Exam Professional Cloud Database Engineer topic 1 question 48 discussion

D. Start the original primary instance and resume replication.

Agree with C

For MySQL, this should be C

For MySQL, this should be C

When a Cloud SQL read replica becomes unavailable, Cloud SQL automatically tries to create a new replica to replace the failed one. This process is part of the automated failover mechanism in Cloud SQL. B. Manually creating a new failover replica from backup might involve additional steps and may not be necessary for a temporary unavailability of the read replica.

"During a zonal outage, traffic stops to read replicas in that zone. After the zone becomes available again, any read replicas in the zone resume replication from the primary instance." https://cloud.google.com/sql/docs/mysql/high-availability#read_replicas

C is the correct answer. "During a zonal outage, traffic stops to read replicas in that zone. After the zone becomes available again, any read replicas in the zone resume replication from the primary instance." https://cloud.google.com/sql/docs/mysql/high-availability#read_replicas

B is correct answer. Creating a new failover replica from a backup is a reliable way to restore replication and ensure that the read replica is up-to-date.

C Recovery Process: Once Zone-B becomes available again, Cloud SQL will initiate the recovery process for the impacted read replica. The recovery process involves the following steps: 1. Synchronization: Cloud SQL will compare the data in the recovered read replica with the primary instance in Zone-A. If there is any data divergence due to the unavailability period, Cloud SQL will synchronize the read replica with the primary instance to ensure data consistency. 2. Catch-up Replication: The recovered read replica will start catching up on the changes that occurred on the primary instance during its unavailability. It will apply the necessary updates from the primary instance's binary logs (binlogs) to bring the replica up to date. 3. Resuming Read Traffic: Once the synchronization and catch-up replication processes are complete, the read replica in Zone-B will resume its normal operation. It will be able to serve read traffic and stay updated with subsequent changes from the primary instance.

By verifying that the new replica is created automatically, you can ensure that the read replica instance is functioning and replication is maintained even after the temporary unavailability of the replica in Zone B.

https://cloud.google.com/sql/docs/mysql/replication#:~:text=If%20replication%20is,a%20new%20one.

Resume replication

B is wrong, failover replica is NOT a read replica A and C makes no sense

How is B correct? You don't create a replica from backup, and there's no mention of HA that would point to "failover" distinction. But don't like other answers either. Will go with B unless enlightened by subsequent contributor,

B is the correct answer

Vote for B

C - makes no sense D - nobody said primary instance was offline, plus you can't stop/resture replication on the primary instance, only on read replica A makes no sense: you can't create clone database from read replica, and if it means to create one from the primary instance - how would that help to insure that replica is still working? I'll go for B