Now I read the copy there.
I want to delete the one that failed.
Is this just a case of terminating it? Does it affect the master device or other read-only replicas?
There has never been such a thing before, so I want to make sure to do it the right way
In MySQL replication, the connection between the master server and the replica server is initiated from the replica server (“slave server”) server side-if the replica server goes offline, the master server just closes the connection and continues its business; even It is random, and the owner in the copy basically has no risk of disappearing.
Specifically, in RDS, the copy seems to be obviously designed as a one-off, and RDS also provides “promotion” The ability to replicate – this is not to promote becoming a “master” (replace the existing one), but to promote becoming a “one” master, an independent system that is completely disconnected from the previous relationship with its predecessor.
If it makes you more comfortable, you can “upgrade” the copy first, and then delete it when you see that the process is complete, but simply deleting the copy should be harmless.
http:// aws.amazon.com/about-aws/whats-new/2012/10/11/amazon-rds-mysql-rr-promotion/
I recently had an RDS The read copy crashed, so a new read copy needs to be started.
Now I read the copy there.
I want to delete the one that failed.
p>
Is this just a case of terminating it? Does it affect the master device or other read-only replicas?
There has never been such a thing before, so I want to make sure to do it the right way
This should be a Very safe operation, but the last-minute snapshot of the master (or even another copy, if it is online, copy and sync) will not be harmed.
In MySQL replication, the master server and The connection between replica servers is initiated from the server side of the replica server (“slave server”)-if the replica server is offline, the master server just closes the connection and continues its business; even if it is random, the master in the replica basically does not The risk disappears.
Specifically, in RDS, the copy seems to be clearly designed to be one-off, and RDS also provides the ability to “promote” the copy-this is not to promote becoming the “master” (Replace the existing one), but to promote becoming “a” master, an independent independent system that is completely disconnected from the previous relationship with its previous master.
If it makes you more comfortable, you can first ” Upgrade the” copy, and then delete it when you see that the process is complete, but simply deleting the copy should be harmless.
http://aws.amazon.com/about-aws/whats- new/2012/10/11/amazon-rds-mysql-rr-promotion/