Oracle DataGuard - Step-by-Step - Architecture, Configuration, Switchover, Failover, DG Broker

Welcome to the Part-2 of this new series on oracle DataGuard. I will upload more tutorials in this series that will cover everything you need to know to effective manage and use Oracle DataGuard and become a master in it. My plan is to make this a minimum 10 part series but
I may increase it based on the demand and requirements. In this video I am explaining the Architecture, Configuration, Switchover, Failover, DG Broker and related commands.
Oracle dataguard is an indispensable part of oracle's high availability solutions. In a real world scenario the Primary and Phsyical standby databases reside on geographically different locations or different data centers to make sure that even if one region goes completely down, your database is still available and intact. In this tutorial and the series I am using a virtual environment for demonstration purpose.
Please note that in the DG architecture diagram, I have mentioned about the process called LNS for simplicity. It was used in older version but 12c onwards, it has been replaced by two other processes called NSS (Network Server Sync) and NSS (Network Server Async). They are used depending on where Sync or Async mode of Redo transport is used.
Steps, Commands, SQLs used :
docs.google.co...
Detailed Explanation on log_archive_dest_n parameter options:
docs.oracle.co...
DataGuard Broker Properties:
docs.oracle.co...
My other series on RUclips:
Mastering OEM 13c (Full Series):
• Mastering OEM 13c
Oracle VirtualBox Virtual Machines:
• VBox Virtual Machines

[@joseluisdelarosa728]

Great explanation on the Data Guard architecture! However, I have a question regarding the Managed Recovery Process (MRP) and its role in Real-Time Apply:
Does the MRP always apply changes directly from the redo log buffer in a real-time fashion? There seems to be a general belief that it's the
archived logs that are being applied to the standby database. Or is it that the archived logs are only applied during a 'Gap Resolution Scenario'
when there are missing redo records? I'm trying to clarify this specific aspect of how Data Guard operates. Thanks in advance for shedding light on this

[@YouVolve]

Hi @joseluisdelarosa728 - thanks for watching my video and providing your feedback.
Actually MRP does not read from redo log buffers. In real time apply, it reads from the SRLs and apply to the standby DB. If it not real-time apply, then it reads from the archived logs in the standby.
Your remaining understanding is correct and yes it is a general belief that it's the archived logs that are being applied to the standby database but in a real time apply configuration, the archived logs are applied only in a situation when the primary is generating a very huge redo data very fast causing the redo logs in primary (and the SRLs in standby) to switch and archive even before the MRP is able to read and write that to the standby DB. In that case the MRP has no option other than to go and read the archived logs for the changes it was not able to keep up the speed with.

[@joseluisdelarosa728]

Thank you so much for your detailed explanation and for taking the time to clarify my doubts.
I really appreciate the depth of knowledge you've shared, and it has enhanced my understanding
of the Data Guard's real-time apply mechanism. Your videos are incredibly helpful and educational.

[@YouVolve]

@@joseluisdelarosa728 - Glad that it helped.