Describe virtual machine resizing
There are many size options for Azure Virtual Machines. For SQL Server workloads the main characteristics to look for are the amount of memory available, and the number of input and output operations (IOPs) the virtual machine can perform.
Using constrained cores
Typically SQL Server is licensed by the core and Azure will use a fixed ratio of CPU cores to memory. However, you may have workloads that require large amounts of memory, but do not require the default amount of allocated CPUs to get that amount of memory. In these cases it may be useful to use Azure's constrained cores.
With constrained cores, you can reduce the cost of software licensing while still getting the full amount of memory, storage, and I/O bandwidth. This is good for database workloads that are not CPU-intensive and can benefit from high memory, storage, and I/O bandwidth, while using a constrained vCPU count.
Using general purpose virtual machines
Most SQL Server production workloads will run on the general purpose or memory-optimized families of Azure Virtual Machines. Larger workloads requiring more memory and/or CPU resources will land in memory-optimized virtual machines, but many production applications can run comfortably on general purpose virtual machines.
Resizing virtual machines
Azure supports resizing your virtual machine. This operation does require a restart; however, restarting a virtual machine is typically a fast process. In some cases, depending on what virtual machine type you are switching to and from, you may need to deallocate your virtual machine and then resize. This operation does extend the duration of the outage but should not take more than a few minutes.