SNS SQS STS - AWS Free Practice Test

Page 1 of 1

1. You are creating a static web application which is used for storing images uploaded by the user. The images are then run by some filters and after processing them, different filtered images are to be stored in some places, for the user to be downloaded. What will be the best architecture which can sustain certain spike in user activity and cater to the varying number of images uploaded with reliable performance?

Use EC2 m4x large to host the application and process the filtering and store the processed files in S3 for the users to download

Use S3 to host the application and configure SNS to distribute the load to multiple EC2 instances to handle the varying load and store the processed images in S3

Use S3 to host the application and store the messaged in SQS, have a load balancer backed by multiple EC2 instances to handle the varying load and store the processed images in respective EBS volumes

Use S3 to host the application and store the messaged in SQS, have a load balancer backed by multiple EC2 instances to handle the varying load and store the processed images in S3

2. You have created an application where users are uploading photos and they are stored in SQS and then an Ec2 instance is reading the messages and applying some filters on them and storing the filtered phots in S2 bucket. You have seen this entire process is taking more time to get completed, to solve this you have added another similar EC2 instance which reads the same message from the queue. But you can see the second Ec2 instance is not receiving any message to consume what can be the possible reason?

The second Ec2 instance is making a call before the visibility timeout is elapsed

The second Ec2 instance is not having access to read the message

The messages are getting deleted automatically as the first EC2 instance has processed it

The processing power of the second Ec2 instance is low so the first Ec2 instance is consuming all the messages

The first EC2 instance is processing the messages and then delete them

3. As part of the new business requirement, you need to make sure you store the clickstream ( which store the clicks made by the user in sequence and streams them in realtime ) data of all the users in real-time and analyze them. What is the best possible way you can do that?

Store the data in Dynamodb and then use ELK stack to analyze the data

Store the data in FIFO queue and then use multiple EC2 instances to read the data form the que and do the analysis

Store the clickstream data in S3 and then AWS lambda function which has an event on every object created in the S3 bucket to analyze the data

Use Kinesis stream to store the data and then use Kenesis worker to analyze the data

4. You have an enterprise application which is having SQS and thousands of messages are getting dropped in it and to consume it there are multiple EC2 instances that are processing these messages and deleting them. Your DevOps engineers are reporting of huge bills are getting incurred for this entire process. What are the possible solutions to reduce the bills without taking many hits of the performance? [Choose as many as apply]

send delete message request in the batch

Increase the visibility time out of the queue

Reduce the number of instances

Use SQS short polling

Use SQS Long polling