DataTerminology.comNEW
Centralized, scalable storage that holds raw structured and unstructured data in native formats for flexible analytics.
DESCRIPTION
A data lake stores vast volumes of raw data at low cost, typically on cloud object storage. It supports schema-on-read, enabling varied analytics and ML workloads without upfront modeling. Lakes ingest batch and streaming data and often integrate with table formats like Delta Lake, Iceberg, or Hudi for ACID and governance. Catalogs, lineage, and quality controls mitigate the risk of a ‘data swamp.’ Lakes frequently serve as the foundational layer for lakehouse architectures.
ALTERNATIVE DEFINITIONS
Wikipedia
A data lake is a system or repository of data stored in its natural/raw format, usually object blobs or files.
Centers on natural/raw formats stored as objects or files.
AWS
A data lake is a centralized repository that allows you to store all your structured and unstructured data at any scale.
Emphasizes centralization, any scale, and mixed data types.
RELATED TERMS
Object Storage Extract-Transform-Load ELT Schema-less Delta Lake Apache Hudi Apache Iceberg Data Discovery Platform
COMPETING TERMS
RELATIONS
KNOWLEDGE AREAS
Data Architecture Data Engineering Cloud Data Warehousing & BI Data Storage & Operations
TARGET AUDIENCE
Data Engineer Data Architect Cloud Architect BI Specialist Data Scientist Chief Data Officer
EXTERNAL SOURCES
- Wikipedia - Data lake
- AWS - What is a data lake overview
- Google Cloud - Architecture for building a data lake
- Microsoft Learn - Azure Data Lake Storage Gen2 intro
TOP QUESTIONS
What makes a Data Lake different from a Data Warehouse?
Lakes store raw data with schema-on-read for flexibility; warehouses store modeled, structured data optimized for analytics performance.
Category: Distinction
What risks can turn a Data Lake into a data swamp?
Lack of metadata, governance, and quality controls leads to unusable, untrusted data.
Category: Risk
How do open table formats improve Data Lakes?
Delta Lake, Iceberg, and Hudi add ACID, schema evolution, and time travel, enabling reliable analytics.
Category: Capability
What workloads run on a Data Lake?
Batch ETL/ELT, streaming ingestion, machine learning feature generation, and Ad-hoc analytics.
Category: Example
How is cost managed in a Data Lake?
Use lifecycle policies, partitioning, compaction, and query optimization to reduce storage and compute spend.
Category: BestPractice
How to secure a Data Lake?
Apply IAM, encryption, network controls, and fine-grained access with audit logging and policy enforcement.
Category: Security
Is schema-on-read always best for Data Lakes?
Not always; many teams mix schema-on-read with enforcement at write time via table formats and validations.
Category: Opinion
What is the role of a catalog in a Data Lake?
Catalogs provide discovery, schema, lineage, and access control—key to preventing swamps.
Category: Concept
Can a Data Lake support BI directly?
With table formats, query engines, and semantic layers, many BI workloads run directly on lakes.
Category: Practice