Basic Information
| Field | Details |
|---|---|
| English name | Data Analytics Core Concepts |
| Full name | Data Lake / ETL / ELT / Zero-ETL / Data Pipeline |
| Chinese description | 数据分析核心概念 |
| Japanese description | データ分析の基本概念 |
| Exam frequency | ⭐⭐⭐⭐⭐ |
| Often confused with | Data Lake / Data Warehouse / ETL / ELT / Zero-ETL |
In one sentence
A data pipeline continuously turns raw data into assets that can be queried, visualized, and used for model training.
Key points
- A data lake commonly uses S3 to store structured, semi-structured, and unstructured data at low cost.
- ETL transforms before loading; ELT loads first and transforms with the target system.
- Zero-ETL reduces traditional pipeline management; it does not mean transformation disappears.
- A typical flow is ingest → store → catalog → process → query → visualize or ML.
Exam focus
- Data lake points to S3; data warehouse points to Redshift.
- Glue Catalog stores metadata, not the primary business data.
Common pitfalls
- Do not choose by product name alone; confirm data type, latency, control, operations, and cost constraints.
Remember
Ingest, store, catalog, process, query, then visualize or train.