You are the IT administrator for a small law firm. The company has one lawyer and one legal assistant. The company has two Windows 10 Professional desktop computers and a Linux server that hosts a web-based case management system.
The two desktop computers and the Linux server are connected by a network hub. The hub itself is connected to a router, which connects directly to the Internet via cable. No inbound ports are open on the router. The desktop computers host client applications that connect to the case management system at IP address 10.10.10.10 over TCP port 24000.
The owner of the firm wants it to transition to a virtual firm. The lawyer and the assistant must be able to work from home by connecting to the Windows 10 desktop computers from any device. The owner wants you to move the existing infrastructure to Azure and make the system work as if it were in the physical office. However, the owner wants to use the minimum amount of resources and the least expensive options.
The two computers and server should be imported into Azure as virtual machines (VMs). The VMs for the lawyer and assistant should be always available, even during periods of upgrades or maintenance. As more cases are imported into the case management system, the disk attached to the Linux VM should automatically resize to ensure that it always has 20 percent of free space.
You need to meet the availability demands for Windows computers.
What should you do?
You should create one availability set for each VM. An availability set allows you to group VMs for availability. For example, the first availability set can contain the Windows 10 computer for the assistant, with additional VM instances for failover support. The second availability set can contain the Windows 10 computer for the lawyer, with additional VM instances for failover support.
You should not create one availability set for both VMs. This would cause the lawyer's VM to be used when the assistant's VM is being upgraded, and vice versa.
You should not implement horizontal auto-scaling. Horizontal auto-scaling allows more VMs to be created as load on a particular VM increases. It does not provide failover support.
You should not implement vertical auto-scaling. Vertical auto-scaling allows more resources to be added to a VM as load on a particular VM increases. It does not provide failover support.