Skip to content
Sample code for dynamodbstreams to Lambda to SNS to SQS.
Branch: master
Clone or download
Latest commit 7648dde Nov 19, 2018
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
example-consumer
template add consumer Nov 19, 2018
.gitignore add consumer Nov 19, 2018
LICENSE Initial commit Nov 19, 2018
Makefile
README.md
go.mod
go.sum
main.go

README.md

Overview

This is an example of tracking a dynamodb table's changes using dynamodbstreams. Table item updates will trigger a lambda function that will then forward the event payload to an SNS topic. An example SQS subscriber is also provided that will subscribe to the SNS topic and print all event data.

A cloudformation template is provided to create the resources needed in this example. The example-consumer will create the SQS queue that subscribes to the SNS topic upon execution, if needed. The lambda code is in main.go.

How to run

You need to have the following required environment variables:

AWS_ACCT_ID={your-aws-account-id}
AWS_REGION={your-aws-region}
AWS_ACCESS_KEY_ID={key}
AWS_SECRET_ACCESS_KEY={secret}

You also need to create an S3 bucket named lambda-deploy-development for lambda deployment.

This example assumes that you will use an existing table. In my case, I used a test table named TESTSTREAM with dynamodbstreams enabled. View type is "New and old images". The streams' ARN is in the template (EventSourceArn) so you might want to update that part.

Then run the following command from the repo's root folder:

# Note that the GO111MODULE=on variable is enabled during the build.
$ make deploy

Your stack should be ready at this point, provided that the AWS credentials has the permission to create the resources. You can now run the example consumer for testing.

$ cd example-consumer/
$ go build -v
$ ./example-consumer

While the consumer is running, you can update your test table by inserting new items, modifying or removing existing items. The running consumer should print all event data.

Finally, to cleanup the stack, run the following command from the repo's root folder:

# This command will delete the cloudformation stack. It will not delete the SQS queue
# created by example-consumer so you need to delete that manually.
$ make remove
You can’t perform that action at this time.