In the Create availability set dialog, not only is it possible to create an Availability Set with 1 Fault Domain/1 Update Domain, but if you select 1 Fault Domain, you are forced to select 1 Update Domain.
I don't understand what redundancy is achieved with this. Is there a hidden third criteria that azure isolates the VMs on?
Is there a hidden third criteria that azure isolates the VMs on?
There is no hidden third criteria that Azure uses to isolate virtual machines in an availability set with 1 fault domain and 1 update domain.
While creating availability set with 1 fault domain by default update domain must be set as 1 and the same physical infrastructure is used by all of the virtual machines in the availability set, and they are all updated or restarted instantly like below.
By combining 1 fault domains and 1 update domains, basically means creating one logical group of virtual machines that are updated or rebooted at the same time and using same physical infrastructure. This does not give any redundancy or protection over hardware failures or scheduled maintenance events.
It is good to use multiple fault domains and update domains for optimal redundancy and high availability in your Azure infrastructure.
You can configure up to 3 fault domain and 20 update domains by increasing these values, it reduces all your VMs being affected by a single fault or maintenance event. This enhances the availability and resilience of your Azure-based application.