Elastic Compute Cloud
You are designing a system which needs at minimum, 8 m4.large instances operating to service traffic.While designing a system for high availability in the us-east-1 region having 6 Availability Zones, your company needs to be able to handle the death of a full availability zone. How should you distribute the servers to save as much cost as possible, assuming all of the EC2 nodes are properly linked to an ELB? Your VPC account can utilize us-east-1’s AZs a through f, inclusive.
A. 3 servers in each of AZs a through d, inclusive.
B. 8 servers in each of AZs a and b.
C. 2 servers in each of AZs a through e, inclusive.
D. 4 servers in each of AZs a through c, inclusive.
C. 2 servers in each of AZs a through e, inclusive.
The best way is to distribute the instances across multiple AZs to get the best performance and to avoid a disaster scenario.
With this solution, you will always have a minimum of more than 8 servers even if one AZ were to go down.
Even though options A and D are also valid, the best solution for distribution is Option C.
For more information on High Availability and Fault tolerance, please refer to the below link:
https://media.amazonwebservices.com/architecturecenter/AWS_ac_ra_ftha_04.pdf
Note:
In option A, we need to distribute 3 servers in each AZ’s.
so, A=3, B=3, C=3, D=3. Total servers used=12.
In option B, we will not get high availability.
In option C, we need to distribute 2 servers in each AZ’s.
so, A=2, B=2, C=2, D=2, E=2. Total servers used=10.
In option D, we need to distribute 4 servers in each AZ’s.
so, A=4, B=4, C=4. Total servers used=12.
In the question, it’s clearly mentioned that “the company needs to be handle death of full AZ and save as much cost as possible.” In option C we are using less number of servers i.e 10 servers distributed in more AZ’s.
