Data Analytics - Specialty - AWS Free Practice Test

I found this certification to be quite difficult .[More than Solution Architect Professional]

As of now there are 18 questions I am adding more questions daily.

This is a sectional test for Data Analytics

The questions are with similar difficulty levels with the Speciality exam.
Please try to find the answer on your own
There some really updated service related questions
I have not kept any question questions from professional level but this can really help for that level
If you want a time-bound quiz with exact difficulty check for the practice test sets

For any help, you can reach out to me at [email protected]

100%

Page 1 of 1

1. Your company has newly acquired a company that has an in-house PostgreSQL with several tables all having an indexed Timestamp column. As a data architect, you need to make sure you integrate this with your existing Datawarehousing Solution which runs on Redshift as a staging layer. The business has given a requirement of having a data lag of a maximum of 3 hours. What is the most optimal way to achieve this?

Configure a trigger in the Postgre tables and write it to a Kinesis Stream and build an App that uses the KCL library to read the data from Kinesis stream and put it into Redshift.

Write one lambda function which extracts the table data into multiple S3 directories as per the tables. Create an external table of this S3 bucket, redshift would automatically scan all the files in the Bucket.

Write two lambda functions one of which extracts the files in different S3 buckets and compresses the CSVs as gzip. Enable an event trigger on the S3 bucket which calls the second lambda to use the COPY command to put the data in Redshift.

Write one lambda function which extracts the files in one S3 bucket into one compressed gzip and use the COPY command to put the data in Redshift.

2. There is an AWS Batch application that drops a file in excess of 50 GB into an S3 bucket and another batch application picks this file up and stores in Redshift. The size of the data is increasing daily and you can see the data loading is getting slow day by day. Which is the optimization technique that you should take to remove this slowness?

Split the data 2GB each and use Gzip compression. Use multiple COPY operation and run concurrently

Split the data 1GB each and use Gzip compression. Use multiple COPY operation and run concurrently

Split the data in 500 MB each and use a single COPY operation. and use Gzip compression

Load the data as randomly as possible

3. You are responding to a business requirement that has been floated by your client. Your client is having data in various data sources CSVs stored in S3, Logfiles files stored in S3, some structured data stored in Redshift. What business solution you should propose to your client which can give him an analytical Dashboard in the most cost-optimal way and in minimal time?

Use Glue jobs to load the data from S3 to Redshift and run Quicksight on top of Redshift

Use Quicksight directly on top of S3 and Redshift to come up with visualization dashboard

Write Python code on Jupyter notebook that runs on the EMR cluster which generates some graphs

Write a Lambda function which takes the data from S3 and Redshift and create some API gateway in front of the Lambda function. Create a javascript application hosted in S3 which renders the charts.

4. Choose the best option to create a real-time replica for a Redshift cluster.

Create an automated periodic snapshot in S3 in a different availability zone, and restore it when required.

Spinup two separate redshift cluster in different availability zone and write to each of it simultaneously

Configure the multi-AZ option in the redshift cluster.

Share the automated snapshot to a different availability and create a new cluster from it.

5. A large eCommerce company is using historical search data to create a recommendation engine. The data is fairly consistent apart from some predictable surge during sales. What is the best possible cost-effective solution to come up with the solution?

Publish click data into Amazon S3 using Kinesis Streams, and process the data in real time using Splunk on an EMR cluster with spot instances added as needed. Publish the modelâ€™s results to DynamoDB for producing recommendations in real-time.

Publish click data into Amazon S3 using Kinesis Firehose, and process the data nightly using Apache Spark and MLLib using reserved instances in an EMR cluster. Publish the modelâ€™s results to DynamoDB for producing recommendations in real-time

Publish click data into Amazon S3 using Kinesis Firehose, and process the data nightly using Apache Spark and MLLib using spot instances in an EMR cluster. Publish the modelâ€™s results to DynamoDB for producing recommendations in real-time.

Publish click data into Amazon Elasticsearch using Kinesis Firehose, and query the Elasticsearch data to produce recommendations in real-time

6. You need to analyze AWS cloud trial logs data which are stored in the S3 bucket, which is the best and most cost-effective way to analyze these logs using Athena.

