AZ-104 Question 153
Select 2You are deploying a new Azure Virtual Machine Scale Set to support a CPU-intensive back-end service for an e-commerce application that often experiences rapid spikes in traffic. You want to ensure the service can automatically scale out during peak usage and scale in when demand decreases. Which two actions must you take to configure the scale set for dynamic auto-scaling?
- A
Set the orchestration mode of the scale set to Flexible
- B
Define autoscale rules with CPU threshold-based scaling policies
- C
Attach a public IP address to each VM instance
- D
Enable Boot Diagnostics for every VM instance
- E
Configure the default capacity to zero and set an upper limit for maximum capacity
Show answer and explanation
Correct answers: B, E
Explanation
To enable automatic scaling based on resource usage, you must create autoscale rules (such as CPU-based thresholds) and configure the minimum, default, and maximum instance counts. This combination ensures the scale set can scale out rapidly during spikes and scale in when demand subsides.
- A. Incorrect.
While Flexible orchestration offers more granular control over VM instances, it is not strictly required to enable dynamic auto-scaling based on service demand.
- B. Correct.
Defining autoscale rules with CPU thresholds enables the scale set to automatically add or remove instances when CPU usage meets defined criteria.
- C. Incorrect.
Attaching a public IP to each VM instance is not necessarily related to auto-scaling. You can configure load balancers and NAT rules for external connectivity without assigning public IPs individually.
- D. Incorrect.
Enabling Boot Diagnostics helps in troubleshooting, but it doesn't impact scaling rules or capacity adjustments.
- E. Correct.
Setting the scale set's default capacity to zero and defining a maximum limit allows the service to automatically scale out from zero instances when demand is high and scale back in when demand is low.