I understand that for Azure SQL PaaS offering, for redundancy purposes, we have Geo-replication feature.
1)Can we create a read-replica of the target replicated database ? or is it only Active-active geo replica
2)Also for geo-replication of an Azure SQL DB, is there any requirements or prerequisites to be met like all tables should have primary key or something like that ?
3) Also i think geo-replication is asynchronous . So how much delay is there in replicating data to secondary replica ?