Compress using gzip and rotate on a daily basis

Covert it into AVRO format and sort the data by date

Convert the data into Apache Parquet format, compressed with Snappy, stored in a directory structure of year=XXXX/month=XX/day=XX/

7. Due to some regulatory requirements, one of the health care clients of yours needs to make sure they back up their encrypted Dataware housing solution which is built in Redshift to a different region. What is the easiest way to implement this making sure the solution preserves the encryption at rest and in transit?

Configure cross-region snapshot of Redshift. Make sure the AWS KMS customer master key is used in the destination region.

Create a different Redshift cluster in different regions and configure a trusted connection between two redshift clusters and HSM. Configure realtime sync between the two redshift clusters.

Configure another Redshift cluster in a different region. Configure a Lambda function which is triggered daily by a cloudwatch event as cron which uses COPY command to put the data from one region to another.

Move data into an S3 bucket and use multiple Redshift clusters in different regions to query it.

8. Your client installed an elastic search node into their VPC. You have been employed as a cloud security consultant and your job is to secure the connection between any third-party service and elastic search node. What is the possible solution you can implement? CHOOSE TWO.

Create an SSH tunnel with port forwarding at 5601

For the web application use KMS between the Browser and the elastic search node.

Use the Ngnix reserve proxy at 443 port to use https for the browser-based application.

Set up a reserve proxy between browser and the elastic node.

Configure an IAM policy to allow your browser to access Kibana

9. You have got a ticketing application that puts the tickets dumps in the S3 bucket on a daily basis but in irregular time. The volume of the ticket dump is also not constant. The data dropped in S3 needs some transformation before being sent to the Elasticsearch cluster. What is the solution you will implement so that you can put these data in an Elastic Search cluster with minimum maintenance and maximum scalability?

Use Kinesis FIrehose to connect to S3 to directly send the data to Elastic Search

Use a Lambda function which is being triggered by an event in S3 object creation, that does the required transformation before sending the data into Elastic Search cluster.

Create a Glue job and call elastic search context in that to load the data from S3 to it. The transformation will be done in Glue job.

Implement SPARK streaming on an EMR cluster which is hosted in a reserved instance. Look into the S3 object creation and send the data to elastic search after transformation.

10. You are creating a sensor monitoring system that reads the sensor (which is installed in power plants ) data and logs them in Amazon S3. What the system you will build which will read this data in a continuous manner and report any anomaly or unpredictable behavior is detected in the data that will be most easily implementable.

Log the sensor data in S3 and write an Athena query to check for the sensor vitals. Write Lambda function which periodically is triggered from cloud watch events and executed the quires and calls SNS if the query breached any threshold.

Stream the data using a Kensis stram which sends the data in Apache Flink which is hosted in the EMR cluster. Use SNS to raise the alarm and Elastic Search as the monitoring dashboard.

Stream the data in Kinesis stream and use Kinesis data analytics and use RANDOM_CUT_FOREST to detect an anomaly, when detected use Lambda to raise SNS alarm.

Stream the data in Kinesis stream and build a custom application in EC2 which will validate the incoming data against a predefined model. Raise alarm if the data does not fit into a predictable model .

11. You have created a real-time IoT monitoring system. The sensors send the data in a Kenesis stream and then Lambda function reads the data from the streams and persist them into Dynamo DB table. You are using chard.js to create the dashboard for the system. It is observed that in peak time you are losing some data, which can be a possible cause. You are sure that you have allocated enough shards and read and write capacities. What can be possible causes for this situation? CHOOSE ALL THAT APPLY

There is some other Lambda function is your account which has some provisioned concurrency.

The lambda execution timeout value is too low.

The data that is getting process through the Lambda size is too low.

Not enough Kenesis shards are provisioned.

The code is not optimized

The memory allocated to the Lambda function is too low.

12. Because of GDPR compliance, you need to make sure the data your Redshift cluster is analyzing is residing in an S3 bucket that belongs to another AWS account and in a different Region. How can you configure your system to achieve this solution?

Create an IAM role in the Amazon S3 account and create another IAM role in the Amazon Redshift account that has permissions to assume the role of S3 account.

