What exactly happens when I use the console and create a replica?
The new instance is launched as part of the cluster and has access to the main data - or, more precisely, to the cluster data. All Aurora instances are members of a "cluster", even if it is only a cluster of one core server. Aurora replication in the same region is very different from basic MySQL replication.
Will a new instance appear and start retrieving data from the primary instance?
Not really. As described above, a new instance will appear and be able to read from the main backup storage - it does not have its own storage.
Aurora operates on 3 sets of 2 copies of operational data, is mirrored and replicated in the availability zones in the region. This logical entity is called a cluster volume.
The cluster volume covers several availability zones in one region, and each availability zone contains a copy of the cluster volume data.
http://docs.aws.amazon.com/AmazonRDS/latest/UserGuide/Aurora.Managing.html
(Documents say that each AZ contains a “copy” that is true but mirrored.)
Aurora replicas read from this data - for all practical purposes synchronously.
Q: How far behind will my cues be?
Because Amazon Aurora Replicas uses the same amount of data as the primary, there is virtually no replication delay. Usually we observe lag times of 10 s milliseconds.
- https://aws.amazon.com/rds/aurora/faqs/
Could this affect database performance?
It should not be.
I am sure it will take some time before the replica “catches up” and loads 250 GB; How do I know when he "finished"?
No, it really shouldn't. Once the replica instance is available, it needs to be updated as it reads the same data from the same place as the master. Aurora replica latency metrics are available in the console.