Skip to main content

[ep 5] How to Navigate Data Issues and Isolation Levels

In this episode of The Cockroach Hour, you'll learn how to navigate data issues and isolation levels. Specifically, we'll explain how data issues and isolation levels impact data security and application performance. Please use the chapter headings below to jump around in the video if need be. Dirty reads, phantom reads, write skew, ACID rain… These are often overlooked, but troubling data issues that might plague your database. Each is important to comprehend so you can know what and how data can be wrong in your database. More importantly, are you familiar with isolation levels in a database and the potential problems for each setting? In this session, we walk through a definition of each of these potential data issues and then map them to isolation levels. We will then speak to how CockroachDB delivers serializable isolation in a distributed environment to ensure these issues won’t plague your application. -------------------------------------------------------------------------------------------------------------------------------- 00:00 Introductions 3:07 Why do isolation levels matter in the database? 4:24 What is the difference between Isolation and Consistency in databases? 5:47 What does ACID stand for? 6:15 What is atomicity? 7:44 What are the database isolation levels? 8:45 What is serializable isolation? 9:47 what is read committed isolation? 10:23 Linearizable isolation vs Serializable Isolation 11:46 What developers need to know about isolation levels 14:25 Impact of isolation levels on data security 16:09 What is CockroachDB? 17:33 Impact of database isolation levels on application performance 20:05 Why does CockroachDB use Serializable Isolation? 23:00 How does CockroachDB do Serializable Isolation? 27:04 What to do with dead locks? 27:54 When to use retry logic 31:04 What is MVCC? --------------------------------------------------------------------------------------------------------------------------- Careers: CockroachCloud: Blog: Docs: Community Slack: Twitter: