Explore the versatility of FoundationDB as a database construction kit in this 25-minute conference talk. Learn how to build custom database solutions from scratch using an empty key-value store while maintaining safety and reliability. Discover the process of creating a database with a schema, implementing non-blocking schema changes, constructing secondary indexes, and developing a change data capture system for auditing and ETL. Delve into enhancing object storage systems like Amazon S3 by adding strongly consistent object listing and the ability to append to existing objects. Gain practical insights on mapping ideas into keys and values, enabling you to craft innovative solutions using FoundationDB's flexible architecture.
Overview
Syllabus
Intro
Database Construction
Detour: CAP
Consistency using Foundation DB
(Relational) Schema Objects
Tuple Encoding CREATE TABLE Students
Index Tuples
Index Key Tuples
Non-blocking Schema Changes
Metadata Version Feature
Online, Asynchronous Schema Change in F1
Adding an Index
Change Data Capture
Implementing CDC in FDB
What are "old" and "new"?
CDC Key Structure
Feature Request
Amazon S3 Limitations
So what can S3 do?
How to fix S3: Listing
How to fix S3: Put
How to fix S3: Delete
Taught by
Linux Foundation
