We are all well-aware fo 𝗔𝗪𝗦 𝗔𝘃𝗮𝗶𝗹𝗮𝗯𝗶𝗹𝗶𝘁𝘆 𝘇𝗼𝗻𝗲𝘀 𝗮𝗻𝗱 𝗥𝗲𝗴𝗶𝗼𝗻𝘀, here are a few 𝘀𝗰𝗲𝗻𝗮𝗿𝗶𝗼 𝗯𝗮𝘀𝗲𝗱 𝗶𝗻𝘁𝗲𝗿𝘃𝗶𝗲𝘄 𝗾𝘂𝗲𝘀𝘁𝗶𝗼𝗻𝘀 for you. You will get tons of posts stating real-world interview questions, but most of them I found are very basic, here I will present some interesting questions, which help you cover the topics well.
Q: You have deployed your application within a single Availability Zone, and you now want to make the same highly available. What would you do?
A. I would increase the availability of the applications by at least another availability zone, with at least another instance of the same application in the same region. Subsequently, I would configure an Elastic Load Balancer that will distribute incoming traffic across these instances across the AZs. Finally, Auto Scaling should be configured to further enhance the increase in availability under increased loads or instance failure.
Q: One of the services in one Availability Zone has failed. How do you ensure that your application hasn't stopped working and no manual intervention is required?
A: I'll ensure that my application is spread across various AZs and ELB would redirect traffic to the healthy instances. In addition to that, I will have Auto Scaling configured so that new instances would be automatically launched in another AZ in case one zone fails.
Q: You have an RDS instance running in one AZ. How do you make it highly available and minimize downtime?
A: To make the RDS instance highly available and minimizing downtime, I'd make changes to the RDS instance to enable Multi-AZ deployment. This will create a standby replica of the instance in another AZ. In case the primary instance fails, RDS will failover automatically on the standby instance, hence preventing downtime.
Q: Ensure that your data is always accessible, even if a complete AWS region is unavailable. How do you approach this solution?
A: For multi-region architecture, I would always replicate data and applications across multiple regions. For S3, I can enable cross region replication. For databases, I can use either Amazon Aurora Global Databases or cross region replication in RDS. Aside from that, I would use Route 53 for DNS failover to redirect traffic to a healthy region in case of an outage in one region.
Q: You have an application with very high read traffic. How can you take advantage of Availability Zones to improve performance?
A: I would deploy the read replicas in different AZs to distribute the read traffic. This approach would improve latency and performance because read traffic could now be load-balanced across multiple AZs. As for databases like RDS, I can have the read replicas in multiple AZs and configure my application for forwarding the read requests to those replicas.
Q: Your web application needs to provide low latency to users in different geographies. How would you address this?
A: I would distribute the application across multiple regions on AWS, along with using Amazon Route 53-enabled Latency-based routing capabilities. Route 53 would then route end users to the nearest region offering low latency to them, thus minimizing the response time. Finally, I would cache the content close to the users through globally using CloudFront.
Q: A natural disaster affects the data center that hosts your primary AZ. Which AWS service would ensure your EC2 instances were still reachable?
A: If using more than one AZ in EC2 instance launch, with an ELB, the ELB will automatically route traffic to instances in good AZs and the application will be kept alive. Auto Scaling can automatically launch new instances of an AS group where at least one has failed, in healthy AZs.
Q: You notice that more geographically dispersed users hitting your application provide higher latency levels. What changes would you implement to minimize latency levels?
A: I would use Amazon CloudFront to distribute content at edge locations around the globe. I would be caching static and dynamic content for the reason that it would take less latency to the distinct geographically-dispersed locations now, therefore improving response times.
Q: You want to protect your S3 data in case there is a failure of an Availability Zone. What can you do?
A: In general, S3 stores data redundantly across multiple AZs within a region. If I wanted that extra layer of protection I could enable cross-region replication, by replicating that data into a different region if ever there was a full failure at the regional level.
Q: Your company wants to move to a new AWS region in order to serve a different market. How would you do it with little downtime?
A: First, I will replicate the application in the new region, deploying resources needed there (EC2, RDS, etc.). Then I would make use of data replication services like DMS (Database Migration Service) or cross-region S3 replication. Finally, I will update Route 53 to route traffic to the new region to do a failover swap-over with no downtime.
For more such stuffs, follow my page.
Top comments (0)