We are using Peer to Peer (P2P) replication on SQL 2012.
Setup is with 3 nodes NodeA, NodeB and NodeC ( on prem )
We isolation the users by geo location and we are using P2P replication as DR/HA and to improve the performance.
I am planning to upgrade them to SQL Server 2017 or SQL 2019 but planning to add one more node to expand the SQL Server to Azure.
Adding additional node will add extra replication activity on each node because each node have to talk to other three nodes in four node P2P and
I believe we can't control the flow of replication from one to another in P2P setup unless there is any change in SQL 2017 or SQL 2019.
Is it possible setup P2P replication between three nodes (NodeA, NodeB and NodeC ) and
create separate two node P2P replication between to nodes ( NodeC and NodeD (Azure) ) ?
In this model NodeA and NodeB will not get the data if something goes wrong to NodeC ( I am OK with it).
Second question is.
Is it possible to create each P2P Peer node on a cluster within each datacenter for local HA while using the P2P replication using multiple datacenters?
Thanks,


