Hello,
I setup SQL HA in Azure with replicas in different DCs. Obviously issue come with what IP I use for connectivity. There were so many different things to setup (internal load balancers, new listeners etc). My question is why to go to all this complexity instead of just using SQL mirroring connection string instead where Primary and Secondary servers are specified instead with direct internal IPs for each SQL server? It does in fact work very well and I'm confused what is wrong/drawback of this setup compared with connection string going to AG listener instead?