Availability sets for redundancy in Azure - Step-by-Step Execution

Process Flow - Availability sets for redundancy

Create Availability Set

↓

Assign VMs to Availability Set

↓

VMs distributed across Fault Domains

↓

VMs distributed across Update Domains

↓

If hardware or update failure

↓

Only subset of VMs affected

↓

Other VMs remain running

↓

Service stays available

Create an availability set, place VMs inside it to spread them across fault and update domains, so failures affect only part of the VMs, keeping service running.

Execution Sample

Azure

az vm availability-set create --name myAvSet --resource-group myRG --location eastus
az vm create --name vm1 --resource-group myRG --availability-set myAvSet --image UbuntuLTS
az vm create --name vm2 --resource-group myRG --availability-set myAvSet --image UbuntuLTS

Creates an availability set and two VMs inside it, ensuring they are spread across fault and update domains for redundancy.

Process Table

Step	Action	Result	VM Distribution	Service Status
1	Create availability set 'myAvSet'	Availability set created with 3 fault domains and 5 update domains	No VMs yet	No service yet
2	Create VM 'vm1' in 'myAvSet'	VM 'vm1' created	vm1 in Fault Domain 0, Update Domain 0	Service partially available
3	Create VM 'vm2' in 'myAvSet'	VM 'vm2' created	vm2 in Fault Domain 1, Update Domain 1	Service fully available
4	Simulate hardware failure in Fault Domain 0	VM 'vm1' affected	vm2 unaffected in Fault Domain 1	Service still available via vm2
5	Simulate update in Update Domain 0	VM 'vm1' rebooted	vm2 running normally	Service still available via vm2
6	All updates and failures handled	Both VMs running after recovery	vm1 and vm2 running	Service fully available

💡 Execution stops after all VMs are created and redundancy scenarios show service availability maintained.

Status Tracker

Variable	Start	After Step 2	After Step 3	After Step 4	After Step 5	Final
Availability Set	None	Created with 3 fault domains, 5 update domains	Same	Same	Same	Same
VM vm1 Status	Not created	Running in FD0 UD0	Running in FD0 UD0	Failed (hardware)	Rebooting (update)	Running
VM vm2 Status	Not created	Not created	Running in FD1 UD1	Running	Running	Running
Service Availability	None	Partial (1 VM)	Full (2 VMs)	Full (vm2 only)	Full (vm2 only)	Full (both VMs)

Key Moments - 3 Insights

Why does the service stay available even if one VM fails?

Because VMs are placed in different fault domains, failure in one domain affects only some VMs, others keep running (see execution_table step 4).

What is the difference between fault domains and update domains?

Fault domains separate hardware to avoid simultaneous failures; update domains separate VMs for rolling updates. Both ensure not all VMs go down at once (see concept_flow).

Can VMs outside an availability set get the same redundancy?

No, only VMs in the same availability set are distributed across fault and update domains for redundancy (see execution_table steps 2 and 3).

Visual Quiz - 3 Questions

Look at the execution_table, at which step does VM 'vm2' get created?

A Step 1

B Step 3

C Step 2

D Step 4

💡 Hint

Check the 'Action' and 'Result' columns in execution_table rows for VM creation.

According to variable_tracker, what is the status of VM 'vm1' after step 5?

A Rebooting (update)

B Failed (hardware)

C Running

D Not created

💡 Hint

Look at the 'VM vm1 Status' row under 'After Step 5' column in variable_tracker.

If we add a third VM to the availability set, how would the service availability change after a fault domain failure?

A All VMs would fail simultaneously

B Service would be unavailable

C Service would still be available via VMs in other fault domains

D Service availability depends on update domains only

💡 Hint

Recall from concept_flow and execution_table how fault domains isolate failures.

Concept Snapshot

Availability sets group VMs to spread them across fault and update domains.
Fault domains isolate hardware failures.
Update domains isolate planned maintenance.
This setup keeps service running if some VMs fail or reboot.
Use 'az vm availability-set create' and assign VMs to it.

Full Transcript

An availability set is created first. Then VMs are assigned to it. The availability set divides VMs into fault domains and update domains. Fault domains separate hardware racks to avoid simultaneous hardware failure. Update domains separate VMs for rolling updates. When a hardware failure or update happens, only VMs in one domain are affected. Other VMs keep running, so the service stays available. This redundancy improves reliability. The example shows creating an availability set and two VMs inside it. When one VM fails or reboots, the other VM continues serving. This keeps the service up during failures or maintenance.