You are designing the compute infrastructure for a media streaming company's transcoder service. The service uses stateless virtual machines that dynamically scale based on CPU utilization. The design must satisfy the following requirements:
- Maintain an uptime SLA of 99.99% for the compute instances.
- Support both Spot VMs for cost-effective scaling and On-Demand VMs for baseline capacity within the same logical group.
- Allow the operations team to interact with and manage the individual underlying virtual machines using standard VM APIs.
Which two configurations should you include in the design?
- A Virtual Machine Scale Set configured in Flexible orchestration modeAnswer
- BAn Availability Set containing the virtual machines deployed within a single Availability Zone
- Instance distribution across multiple Availability Zones within the regionAnswer
- DA Virtual Machine Scale Set in Uniform orchestration mode with all instances restricted to a single Availability Zone
Answer
To meet the requirements, you must use a Virtual Machine Scale Set in Flexible orchestration mode and distribute the instances across multiple Availability Zones.
To achieve a 99.99% SLA, virtual machines must be distributed across two or more Availability Zones in the same region. Additionally, using a Virtual Machine Scale Set in Flexible orchestration mode allows mixing Spot VMs and On-Demand VMs within the same scale set while maintaining standard VM API management for individual instances.
Step-by-Step Solution
Key Concept
Designing High Availability compute architecture using Virtual Machine Scale Sets in Flexible orchestration mode distributed across Availability Zones.