Session Abstract: In the age of constantly active applications, attaining high availability with the least possible downtime is of utmost importance. The Oracle Database 23ai introduces a revolutionary feature that enables sub-second failover with zero data loss in distributed database environments. Raft replication allows for this timing of failover.
This session will explore Raft replication in-depth, detailing its consensus-based commit protocol and how it facilitates rapid, automatic failover. Attendees will learn about the active/active replication model, where each shard can process reads and writes, eliminating the need for primary and standby configurations.
We will delve into the architecture of replication units (RUs), leader and follower distributions, and how Oracle manages workload balancing and hardware utilization. The session will also cover the Raft log mechanism, leader election process, and the handling of node failures to ensure seamless operations.
Speaker Bio: As an Oracle DBA Team Leader, I manage complex UNIX-based Oracle databases for mission-critical applications and systems. With over 23 years of experience in database administration, I have developed expertise in high availability and performance tuning and troubleshooting, using technologies such as RAC, Dataguard, and Exadata. I am also an Oracle ACE Pro, a prestigious recognition of my contributions to the Oracle community as an author, speaker, and mentor. Furthermore, I am passionate about sharing my knowledge and best practices with the Oracle community, and I have received multiple awards from Oracle as a Most Valued Contributor.
https://www.aioug.org/events/achieving-sub-second-failover-with-raft-replication-in-oracle-database-23ai
