How Autoscaling and Availability Zones Handle Variable Demand

What Fixed High-Tier Provisioning Is

Fixed high-tier provisioning involves deploying a single, oversized compute instance with sufficient CPU, memory, and I/O capacity to handle the highest projected traffic spikes that the organization might ever experience.

Why It Doesn't Meet the Requirement

This approach fails to meet both cost efficiency and reliability goals:

• Excessive operational costs: The organization pays continuous high pricing for maximum compute capacity 24/7, even during off-peak weekday periods when only a fraction of those resources is utilized.
• Single point of failure: Relying on a single instance creates a severe availability risk. If the host hardware, operating system, or datacenter rack experiences an issue or requires maintenance, the entire application becomes unavailable.

Key Characteristics and Limitations

• No automatic elasticity: Cannot adapt compute footprint dynamically in response to actual user traffic.
• Substantial resource waste: Leads to low average utilization and unnecessary cloud expenditure.
• Zero fault isolation: Fails to provide redundancy across separate fault domains or availability zones.

This Answer Would Be Correct If

The workload were a non-critical, legacy monolithic application that cannot be clustered or scaled across multiple nodes and operates under predictable, continuous maximum compute utilization.

What Manual Vertical Scaling Is

Vertical scaling (scaling up) involves increasing the physical capacity of an existing virtual machine by adding more CPU cores, RAM, or storage to the single running instance through manual configuration changes.

Why It Doesn't Meet the Requirement

• Introduces downtime: Resizing an existing virtual machine typically requires restarting the instance, which causes service interruption rather than minimizing downtime.
• Lacks responsiveness: Manual scaling depends on human administrators noticing traffic changes and executing changes, making it ineffective against sudden, unpredictable spikes.
• Cost inefficiency: Resources cannot be rapidly deprovisioned in real time as traffic fluctuates throughout the day.

Key Characteristics and Limitations

• Architectural ceilings: Physical hardware limits cap the maximum size to which a single machine can scale.
• Downtime during resize: Reallocating hardware to resize a virtual machine incurs system restarts.
• Operational burden: Requires ongoing 24/7 administrative oversight to react to changing traffic levels.

This Answer Would Be Correct If

The application were a database running on a single server that requires occasional scheduled upgrades during maintenance windows rather than automated adjustments during dynamic load surges.

What Cold Disaster Recovery with Manual Failover Is

Cold disaster recovery is a business continuity strategy where a secondary environment is configured in an alternate geographic region, but the secondary compute instances remain turned off or unprovisioned until a catastrophic disaster occurs.

Why It Doesn't Meet the Requirement

• No dynamic elasticity: Cold disaster recovery does not address regular daily traffic fluctuations or scale compute capacity up and down during promotional events.
• Significant recovery time: Activating a cold site and initiating a manual failover takes hours or days, which fails the requirement to minimize downtime during instance failures.
• Targeted at catastrophe, not availability: Disaster recovery protects against region-wide outages rather than ensuring high availability for daily operations.

Key Characteristics and Limitations

• High Recovery Time Objective (RTO): Requires manual spin-up, data restoration, and DNS redirection.
• Static capacity: Does not provide automated responsiveness to application load shifts.
• No load balancing: Primary traffic is not distributed across multiple active nodes during normal operations.

This Answer Would Be Correct If

The primary goal were to provide a cost-effective emergency backup plan for rare, catastrophic regional disasters with an acceptable recovery window of several hours.