Examkin

Question

Difficulty: HardDesign Compute High Availability

You are designing the compute infrastructure for a mission-critical web application on Azure.

The application consists of two tiers:
- A web frontend tier that must host stateless virtual machines, support automatic horizontal scaling, and guarantee a virtual machine uptime SLA of 99.99%99.99\%.
- A backend transaction processing tier that consists of stateful virtual machines, requires a virtual machine uptime SLA of 99.95%99.95\%, and requires sub-millisecond network latency between the virtual machine instances.

Which two configurations should you recommend? (Select two.)

  1. Deploy the web frontend tier using a Virtual Machine Scale Set in Flexible orchestration mode distributed across three Availability Zones.Answer
  2. Deploy the backend transaction processing tier in an Availability Set within a single Availability Zone, and associate the virtual machines with a Proximity Placement Group.Answer
  3. C
    Deploy the web frontend tier in a single Availability Zone using an Availability Set.
  4. D
    Deploy the web frontend tier using Spot virtual machines within a Virtual Machine Scale Set to optimize costs.
  5. E
    Configure a Standard Load Balancer for the web frontend tier and assign Basic SKU public IP addresses to the virtual machine instances.

Answer

Deploy the web frontend tier using a Virtual Machine Scale Set in Flexible orchestration mode distributed across three Availability Zones, and deploy the backend transaction processing tier in an Availability Set within a single Availability Zone associated with a Proximity Placement Group.
To satisfy the frontend tier's requirements of a 99.99%99.99\% SLA and auto-scaling, you must deploy a Virtual Machine Scale Set in Flexible orchestration mode across multiple Availability Zones. To satisfy the backend tier's requirements of a 99.95%99.95\% SLA and sub-millisecond network latency, you must place the virtual machines in an Availability Set within a single Availability Zone and associate them with a Proximity Placement Group. This guarantees the 99.95%99.95\% SLA while keeping the instances physically close enough to minimize network latency.

Step-by-Step Solution

1
Analyze the high availability requirements for the web frontend tier.
The frontend requires a 99.99%99.99\% uptime SLA and auto-scaling. In Azure, a 99.99%99.99\% SLA for virtual machines requires distributing them across multiple Availability Zones. Virtual Machine Scale Sets in Flexible orchestration mode support zonal distribution and scale-out capabilities.
Choosing a configuration that spans zones is necessary to meet the 99.99%99.99\% SLA, and scale sets are needed to handle automatic horizontal scaling.
2
Analyze the high availability and latency requirements for the backend tier.
The backend requires a 99.95%99.95\% SLA and sub-millisecond latency. An Availability Set provides a 99.95%99.95\% SLA. To guarantee sub-millisecond latency, the instances must be placed in a single Availability Zone and associated with a Proximity Placement Group to ensure they are physically colocated in the same data center.
Spreading backend virtual machines across different Availability Zones would increase latency beyond the sub-millisecond requirement, so keeping them in a single zone with a Proximity Placement Group is necessary.
3
Verify load balancing and SKU compatibility rules.
Standard Load Balancers are required for zonal deployments but they only support Standard SKU resources. Assigning Basic SKU public IPs to virtual machines behind a Standard Load Balancer is invalid.
This confirms that options attempting to mix Standard and Basic SKUs are invalid.

Key Concept

Designing compute high availability in Azure by balancing Availability Zones (for 99.99%99.99\% SLA), Availability Sets (for 99.95%99.95\% SLA), and Proximity Placement Groups (for low latency), while respecting resource SKU compatibility rules.
Rate this question

Topics

Design Compute High Availability

More questions from this topic