After pouring over the interwebs for several days - I tried something that works!
Let's say the corrupted instance is called "badinstance". It has an endpoint called: badinstance.c5taqjjaja1d.us-west-2.rds.amazonaws.com:3306
Once you have restored the database to point in time, RDS will create a new instance, let's say you called it "newinstance".
Your Elastic Beanstalk will still be pointing to badinstance.c5taqjjaja1d.us-west-2.rds.amazonaws.com:3306
First delete "badinstance" - scary I know - I did multiple backups before I actually hit "delete.
Then click on "newinstance" and select modify. You will be able to change the instance identifier to "badinstance". Make sure at this point that the security groups on the db instance match the group you have to Elastic Beanstalk.
The name changing took a long time to take place (I made the request around 3 pm or so and saw that it went through at around 3:50 am the next morning).
I restarted the Elastic Beanstalk app-server and voila: the data had been restored.
Hopefully this will be helpful to someone down the road.
Sam
与恶龙缠斗过久,自身亦成为恶龙;凝视深渊过久,深渊将回以凝视…