Oracle Data Guard Fast-Start Failover - The big deep dive session

👇👇Too long a video? Check the table of contents below and jump to the point of interest. 👇👇 How does Fast-Start Failover ensure that the data loss exposure doesn't pass the required RPO? How is a standby promoted to primary? Under which conditions a primary can keep committing without protection? How is a new observer promoted after the active one fails? What about split-brain? Or observer communication with the primary? In this very long video, the Oracle Data Guard product manager answers these and other fundamental questions that all Data Guard users want to understand before implementing Fast-Start Failover. 00:00 - Introduction 01:19 - The Data Guard redo transport modes overview 02:02 - The ASYNC redo transport 04:47 - The FASTSYNC redo transport 06:23 - The SYNC redo transport 07:40 - Stalling SYNC destinations 09:35 - Lagging transport vs. no transport 12:53 - Fast-Start Failover overview 22:44 - Overview of Fast-Start Failover Protection modes 24:12 - Deep Dive: Fast-Start Failover Maximum Performance mode 45:55 - Deep Dive: Fast-Start Failover Maximum Availability mode 56:11 - Deep Dive: Fast-Start Failover Maximum Protection mode 1:00:00 - Observer placement 1:06:02 - Observer high availability 1:11:50 - Multiple standbys: FSFO Target and FSFO Candidates 1:16:47 - Fast-Start Failover 21c features (that will be there in 23c as well!) If you need to implement failover automation and ensure that you comply with your RPO and RTO objectives, this video is for you. ---------------------------------------------- Copyright © 2023, Oracle and/or its affiliates.