Hi, I have a query regarding VMSS behavior. Let's assume I have a VMSS configured with Multi-Availability Zone (1,2,3) but with 1 fixed instance. Now I will see my 1 fixed instance configured in any one of the availability zones from 1,2,3 for which I do not have any control. Let's assume the single instance is running on Zone2, what happens when Zone2 is down? Will the instance come up automatically in Zone1 or Zone3 which must be available, because I configured with fixed single instance? or I will see the VMSS is down.

Hi Jeyasingh, Vijayalingam , Welcome to the Microsoft Q&A Platform! Thank you for asking your question here. Behavior in the Event of Zone 2 Going Down In case, you have configured a fixed single instance, and it is running in Zone 2 . If Zone 2 goes down, that specific instance will become unavailable. The instance will not automatically come up in Zone 1 or Zone 3. The VMSS does not have the capability to relocate or recreate the instance in another zone when only one instance is specified. You will effectively see the VMSS as down because there are no operational instances available to handle requests. The scale set cannot provide redundancy or failover capabilities with a single instance. Here is a more detailed explanation of what happens when a zone goes down: Azure monitors the health of all zones and VMs in a VMSS. If a zone goes down, Azure will automatically detect the failure. Azure will then attempt to restart the VM in the failed zone. If Azure is unable to restart the VM in the failed zone, it will create a new VM in another available zone. Azure will then copy the state of the old VM to the new VM. Once the new VM is ready, Azure will switch traffic to the new VM. To minimize the impact of a zone failure, you can use the following best practices: Use a zone redundant VMSS. This will ensure that your VMs are distributed across multiple zones, so that if one zone goes down, your VMs will still be available in other zones. Use auto scaling. This will ensure that your VMSS automatically scales up to meet demand, even if a zone goes down. Use a load balancer. This will distribute traffic across your VMs, so that if one VM goes down, traffic will be automatically routed to other VMs. https://learn.microsoft.com/en-us/azure/virtual-machine-scale-sets/virtual-machine-scale-sets-use-availability-zones?tabs=portal-2 If you have any further queries, do let us know. If the answer is helpful, please click "Accept Answer" and "Upvote it."

How does VMSS behave in the Zonal failure Scenario?

Accepted answer

Mounika Reddy Anumandla 830 Reputation points Microsoft Vendor

2024-10-07T10:47:41.9833333+00:00
Hi Jeyasingh, Vijayalingam,

Welcome to the Microsoft Q&A Platform! Thank you for asking your question here.

Behavior in the Event of Zone 2 Going Down

In case, you have configured a fixed single instance, and it is running in Zone 2. If Zone 2 goes down, that specific instance will become unavailable. The instance will not automatically come up in Zone 1 or Zone 3. The VMSS does not have the capability to relocate or recreate the instance in another zone when only one instance is specified. You will effectively see the VMSS as down because there are no operational instances available to handle requests. The scale set cannot provide redundancy or failover capabilities with a single instance.

Here is a more detailed explanation of what happens when a zone goes down:

Azure monitors the health of all zones and VMs in a VMSS.

If a zone goes down, Azure will automatically detect the failure.

Azure will then attempt to restart the VM in the failed zone.

If Azure is unable to restart the VM in the failed zone, it will create a new VM in another available zone.

Azure will then copy the state of the old VM to the new VM.

Once the new VM is ready, Azure will switch traffic to the new VM.

To minimize the impact of a zone failure, you can use the following best practices:

Use a zone redundant VMSS. This will ensure that your VMs are distributed across multiple zones, so that if one zone goes down, your VMs will still be available in other zones.

Use auto scaling. This will ensure that your VMSS automatically scales up to meet demand, even if a zone goes down.

Use a load balancer. This will distribute traffic across your VMs, so that if one VM goes down, traffic will be automatically routed to other VMs. https://learn.microsoft.com/en-us/azure/virtual-machine-scale-sets/virtual-machine-scale-sets-use-availability-zones?tabs=portal-2

If you have any further queries, do let us know.

If the answer is helpful, please click "Accept Answer" and "Upvote it."
Please sign in to rate this answer.
Jeyasingh, Vijayalingam 81 Reputation points

2024-10-07T15:15:37.7233333+00:00

@Mounika Reddy Anumandla Thanks for your response. Yes, it certainly helps. I was thinking to go ahead for VMSS with one fixed instance, as my solution has Windows Services running in an instance and I have the requirement to keep it running always. So, in my case, Is VM a better option where I could leverage the Zone-to-Zone site recovery?

Mounika Reddy Anumandla 830 Reputation points Microsoft Vendor

2024-10-07T17:12:01.4033333+00:00

Hi
Thank you for reaching us back again.
Zone-to-Zone site recovery is a disaster recovery feature provided by Azure Site Recovery. It allows you to replicate and recover virtual machines (VMs) between Availability Zones within the same Azure region. This feature is designed to provide automatic recovery of your VMs in another zone in case of an outage or disaster.
Advantages of Zone-to-Zone site recovery:

Disaster Recovery: Azure Site Recovery allows you to replicate VMs across availability zones within the same region. This setup ensures that your services can continue running even if one zone experiences an outage.

