DDrona4U
Sign inCreate account
My learningData Bricks Interview QuestionsLesson 04
How Databricks Separates Storage from Compute — The Core Architecture Behind Its Scalability

Lesson 04

How Databricks Separates Storage from Compute — The Core Architecture Behind Its Scalability

Databricks separates storage and compute by keeping data in cloud object storage while using independent, scalable compute clusters to process it. This architecture allows organizations to scale compute up or down without moving data, reduce costs, support multiple workloads simultaneously, and enable the flexibility that powers modern data engineering, analytics, and AI platforms.

Get the full lesson

Sign in to unlock everything beyond the preview — it's free.

  • Take timestamped notes as you watch
  • Read the full transcript and download resources
  • Join the discussion and track your progress
Sign inCreate free account

Curriculum

20 lessons · 45m

0/20 lessons done45m left
  1. 0102:00

    Lakehouse vs Data Lake + Data Warehouse: What’s the Real Difference?

    02:00

  2. 0201:59

    What Happens Between the Control Plane and Data Plane When You Run a Databricks Notebook?

    01:59

  3. 0302:01

    Why Databricks Runs in Your Cloud Account — And Why It Matters for Security

    02:01

  4. 02:02

    How Databricks Separates Storage from Compute — The Core Architecture Behind Its Scalability

    02:02

  5. 0502:29

    What Is a Databricks Workspace and How Does It Connect Clusters, Jobs, Catalogs, and Users?

    02:29

  6. 0602:52

    Databricks Workspace: The Central Hub for Users, Clusters, Jobs & Catalogs

    02:52

  7. 0702:13

    Databricks Runtime vs Open-Source Spark: What Do You Lose When You Leave?

    02:13

  8. 0802:21

    Databricks Serverless Compute: What Happens When There’s No Cluster to Manage?

    02:21

  9. 0902:11

    Delta Transaction Log: How ACID Works on Top of Parquet Files

    02:11

  10. 1002:01

    Delta Lake Time Travel: Recovering a Table After an Accidental Truncate

    02:01

  11. 1102:12

    OPTIMIZE vs VACUUM vs ZORDER in Delta Lake: When Should You Use Each in Production?

    02:12

  12. 1202:20

    Liquid Clustering vs Z-Order vs Partitioning: Best Strategy for a 50 TB Multi-Filter Table

    02:20

  13. 1302:15

    Deletion Vectors vs Copy-on-Write: How Delta Lake Speeds Up MERGE and DELETE Operations

    02:15

  14. 1402:32

    Delta UniForm: Querying Delta Tables from Iceberg and Hudi Without Data Rewrites

    02:32

  15. 1502:22

    Delta Lake Schema Evolution: What’s Automatic, What Needs mergeSchema, and What Can Break Your Pipeline?

    02:22

  16. 1602:11

    Concurrent Writes in Delta Lake: What Happens When Two Streaming Jobs Hit the Same Table?

    02:11

  17. 1702:08

    Managed vs External Tables in Databricks: Which One Should You Choose in 2026?

    02:08

  18. 1802:09

    VACUUM with 0-Hour Retention: Why Delta Lake Time Travel Suddenly Broke

    02:09

  19. 1902:13

    Zero-Downtime Migration: Converting a 10 TB Parquet Table to Delta Lake Safely

    02:13

  20. 2002:16

    Small Files in Delta Lake: Every Lever You Can Use to Fix Read Performance

    02:16