I would like to use two availability zones to run two identical EC2 instances with identical data (A side and B side) (hot-hot).
This means that both of the EBS volumes (A and B) will be identical (assuming both AZ's are available) and hence the EBS snapshots will be the same. I am looking for suggestions on how I could only store one EBS snapshot in S3 backed up by the A side, and if something went wrong with the AZ of the A side then I would like the B side to take over and add as an increment to the snapshot previously created from the A side.
I was thinking I could have a lambda function which checks when a successful upload from the A side occurs in S3, and if it fails a snapshot will be created by the B side and the A side snapshot would be deleted but there is a lot of data and will take a long time to run a snapshot over the entire EBS volume hence I am looking for a way to merge another (identical) EBS volume into a snapshot.
Cheers!
