Please enable JavaScript.
Coggle requires JavaScript to display documents.
Databricks Platform Administrator Fundamentals - Coggle Diagram
Databricks Platform Administrator Fundamentals
Architecture
Control Plane
Lives in Databricks own cloud
Managed Backend Services
Repos Notebooks
Encrypted At Rest
Job Scheduling
Web Application
Cluster Management
Data Plane
Where data is processed
All compute resources in the data plane reside in your own cloud account.
Host Clusters and Sql Warehouses
Contains Data Sources that back the workspaces, DBFS, Meta stores and external tables, and conections to external data sources
Major Services
Unity Catalog
Centrilized Hub for managing and tracking data objects
All request from the various Databrick services are authenticated and logged by Unity Catalog
Databricks Machine Learning
Integrated end-to-end machine learning environment
Incorporationg managed services for experiment tracking, model training, feature development and management, and feature and model serving
Databricks SQL
Simple expirience for SQL users.
run quick and adhoc queries on their datalake
Create multiple visualization
Build and share dashboards
Two Flavors:
Classic - Furnish computational resources out of the data plane, needs more administrative effort and incurs soma latency when starting the resources for the first time.
Serverles are provided by Databricks-manged resources. Admon and latencies are virtually non-existent within this model.
Data Science and Engineering Workspace
Classic Data bricks environment for colaborationg data scientist, data engineers, and data analyst
Forms the backbone of hte Databricks Machine Learning
Workspace organizes objects like notebooks, interactive web documents containing code and documentation into folders
Provide access to data and computational resources such clusters and jobs
Security Model