Hi guys!
I have a one question about high availability .
I create a HA design for RDS access, but I need a some clarify.
There're 2 geo-locations (like 2 datacenters) with 2 independent service instances. Each one has a Domain Controller (the second one has a DC Replica). Each one has a server with RDWA, RDGW, RDCB, RD Licensing roles. And a RDSH server farm that store User Profile
Disk on the file share for each datacenter.
Domain Controller has a Domain Controller Replica for HA.
Both servers with a RDWA,RDGW,RDCB,RDLC roles has shared DNS record (3rd party feature that checks availability of web-server and determine a IP address)
UPD are directly synchronized via 3rd party application.
So,
If we have a RDSH server fail - RDCB will reconnect users session to another RDSH server. It's ok.
If server with RDWA role fail then IP address will be changed via DNS - it's ok.
BUT what will happen if both RDWA servers are fine and one of the RDSH farm is totally fails?is it possible if RDSH servers are unavailable to make the RDWA server unavailable too?
Thank you!
Best regards,
Anton