Domain 1 - Design Resilient Architectures Flashcards
A company is designing a human genome app using multiple EC2 instances. The high performance computing (HPC) app requires low latency and high performance communication between the instances. Which solution provides the LOWEST latency between the instances?
A: Launch EC2 instances in a cluster placement group
B: Launch EC2 instances in a spread placement group
C: Launch EC2 instances in an auto-scaling group across multiple regions.
D: Launch EC2 instances in an auto-scaling group across multiple AZs within a single region
A: Correct. Cluster is designed for low latency
B: Wrong. Spread is designed for availability, not latency
C: Wrong. Impossible as auto-scaling groups can’t span regions
D: Wrong. Possible, but would be higher latency than A.
Designing for high availability means:
A: Designing for minimal downtime
B: Designing for zero downtime
A
Designing for fault tolerance means:
A: Designing for minimal downtime
B: Designing for zero downtime
B
What do the following 2 Disaster Recovery Objectives stand for?
RTO
RPO
RTO - Recovery Time Objective
RPO - Recovery Point Objective
What does the Disaster Recovery Strategy of ACTIVE/PASSIVE look at?
Backup and restore, pilot light, warm standby
What does the Disaster Recovery Strategy of ACTIVE/ACTIVE look at?
Multi-site active/active deployment style
A solutions architect must create a disaster recovery (DR) solution for a company’s business-critical applications. The DR site must reside in a different region than the primary site. The solution requires a recovery point objective (RPO) in seconds and a Recovery Time Objective (RTO) in minutes. The solution also requires the deployment of a completely functional, but scaled down version of the app. Which DR strat will meet these requirements?
A: Multi-site active/active
B: Backup and restore
C: Pilot Light
D: Warm Standby
A: Wrong. Although this meets the different region, RTO and RPO requirements, it doesn’t have a scaled down version, it has a full version
B: Wrong. Doesn’t have a scaled down version
C: Wrong. Doesn’t have a scaled down version
D: Correct.
Decoupling
refers to components remaining autonomous and unaware of each other as they complete their work as part of a larger system
Decoupling Techniques - Synchronous
Involves components that must always be available for proper function.
For example, an ELB distributing traffic to multiple EC2 instances across multiple AZs. The EC2s don’t need to know about each other, and as long as the EC2s and ELB are all available, it will function properly.
Decoupling Techniques - Asynchronous
Involves communication between components through durable components.
For example EC2s sending messages to each other, and using SQS to queue these messages. If one of the instances goes offline, the queue would hold the message until it’s ready to be received
A company is migrating it’s on-prem app to AWS and refactoring its design. The design will consist of front-end EC2s that receive requests, back-end EC2s that process the requests, and a message queuing service to address decoupling the app. The solutions architect has been informed that a key aspect of the app is that requests are processed in the order in which they are received. Which AWS service should the SA use to decouple the app?
A: SQS Standard Queue
B: SNS
C: SQS FIFO Queue
D: Amazon Kinesis
A: Wrong. Doesn’t keep the order
B: Wrong. Does not guarantee message delivery, order, or have any queue, and message could get lost
C: Correct.
D: Kinesis is for collecting, processing and analyzing data. Not a decoupling component.
An organization is planning to move from on-prem to AWS and span multiple AZs. You’ve been tasked with the plan to transfer the home directories and other shared, network-attached storage from the data center to AWS. Your migration design should support connections from multiple EC2s running Linux OS to this common, shared storage platform. What storage option best fits your design?
A: Transfer to S3 and access from the EC2 instances
B: Transfer to EC2 instance store attached to the EC2s
C: Transfer the files to EFS and mount that to EC2s
D: Transfer to one EBS and mount to EC2s
A: Wrong. S3 doesn’t provide file systems like home directories
B: Wrong. Can only be attached to a single instance
C: Correct. EFS is designed for parallel shared files
D: Wrong. Can attached to multiple instances, but not if they are in different AZs
