– Intro to Storage Service Durability Architecture
2
– How do we replicate data?
3
– What is erasure coding?
4
– When do we use erasure coding?
5
– How do we select Stretch Factor values N and K?
6
– Replication design summary
7
– How to minimize disk failure impact
8
– Simple three way replication example
9
– Handling disk failures example
10
– Time to recover from disk failure
11
– Magic behind achieving 11 9s durability
12
– Minimize impact of software bugs to durability
13
– Tradeoff testing vs faster feature release
Description:
Explore the architectural approaches used in Oracle Cloud Infrastructure (OCI) Object Storage to achieve high durability in this 20-minute video blog. Dive into the principles of redundancy and recovery as OCI architects Laurion Burchall and Pradeep Vincent discuss efficient data replication, erasure coding, and fast failure recovery techniques. Learn about the replication design, minimizing disk failure impact, and the process of achieving 11 9's durability. Gain insights into handling software bugs, balancing testing with feature releases, and the overall strategy for maintaining exceptional data durability in cloud storage systems.
First Principles: Using Redundancy and Recovery for High Durability in OCI Object Storage