The differences between Data Lake and Data Warehouse

Data Lake and Data Warehouse are types of data storage useful for the infrastructure of dynamic and competitive organizations. Two possibilities with pros and cons to be evaluated according to the needs of the companies.

The expansion of the Big Data world has resulted in various forms of storage, such as the Data Lake and the Data Warehouse. With similar and sometimes confusing functions, both host corporate data for business analysis and reporting, but with different generation systems and access patterns.

Data Lake is a data warehouse – structured, unstructured and hybrid – in one place, with limited quality, but provides the basis for reporting, visualization and advanced analysis. Your data does not relate to predefined objectives (schema-on-read), which means that it can be stored without cleaning, treatment or organization, that is, in its raw state. By storing data at a low cost and in a scalable way and by collecting, importing and processing data from already used analytical infrastructures, you can go through successive upgrades, as the data grows, without having to be out of date in the short term. There are no restrictions for the tool, hence its name, a “lake”, which houses information in a single location: Big Data. Its greatest advantages are: to house any type of data; have flexibility; democratize access; store large amounts of data and algorithms.

Data Warehouse is a central warehouse of integrated and structured data, from two or more sources, mainly used for reporting and analysis and considered as the main component of business intelligence (business intelligence). It implements predefined and distributed analytical standards for a large number of users in the enterprise. Its characteristic is to have a “schema”, with clean, treated and organized data that works like a stock to be consulted periodically, with well located and easily accessed information, about customers and suppliers, but which tend to be out of date in the short term. Storing volumes of data in a Data Warehouse is complex and costly because, before collection, it is necessary to prepare, transform and structure them. Its biggest benefits are: integrating different sources in a single view, storing sanitized data, favoring insights and allowing historical analysis.

According to Gartner, Inc., a certain “fad” in relation to Data Lake is creating some confusion in the area of management and it is necessary to understand how it operates and how to obtain value from it. The fact that the data is in the “lake”, accessible to everyone in the organization does not imply that everyone is qualified for its manipulation and analysis. It takes business management control. Many see Data Lake and Data Warehouse as interchangeable options, but in reality, each has a different primary purpose. When combined, they support complex, diverse and distributed workloads. It is not a matter of knowing which is the best, but which is the best option for a particular company.. And the choice depends on variables such as the size of the company, the limitations and the objectives of the Big Data projects. Despite the technical, conceptual and purpose differences, the tools are complementary and, when they work in an integrated manner, they generate a good cost-benefit ratio and provide organizations with process and time optimization.

https://www.talend.com/resources/data-lake-vs-data-warehouse/

https://planin.com/cuidado-com-a-ilusao-do-data-lake-aconselha-o-gartner/

http://datascienceacademy.com.br/blog/como-diferenciar-data-hub-data-lake-e-data-warehouse/

http://www.gartner.com/document/2805917.