Examkin

Question

Difficulty: HardDesign Compute High Availability

An enterprise is designing the compute architecture for a high-performance distributed database on Azure. The solution must meet the following requirements:

- Achieve a virtual machine uptime SLA of at least 99.99%99.99\%.
- Provide resiliency against a complete datacenter zone failure.
- Minimize network latency between the active database nodes within each region.

Which compute configuration should you recommend to meet these requirements?

  1. Deploy Virtual Machine Scale Sets in Flexible orchestration mode across 33 Availability Zones, utilizing a dedicated Proximity Placement Group (PPG) for each zone.Answer
  2. B
    Deploy all virtual machines within a single Availability Zone and group them using a single Proximity Placement Group (PPG).
  3. C
    Deploy all virtual machines across 33 Availability Zones using a single region-wide Availability Set.
  4. D
    Deploy all virtual machines across 33 Availability Zones associated with a single Proximity Placement Group (PPG) that spans all 33 zones.

Answer

Deploy Virtual Machine Scale Sets in Flexible orchestration mode across 33 Availability Zones, utilizing a dedicated Proximity Placement Group (PPG) for each zone.
Deploying Virtual Machine Scale Sets in Flexible orchestration mode across 33 Availability Zones satisfies the 99.99%99.99\% uptime SLA and zone-level resiliency requirements. Using a dedicated Proximity Placement Group (PPG) for each zone ensures that the virtual machines within each zone are colocated as closely as possible to minimize intra-zone network latency without causing deployment allocation failures.

Step-by-Step Solution

1
Analyze the SLA and high availability requirements.
To achieve a 99.99%99.99\% uptime SLA and survive a zone failure, the VMs must be deployed across at least 22 (and ideally 33) Availability Zones.
Azure guarantees 99.99%99.99\% SLA only when VMs are deployed across multiple Availability Zones in the same region.
2
Evaluate the latency requirement in a multi-zone deployment.
Using a single Proximity Placement Group (PPG) across multiple zones does not reduce latency between zones and can cause deployment failures.
PPGs are designed for low-latency colocation within a single physical datacenter (zone). A single PPG cannot colocate VMs across physically separate zones.
3
Select the optimal configuration that combines zone redundancy and low latency.
Deploy VMSS in Flexible orchestration mode across 33 zones with a dedicated PPG per zone.
This setup allows zone redundancy for the cluster while ensuring ultra-low latency for the subset of nodes operating within each individual zone.

Key Concept

Designing compute high availability using Availability Zones and Proximity Placement Groups.
Rate this question

Topics

Design Compute High Availability

More questions from this topic