Building Data Storage From SaaS Based Applications

About the Customer

Industry: Industry Specific

The USA based client, allied with product designing company providing exceptional services to its clients by blending technology and design. For more than 20 years they are at the forefront of bringing change in their client’s identity with design and delivering measurable value for their business.

The organization is providing the most consistent and reliable designing services and solutions to its customers across the USA, China, London, Europe, and many other countries.

Key Challenges

As per the client’s business requirement, we need to build a centralized data storage for all SaaS-based & on-premise applications accessible over the internet.

Existing Architecture

Bunch of on-premises applications built with some specific purpose, a few third-party systems like Dynamics GP for Finance & Data warehouse for reporting
Most of the backend data reside in Microsoft SQL Server DBs & flow internally through real-time replications

Challenges with Existing Architecture

Difficult to maintain different data servers & databases
Needs to constantly monitor real-time replications & keep them running
Reporting was quite difficult due to decentralized data storages

Our Solution

To fulfill the client’s business requirement and deal with the challenges there is a need to change the working model and folder structure of centralized data storage.

Existing Architecture

Replace on-premise applications with SaaS-based applications to reduce development & maintenance overhead
Migrate existing on-premise databases to the cloud (Microsoft Azure)
Centralized data from all applications into Azure Data Lake Store
Incremental load of data using Azure Data Factory
Generalized formats for storing all systems data – storing data in text files as a comma-separated resulting best utilization of data storages and saving data spaces
Implement raw level & folder level security using Azure Active Directory (following best security policy & standards)

Implemented BI Architecture

Explanation of Diagram Components

API: On-premises applications have been replaced with SaaS-based applications which provides data in JSON format through Web API calls. This raw data is further processed into meaning form for reporting & analytics purposes.

Azure Data Factory: An ETL tool mainly used for EXTRACT, TRANSFORM & LOAD purpose. We used Azure data factory to execute Web API code (using CustomActivity), copy data from on-premise SQL DBs to Azure Data Lake Store & schedule processing on defined intervals.

Gateway: Gateway or Data Management Gateway is an intermediate bridge between on-premises systems and Azure cloud. We configured this to migrate on-premise SQL data to Azure cloud.

Azure Data Lake Store: This is a centralized storage account to store all application data in a generalized format. All the data stored as comma-separated values for smooth retrieval of data.