We're sorry but this page doesn't work properly without JavaScript enabled. Please enable it to continue.
Feedback

Online schema change at scale in TiDB

Formal Metadata

Title
Online schema change at scale in TiDB
Subtitle
How does schema changes work in a distributed SQL database
Title of Series
Number of Parts
542
Author
License
CC Attribution 2.0 Belgium:
You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor.
Identifiers
Publisher
Release Date
Language

Content Metadata

Subject Area
Genre
Abstract
Schema changes at scale must keep the data accessible and cannot block any clients from using a table. TiDB, a distributed MySQL compatible database, solves this by transitioning through compatible states, so clients can transition to the new state asynchronously and continue to use the data. PingCAP, the developers of TiDB, has evolved the schema changes implementation to be both faster and more tuneable, which will be shown in the presentation. The presentation will explain things like why a delete only state is needed, also how something like REORGANIZE PARTITION can be executed online without blocking. New development like generating and ingesting SST files for speeding up data reorganization for ADD INDEX and distributing the data reorganization load.