When creating an Azure Virtual Machine, you will be presented with a wide choice of codes from A0 to M128s. These represent the intended use and configuration of your virtual machine; basically, how many cores, RAM and storage it includes but there are other intricacies to this as well. Your choice depends on the workloads you want to run on the virtual machine. The most important thing is that you understand what the virtual machine will be used for. Once this decision is made, the IT architect can select the series and the size of virtual machine.
How does the process of Virtual Machine selection differ from sizing on premise Virtual Machines? The machine will need as much RAM, CPU and disk as your operating system and applications will consume and in this respect, the selection of Azure Virtual Machine is identical to the process of selecting the sizes and configuration of on-premises physical or virtual machines currently.
One key aspect of Virtual Machine selection that is different, however, is that the Azure cloud environment allows the IT architect to scale. With some restrictions, you can scale your virtual machine up to a more powerful instance or down to a less powerful and cheaper virtual machine. Azure Virtual Machines also offer high availability (HA) via scale-out. For the on-premises architecture, this would require densely packed hardware and the IT team would have to take care of the Virtual Machine hosts, networks and storage whilst also thinking about redundancy and ensuring that the virtual machines were running at all times. Azure is different because the cloud takes care of that work for the IT team and offers high availability as part of that process.
Azure allows organizations to be cost-effective by setting up a group of smaller machines which share workloads and can be turned on or off according to demand or on a timed schedule. Effectively, Azure charges for the compute power you are using when the virtual machines are turned on and doesn’t charge for virtual machines that are turned off. The organization is only paying for any persistent storage or networking of the virtual machines when they are powered off, but not for unused compute power.
Selecting a Virtual Machine Size
To select the correct Virtual Machine series, the IT architect will need to know the intended workload. Each virtual machine type is optimised to run a different workload, so it’s essential that this planning is done first. For example, if you are looking for a virtual machine that can work with Big Data solutions, then the organization should select a virtual machine from the High Performance Compute VM series. At the time of writing, Microsoft offers six virtual machine types:
General Purpose – Balanced CPU-to-memory ratio
Compute Optimised – High CPU-to-memory ratio
Memory Optimised – High memory-to-CPU ratio
Storage Optimised – High disk throughput and IO
GPU – Specialised virtual machines for heavy graphics rendering and video editing
High Performance Compute – Fastest, most powerful CPU with optional high-throughput network interfaces (RDMA)
Once the series has been selected, the IT architect can choose the virtual machine size.
Selecting a Virtual Machine Size
One key piece of advice to note is that if the organization believes that they may need to move up to another larger virtual machine in the future, then it is best to check that the larger machine is available in the same hosting region (e.g. UK South, West US) as the original virtual machine. Otherwise, the organization will have to move the virtual machine to the new region. Although it’s not an onerous task to move a virtual machine from one region to another, it is best to avoid if possible.
The following table will help the IT architect to identify the correct size of virtual machine for the requirements.
To summarise, choosing an Azure Virtual Machine is a crucial part of the transition to cloud. There is a good choice available and you have the ability, with some restrictions, to switch in the future as your needs change.