Recovery Point Objective (RPO) and Recovery Time Objective (RTO): Site Recovery offers a low RPO (approximately 5 minutes) and a low RTO (less than 1 hour), making it suitable for critical applications that require quick recovery.

Complexity and Cost: This option can be more complex and potentially more costly due to the need for replication and failover orchestration.

If your primary requirement is to ensure that your Windows Services are always running with minimal downtime, leveraging a VM with Zone-to-Zone site recovery is likely the better option. It provides robust disaster recovery capabilities and ensures that your services remain available even if one zone experiences an outage.

https://learn.microsoft.com/en-us/azure/site-recovery/azure-to-azure-how-to-enable-zone-to-zone-disaster-recovery

If you need high availability and scalability for a stateless application, VMSS with zone redundancy could be more suitable.

If you have any further queries, do let us know.

If the answer is helpful, please click "Accept Answer" and "Upvote it."

Jeyasingh, Vijayalingam 81 Reputation points

2024-10-08T05:14:04.58+00:00

@Mounika Reddy Anumandla - Thanks much for the clarification provided. That helps.

Jeyasingh, Vijayalingam 81 Reputation points

2024-10-08T11:44:28.0266667+00:00

@Mounika Reddy Anumandla - Another quick question from TCO perspective when I go for Zone-to-Zone Site recovery. Will I be charged for only 1 VM as I deploy VM with 1 Availability Zone (or) I will be charged for 2 VMs as we are configuring Disaster recovery to Secondary Zone?

Mounika Reddy Anumandla 830 Reputation points Microsoft Vendor

2024-10-09T04:19:57.7766667+00:00

Hi Jeyasingh, Vijayalingam,

Thank you for reaching us back again:-)

Azure Site Recovery is billed in units of the average daily number of instances you are protecting over a monthly period. For example, if you consistently protected 20 instances for the first half of the month, and none for the second half of the month, the average daily number of protected instances would be 10 for that month.
When considering Zone-to-Zone Site Recovery in Azure, the billing structure is based on the number of virtual machines (VMs) being protected. In a Zone-to-Zone Site Recovery scenario, you would typically be charged for both VMs.
Pricing for zone-to-zone disaster recovery is identical to the pricing for Azure-to-Azure disaster recovery. You can find more details on the Azure Site Recovery pricing page and in this blog post.

https://learn.microsoft.com/en-us/azure/site-recovery/azure-to-azure-how-to-enable-zone-to-zone-disaster-recovery#faqs
If you have any further queries, do let us know.

Jeyasingh, Vijayalingam 81 Reputation points

2024-10-09T04:58:22.7966667+00:00

@Mounika Reddy Anumandla - Thanks for the details. However, it did not answer me completely and I see some missing dots.

Let me take the scenario mentioned in the blog that you have shared me, where I have the requirement to Protect 6 VM's, 12 HDD, 6 Premium SSD. Assuming I am going for a Zone-to-Zone site recovery (between Zone1 & Zone2), in this case the egress cost will not apply for me as all the network traffic is within the same region. So, my understanding is, I will incur cost for the following:

Azure Site Recovery ($25 per instance)

Storage cost of 12 HDD (in both Zone1 and Zone2)

Storage cost of 6 Premium SSD (in both Zone1 and Zone)

Other than the above cost, I am expecting the Cost for VM that will be running in any one of the Zone (Zone1 or Zone2) because the VM in another Zone would be down and stopped status. So preciously only 1 VM will be running at any given point of time.

Is my understanding being correct?

Mounika Reddy Anumandla 830 Reputation points Microsoft Vendor

2024-10-09T16:05:16.12+00:00

Hi Jeyasingh, Vijayalingam,

Thank you for reaching us back again for further clarification.

Yes, your understanding is correct. During an actual disaster recovery event, you will incur costs only for the active VM that is running in one of the zones, along with ongoing ASR licensing and storage costs for all replicated data. The other VMs will not generate compute charges while they are stopped or deallocated. But the cost will be incurred on services associated with the VM. Egress costs do not apply in your case since all traffic remains within the same region. However, be mindful of additional resource costs for any resources such as load balancers and public IPs created beforehand which are required for the application along with the VMs to work properly when failed over.
you can use this pricing calculator: https://azure.microsoft.com/en-us/pricing/calculator/

Jeyasingh, Vijayalingam 81 Reputation points

2024-10-10T04:27:06.2533333+00:00

@Mounika Reddy Anumandla - Thanks much for the clarification provided. Appreciate your timey help.
Sign in to comment

Use comments to ask for clarification, additional information, or improvements to the question.

Share via

How does VMSS behave in the Zonal failure Scenario?

0 additional answers

Your answer