Skip to main content

Surviving IT Downtime | What is RTO? | Traditional Failover Prep

In this live conversation, we discuss the reality that Everything Fails and we'll talk about how ways to survive database disasters and the unfortunate constraint of planned downtime. The point of this conversation is to shift your thinking from planning for failure to planning for survival. With CockroachDB you will survive any failures and you will not have any planned downtime. Learn how we're making that possible. 00:00 Everything Fails 00:59 What is the cost of IT downtime? ($56,000/minute) 4:15 What is RTO? 4:54 What is RPO? 6:22 How do traditional database systems prepare for failure? 9:55 What do databases need to be able to survive? 10:20 What's the most common cause of database failures? 12:30 What's the difference between Relational, NoSQL and Distributed SQL databases 14:31 What are the characteristics of a Distributed SQL database? 16:55 What is the problem with eventual consistency? 20:25 Why is CockroachDB named after a cockroach? 21:30 Why do developers like CockroachDB? 22:53 How does CockroachDB survive failures? 24:00 How do all nodes in CockroachDB act as gateway nodes? 26:30 What is the Key-Value store that supports CockroachDB? 29:19 Why do people choose CockroachDB? 31:47 How does Cockroach guarantee consistent transactions when a read comes from one reason and the write comes from a different region? 37:44 What can CockroachDB survive? 39:37 How does CockroachDB save you money? 41:25 How do nodes recover after an outage? 44:20 What is the difference between legacy database architecture and modern database architecture? Relevant Links: What is distributed SQL?: https://www.cockroachlabs.com/blog/what-is-distributed-sql/ How important is RPO/RTO: https://www.cockroachlabs.com/blog/demand-zero-rpo/ Fault Tolerance and Recovery in CockroachDB: https://www.cockroachlabs.com/docs/stable/demo-fault-tolerance-and-recovery.html