Panoply Blog - Data Management, Data Infrastructure, and Data Analysis.

Data Warehouse or Data Lake: How to Select the Right Solution

Written by An Bui | Nov 21, 2017 7:38:29 PM

Data warehouses and data lakes are two different types of data storage repositories. Data warehouses integrate data from various sources and structures them for business reporting. Data lakes store raw structured and unstructured data in whatever form the data source provides. This blog will attempt to define each of these strategies, identify fundamental differences, and provide insight as to which approach is best for big data integration.

What is a data lake?

A data lake is a storage repository that holds massive amounts of structured and unstructured data in whatever form the data source provides until needed. Data lakes do not require users to have knowledge of the analyses they want to perform.

What is a data warehouse?

Data warehouses store summarized historical data from many different applications. There is a one to many relationship between data warehouses and the applications that serve as data sources. Examples of data sources include but are not limited to customer relationship management (CRM), enterprise resource management (ERP), or even social media data.

You Deserve it All—See How Easy the Hybrid Approach Is


Key differences

Data types

Data warehouses stores data from specific traditional sources including credit card transactions, customer relationship management (CRM), systems and enterprise resource planning (ERP), systems. Other data sources such as social media activity or web server logs are ignored because storing this type of data for use can quickly become expensive and difficult to manage. Data is organized, defined, and metadata applied before the data is written and stored. This process is officially known as schema on write.

Data lakes take the Life Cereal Mikey approach – They consume EVERYTHING including non-traditional data types such as social media activity, images, and web server logs. Data is stored in its raw form, and the requested information is applied to the schema as it is pulled out of its stored location, rather than as written to storage. This process is officially known as a schema to read.

Data retention

Data retention in a data warehouse requires substantial effort and time to analyze targeted data sources as well as the determining if the data these sources contain will provide useful insights once they are stored in the data warehouse. If the data does not answer a specific business question, it is not included in the data warehouse. Following this data transfer model helps to reduce disk storage space and improve data warehouse performance./p>

Data retention in a data lake is a less complicated process because data lakes retain all data – raw, structured, and unstructured. Data is never purged which always analysis of past, current, and future events at any time. Because there are no data type restrictions, data lakes can be easily created, scaled to petabytes, and maintained using retail servers and inexpensive disk storage devices.


Data Warehouses store historical data and apply structure to the data when it comes in. This approach is great if you only want to use the data to answer specific business questions. However, if your questions change or your organization wants access to more data to accommodate more in-depth analysis, the limitation of a data warehouse quickly surface. While a good data warehouse design will have can adapt, the loading the data and the development effort needed to make analysis and reporting less challenging is a time consuming and productivity stealing task.

Because data lakes stores all data in its original format, it is immediately accessible to users who can explore the data any way they want. If the information they retrieve is useful and they want to use it again, they can apply a formalized schema to the data, store it, and share it with others in the organization. If the information retrieved is not useful, they can discard it without affecting the stored data or using precious and expensive development resources.

Which strategy is best for your data?

Determining the best approach for your organization can be challenging. As your organization grows and you find it more challenging to access data from across the organization, consider the hybrid approach— create a data lake to run in tandem with a data warehouse. Fortunately, Panoply offers the best of both worlds, through machine learning and natural language processing (NLP) to learn, model, and automate standard data management activities. There’s no reason data warehouse or data lake is an either-or, when you can easily get both.

You Deserve it All—See How Easy the Hybrid Approach Is