Data Platform Terms Study Plan for DBA / DB Platform Engineers

# Data Platform Terms Study Plan for DBA / DB Platform Engineers

AI Summary

Purpose:

• Source plan for the Korean human-facing study series human/study/content/data-platform.
• Organizes core data platform terminology for a DBA or DB platform engineer.

Key points:

• The series explains terminology through operating questions: source of truth, transformation point, consumers, recovery boundary, quality gates, and ownership.
• It covers OLTP/OLAP, Data Lake, Lakehouse, Medallion Architecture, Data Warehouse, Data Mart, Semantic Layer, BI, ETL, ELT, CDC, Orchestration, Governance, Catalog, Lineage, Data Quality, and platform operations.
• The Korean study chapters are presentation outputs grounded in this source plan and public vendor/project documentation.

Relevant when:

• Creating or updating DBA/data-platform study materials.
• Explaining data platform concepts in Korean for practical study sessions.
• Extending the study portal with additional data engineering chapters.

Do not read full document unless:

• You need the exact curriculum sequence or reference list for the data-platform study series.

Linked documents:

• human/study/content/data-platform/series.json
• ai/study/curriculum.md

Open Questions

• Whether to add hands-on labs using local DuckDB/PostgreSQL/MinIO examples.
• Whether to split Lakehouse table formats into a separate advanced series.

Details

Intended audience

• DBA transitioning toward data platform engineering.
• DB platform engineer who already understands relational databases but needs shared terminology for analytics platforms.
• Data engineer who needs to explain platform concepts to operations-oriented engineers.

Curriculum

1. Big picture of data platforms

- Source, ingestion, storage, transformation, serving. - Operating questions: source of truth, transformation boundary, consumers, retention, recovery.

1. OLTP and OLAP

- Why operational databases and analytical systems are separated. - How replication, snapshots, CDC, and events move data out of OLTP systems.

1. Data Lake, Lakehouse, and Medallion Architecture

- Data Lake as raw scalable storage. - Lakehouse as lake storage with warehouse-like reliability and governance. - Bronze/Silver/Gold as progressive quality layers.

1. Data Warehouse and modeling basics

- Fact, dimension, grain, star schema, snowflake schema, SCD, partitioning, aggregates.

1. Data Mart, Semantic Layer, and BI

- Use-case-specific serving datasets, metric definitions, dashboards, refresh, and ownership.

1. ETL, ELT, CDC, and Orchestration

- Movement and transformation patterns, change logs, idempotency, checkpoints, batch/streaming/micro-batch.

1. Governance, Catalog, Lineage, and Data Quality

- Rules, metadata discovery, impact analysis, quality gates, quarantine, freshness.

1. DBA / DB platform operations checklist

- Operational questions for incidents, schema change, CDC lag, mart refresh, quality failures, and cost/performance control.

Reference anchors

• Databricks medallion architecture and lakehouse glossary.
• Microsoft Fabric OneLake medallion lakehouse architecture.
• IBM Think pages for data warehouse, data mart, data governance, catalog, lineage, ETL, and ELT.
• Apache Airflow core concepts for orchestration.
• Debezium documentation for CDC.
• Great Expectations and OpenLineage for quality and lineage examples.

Editorial guidance

• Korean human-facing chapters should avoid abstract platform marketing language.
• Explain each term with operational consequences: failure boundary, reprocessing, schema change, freshness, quality, cost, ownership.
• Use short tables and diagrams because the study portal renders them well.
• Each session should end with questions or exercises the user can apply to current work.