Create a VPC endpoint to create a managed connection between your Amazon Redshift cluster in a VPC and S3

Create an IAM role to allow COPY command and attach it to S3

Create a bucket policy along with the account number of the redshift. Attach the policy to the user agent of Redshift.

13. Being a data analysis manager of an e-commerce company you are looking at the shipment related data. These data are of multidimensional deals with day to shipment, cost of the product, shipment cost, category, destination, season e.t.c. You need to see how the cost of shipping is varying with the destination. What kind of Quicksight chart will be the best suite for this analysis?

Heat Map

Scatter plot

Pie chart

Pivot table

14. Being a data analysis manager of an e-commerce company you are looking at the shipment related data. These data are of multidimensional deals with day to shipment, cost of the product, shipment cost, category, destination, season e.t.c. You need to find out different trends and outliers from the data, what Quicksight visualization chart you will be using?

Heat Map

Pivot Table

Scatter Plot

Stacked horizontal bar chart

15. You are working as a solution architect in an online streaming music app company. The present architecture is using apache spark to make some digital filters on top of audio streams that are consumed using on-prem Apache Kafka and then they are stored in S3. You want to migrate the entire echo system to Amazon web service, which will be the best approach for this?

Produce data using spark streaming to Kinesis Data streams, read the data using Spark streaming from Kinesis streams, and write to S3.

Produce the data using Spark streaming to Kinesis Firehose, read using Spark streaming from Kinesis firehose and write to S3

By building your streaming data solution on Amazon EC2 and Amazon EMR and store the data in S3

Produce data using spark streaming to Kinesis Data streams, consume the data using Lambda from Kinesis streams to add the filters, and write to S3.

16. You are working as a solution architect in an online streaming music app company. The present architecture is using apache spark to make some digital filters on top of audio streams that are consumed using on-prem Apache Kafka and then they are stored in S3. You want to migrate the entire echo system to Amazon web service, which will be the best approach for this?

Produce data using spark streaming to Kinesis Data streams, read the data using Spark streaming from Kinesis streams, and write to S3.

Produce the data using Spark streaming to Kinesis Firehose, read using Spark streaming from Kinesis firehose and write to S3

By building your streaming data solution on Amazon EC2 and Amazon EMR and store the data in S3

Produce data using spark streaming to Kinesis Data Firehose, consume the data using Lambda from Kinesis streams to add the filters, and write to S3.

17. You are working as a solution architect in an online streaming music app company. The present architecture is using apache spark to make some digital filters on top of audio streams that are consumed using on-prem Apache Kafka and then they are stored in S3. You want to migrate the entire echo system to Amazon web service, which will be the best approach for this?

Produce data using spark streaming to Kinesis Data streams, read the data using Spark streaming from Kinesis streams, and write to S3.

Produce the data using Spark streaming to Kinesis Firehose, read using Spark streaming from Kinesis firehose and write to S3

By building your streaming data solution on Amazon EC2 and Amazon EMR and store the data in S3

Produce data using spark streaming to Kinesis Data Firehose, consume the data using Lambda from Kinesis streams to add the filters, and write to S3.

18. You have multiple sensors sending the data from a medical device that monitors some human vitals as part of some specific medical condition investigation. The data coming from the sensors takes approximately two minutes to get analyzed and then sent to some analysis dashboard. The number of sensors sending the data and the data packet load is not constant and there is no predictable load of the data to be analyzed. What will be the best solution for this kind of situation?

Define some IoT rules to send the data to Kenisis stream and use Lambda to consume them and do that transformation and then send it to S3. Use Quicksight to read the data from S3 and do the visualization dashboard.

Define some IoT rules to send the data to Kenisis firehose and use Lambda to consume the data from the firehose and do the transformation and then send it to S3. Use D3js to do charting and visualization.

enable Kinesis Data Firehose data transformation, Kinesis Data Firehose then invokes the specified Lambda function asynchronously with each buffered batch using the AWS Lambda synchronous invocation mode. The transformed data is sent from Lambda to Kinesis Data Firehose. Kinesis Data Firehose then sends it Quicksight for visualization.

Define an IoT rule action to send the data in SQS and use EC2 backed bu an auto-scaling group to consume the data as per the load.