Hi everybody!
I need a some help with a Cloud project on the Azure.
I want to create maximum high availability Terminal Server (VDI) service.
Everything should be available, consistent and partition tolerance. (yeah, I know about the CAP Theorem. Just as much as possible.
So, my goal is: every service role should be fault tolerance:
* AD Domain Controller, via AD Controller server and AD Replica Controller server;
* SQL Database (for HA services), via SQL Server AlwaysOn Group;
* RD Gateway, via NLB
* RD Connection Broker, via Failover Cluster feature
* RD Session Host, via Failover Cluster feature
* File Share for a user profile disk, user shares via Storage Servers with a Failover Cluster feature
Of course, a disk array for hosting SQL, RDVS, UPD, file shares.
All parts should be easy scalable.
Perhaps you can give me some advises.
I want to know:
* will it be as stable as i think?
* what roles can I combine on a instance together?
* What a sufficiently reliable disk array I can use? Can I use only a Azure Storage for my purposes? Is it reliable?
* Should I fully trust to the Azure SQL service and don't implement my solution?
* Is it some service in the Azure instead of NLB or DNS round-robin?
Thank you very much!
Best regards,
Anton