Case Study

Solution

From an architectural point of view, the main challenge was to create a structure capable of handling huge daily amounts of data, while maintaining its integrity and resilience, since there is a very small hourly window to recover this data in the event of a disaster.

The solution involved the use of Microsoft Azure Data Factory, programmed to ingest raw data and subsequently transform it in two ways:

Sending part of the data to Microsoft’s main relational database, Azure SQL Server;
Sending the remaining data to the non-relational database Azure Cosmos DB.

However, that was only part of the challenge. In order to meet expectations and control such a volume of data, the application would need to achieve a high speed when made available to the client. This speed was achieved by using techniques such as in-memory cache.

The cache policy takes into consideration that after the user accesses the platform for the first time, most of its data becomes available in memory and all subsequent accesses do not require the database, making the response highly efficient.

From a security point of view, Imovirtual has strict access and validation rules. In order to deliver this project in production, penetration tests were performed and the solution was adapted to the security policies, by using security vaults to store secrets, passwords and connections to the database.