You are asked to deploy a new application that has been designed to scale horizontally.
The business stakeholders have asked that the application be deployed in us-phoenix-1
Normal usage requires 2 OCPUs.
You expect to have few spikes during the week, that will require up to 4 OCPUs, and a major usage uptick at the end of each month that will require 8 OCPUs.
What is the most cost-effective approach to implement a highly available and scalable solution? (Choose the best answer.)
Click on the arrows to vote for the correct answer
A. B. C. D.A.
The best answer for this scenario is A. Create an instance pool with a VM.Standard2.2 shape instance configuration. Setup the autoscaling configuration to use 2 availability domains and have a minimum of 2 instances, to handle the weekly spikes, and a maximum of 4 instances.
Here is the detailed explanation of why this is the best answer:
Firstly, the requirement is to deploy a horizontally scalable application in the us-phoenix-1 region. Horizontal scaling is achieved by adding more instances of an application to handle increased traffic. Hence, we need to create an instance pool that can scale horizontally.
Secondly, the normal usage requires 2 OCPUs. This means that we need to create an instance pool with a minimum of 2 OCPUs.
Thirdly, there are spikes during the week that require up to 4 OCPUs. This means that we need to scale up the instance pool to handle the increased traffic during the spikes.
Fourthly, there is a major usage uptick at the end of each month that requires 8 OCPUs. This means that we need to scale up the instance pool further to handle the increased traffic during the monthly uptick.
Considering all the above requirements, we need to create an instance pool that can handle a minimum of 2 OCPUs, scale up to 4 OCPUs during the weekly spikes, and scale up to 8 OCPUs during the monthly uptick.
Option A suggests creating an instance pool with a VM.Standard2.2 shape instance configuration. This configuration provides 2 OCPUs, which is the minimum required for normal usage. The autoscaling configuration is set up to use 2 availability domains and have a minimum of 2 instances, to handle the weekly spikes, and a maximum of 4 instances, which will handle the increased traffic during the spikes.
This option is cost-effective because it provides the necessary resources only when required, and scales down when the traffic decreases. This means that we only pay for the resources that we use, and do not incur unnecessary costs.
Option B suggests creating an instance with 1 OCPU shape and using the Resize Instance action to scale up to a larger shape when more resources are needed. This option is not suitable because we need to scale horizontally, and creating multiple instances is necessary to handle increased traffic.
Option C suggests creating an instance with 1 OCPU shape and using a CLI script to clone it when more resources are needed. This option is not suitable because it requires manual intervention to scale up the application, which is not efficient when there is a sudden increase in traffic.
Option D suggests creating an instance pool with a VM.Standard2.1 shape instance configuration, with a minimum of 2 instances and a maximum of 8 instances. This option is not suitable because the VM.Standard2.1 shape only provides 1 OCPU, which is insufficient for normal usage. Scaling up to 8 instances also means that we are over-provisioning resources, which will incur unnecessary costs.