Examkin

Question

Difficulty: HardDesign Compute High Availability

An organization is designing a high-availability compute infrastructure for a mission-critical financial transactions processing application in Azure. The compute tier will run on Azure Virtual Machines.

The solution must meet the following requirements:
- Provide a virtual machine uptime SLA of at least 99.99%99.99\%.
- Support automatic scaling of the compute tier using Virtual Machine Scale Sets (VMSS) with Flexible orchestration mode.
- Ensure the lowest possible network latency between the application VMs and a database cluster.

Which two actions should you include in the design? (Select two.)

  1. Deploy the Virtual Machine Scale Set across multiple Availability Zones.Answer
  2. Create a Proximity Placement Group in each Availability Zone to co-locate the application VMs and database VMs within each zone.Answer
  3. C
    Deploy the Virtual Machine Scale Set within a single Availability Zone using five fault domains.
  4. D
    Configure the Virtual Machine Scale Set to run on Spot Virtual Machines.

Answer

To meet the requirements, you should deploy the Virtual Machine Scale Set across multiple Availability Zones and create a Proximity Placement Group in each Availability Zone to co-locate the application VMs and database VMs within each zone.
To achieve the 99.99%99.99\% SLA, Azure virtual machines must be distributed across multiple Availability Zones. To ensure the lowest possible latency between the application and database VMs within those zones, a Proximity Placement Group should be created for each zone. This combination satisfies both the high-availability SLA and the low-latency networking requirements.

Step-by-Step Solution

1
Identify the SLA requirement.
A 99.99%99.99\% virtual machine uptime SLA is required.
This SLA level can only be achieved in Azure by deploying VMs across two or more Availability Zones.
2
Determine the physical placement of resources to meet the latency requirement.
A Proximity Placement Group (PPG) must be used inside each Availability Zone.
A single PPG cannot span zones without risking latency issues or deployment failures. Therefore, separate PPGs are created for each Availability Zone to keep resources close within their respective zones.
3
Evaluate workload stability and VM pricing model.
Avoid Spot VMs and stick to standard VMs.
Spot VMs do not have an SLA and can be evicted randomly, which violates the high-availability requirement of a mission-critical platform.

Key Concept

Designing compute high availability using Availability Zones and Proximity Placement Groups to meet specific uptime SLAs and performance latency requirements.
Rate this question

Topics

Design Compute High Availability

More questions from this topic