Test Workload

Test the Workload

In this section we’ll make sure that the workload is running properly, and that data is replicating to the backup region. We use a python script to simulate a stream of incoming tweets.

Open a Session Manager connection to the EC2 instance sample producer in the primary region. You can find the instance ID and the Global Accelerator DNS endpoint in the CloudFormation output.

  • Open the EC2 console and navigate to Instances.
  • Select the producer instance and click Connect.
  • Choose Session Manager.

Session Manager

Once connected, run:

sudo su - ec2-user
python3 tweetmaker.py --endpoint <Global Accelerator endpoint>

Now, navigate to Kinesis Analytics in the AWS Console. Click on the radio button for the application called <stack name>-KinesisAnalyticsApplication and select Run.


After a few minutes you should be able to preview the table in Athena and see some output. The Glue database is called backuprestoredb and the raw tweets are in a table called rawdata.

To produce the nightly compacted files, run the Glue job called CompactNightly. Then you can preview the table compacteddata.

To verify data replication, you can look for files in the bucket in the backup region.

You can also verify that data is getting populated in the DynamoDB table - processed_tweets. We have set up an AWS backup plan where we would be backing up this table every hour.