A company would like to migrate its current on-premises workloads to the public cloud. The current platform requires at least 80 instances running at all times to work properly. The company wants the workload to be highly available, even if the cloud provider loses one region due to a catastrophe, and the costs to be kept to a minimum. Which of the following strategies should the company implement?
The company should create /26 subnets in three regions and run 40 instances on each one. This configuration ensures that at least 80 instances are running at all times even if one entire region goes down, thereby providing high availability. Distributing the instances across three regions also provides redundancy and fault tolerance, while keeping costs lower compared to running 80 instances in each of two regions, which would result in a total of 160 instances.
I like A because of the simplicity. C would be my alternate. As to cost I'm not sure if it's cheaper to run the extra instances (A) or if it's cheaper to run in an extra region (C).
You cant run 80 on a /26
First, the question states that in order to operate correctly, the platform needs 80 instances at ALL times in order to operate correctly. This rules out "B" (if one of your 2 regions goes down, you're left with only 40). You could have 40 instances in 3 regions and if 1 goes down you still have your 80, but it states to keep costs down so you're not going to have 3 regions. This rules out "C" and "D". In addition, the /26 unequivocally rules out "D". Answer "A" fits the bill perfectly with regard to the subnet requirement, the low cost requirement and the 80 instances requirement should one region go down.
To migrate the current on-premises workloads to the public cloud, with high availability and cost minimization, the company should implement the strategy of "Create /26 subnets in three regions and run 40 instances on each one" (C). Running 80 instances on each region would be more expensive and less efficient in terms of resource utilization. Creating two regions with /25 subnets or /26 subnets with 40 instances each would not provide sufficient redundancy in case of a catastrophe in one region. By creating /26 subnets in three regions and running 40 instances on each one, the company would have a total of 120 instances running, which satisfies the requirement of at least 80 instances running at all times. This strategy also provides high availability since the workload can be replicated in multiple regions, ensuring that it continues to function even if one region is lost due to a catastrophe.
I think the answer is C. You'd be running a total of 120 instances vs 160 instances and you'd have high availability because if one subnet goes down you still have 80 instances.
40 in 3 regions (40x3=120) vs 80 in 2 regions (80x2=160) -> means that it is cheaper to host the 40 in 3 regions, while also being able to lose an entire region and still have the required 80 instances.
"at all times" Implies that less than 80 will not suffice. Thus we need at least 2 sets of 80 for HA
The question said : 1.at least 80 instances running at all times to work 2.wants the workload to be highly available, even if the cloud provider loses one region due to a catastrophe 3.costs to be kept to a minimum The choice of /26 subnets allows for up to 64 IP addresses, which is sufficient for running 40 instances in each region - Respect al 3 requiraments By running 40 instances in each region, the company can achieve the desired level of availability while keeping costs lower - So, we have cost optimization ! By distributing the workload across three regions, the company ensures redundancy and fault tolerance. They want AT LEAST 80 instances.
VVV4WIN explains it best.
C is the correct answer
I think A is correct. Its not B or D as the maths just dont work. The question specifically states they want redundancy if one region goes down, so A covers that. Creating 3 regions cost MORE than 2 regions so thats my logic for choosing A.