not only is this technically difficult to perform due to data migration and rollback, but it can also slow down the deployment process. You should carefully consider what best suits your needs and implement the appropriate solution. As you build your application and design your deployment, you may want to consider several deployment approaches.