Microsoft Azure Infrastructure as a Service (IaaS) provides Virtual Machines (VMs) workload with powerful resiliency features.
Those features will help you to make your Virtual Machine workloads available during planned maintenance or unplanned failures.
Following are the availability options Azure provides
Availability Sets
We can implement an Availability Set and place multiple virtual machine nodes in order to make them available despite rack level or host level failures. There are two important concepts to consider.
Fault Domain : Group of virtual machines that share a common power source and network switch. Essentially that is the server rack.
Update Domain : Group of virtual machines and underlying physical hardware that can be rebooted at the same time. Essentially that is the physical server in a rack. Microsoft might reboot the server after a planned maintenance.
If we had all of our virtual machine nodes in a single update domain, the resource will not be available during a planned maintenance activity.
And if we had all of our virtual machines in a single fault domain, our resource will be unavailable during a rack level failure.
We can place our resources in maximum 3 Fault Domains and 20 Update Domains in order to avoid such down times.
Availability Zones
Availability Sets is not the only availability option we have. Availability set will help us during rack level or server level failures and maintenance. What would happen if the entire data center goes down.
Availability Zones will come to the rescue!
Many regions support the concept of Zones. That means, a specific region supports multiple data centers. In a situation where even an entire data center goes down you have another set of data centers to support your workloads.
You can distribute your virtual machine workloads up to 3 data centers
I will discuss the concept of Virtual Machine Scale Sets (VMSS) in a separate post
