awssqssource

AWS SQS event source for Knative Eventing

This event source consumes messages from a AWS SQS queue and sends them as CloudEvents to an arbitrary event sink.

Contents

1. Prerequisites
2. Deployment to Kubernetes
   • As a AWSSQSSource object
   • As a ContainerSource object
   • As a Deployment object bound by a SinkBinding
3. Running locally
   • In the shell
   • In a Docker container

Prerequisites

• Register an AWS account
• Create an Access Key in your AWS IAM dashboard.
• Create a SQS queue.

Deployment to Kubernetes

The AWS SQS event source can be deployed to Kubernetes in different manners:

• As an AWSSQSSource object, to a cluster where the TriggerMesh AWS Sources Controller is running.
• As a Knative ContainerSource, to any cluster running Knative Eventing.

ℹ The sample manifests below reference AWS credentials (Access Key) from a Kubernetes Secret object called awscreds. This Secret can be generated with the following command:

$ kubectl -n <my_namespace> create secret generic awscreds \
  --from-literal=aws_access_key_id=<my_key_id> \
  --from-literal=aws_secret_access_key=<my_secret_key>

Alternatively, credentials can be used as literal strings instead of references to Kubernetes Secrets by replacing valueFrom attributes with value inside API objects' manifests.

As a AWSSQSSource object

Copy the sample manifest from config/samples/awssqssource.yaml and replace the pre-filled spec attributes with the values corresponding to your AWS SQS queue. Then, create that AWSSQSSource object in your Kubernetes cluster:

$ kubectl -n <my_namespace> create -f my-awssqssource.yaml

As a ContainerSource object

Copy the sample manifest from config/samples/awssqs-containersource.yaml and replace the pre-filled environment variables under env with the values corresponding to your AWS SQS queue. Then, create that ContainerSource object in your Kubernetes cluster:

$ kubectl -n <my_namespace> create -f my-awssqs-containersource.yaml

As a Deployment object bound by a SinkBinding

Copy the sample manifest from config/samples/awssqs-sinkbinding.yaml and replace the pre-filled environment variables under env with the values corresponding to your AWS SQS queue. Then, create the Deployment and SinkBinding objects in your Kubernetes cluster:

$ kubectl -n <my_namespace> create -f my-awssqs-sinkbinding.yaml

Running locally

Running the event source on your local machine can be convenient for development purposes.

In the shell

Ensure the following environment variables are exported to your current shell's environment:

export ARN=<arn_of_my_sqs_queue>
export AWS_ACCESS_KEY_ID=<my_key_id>
export AWS_SECRET_ACCESS_KEY=<my_secret_key>
export NAME=my-awssqssource
export NAMESPACE=default
export K_SINK=http://<url_of_event_sink>
export K_LOGGING_CONFIG=''
export K_METRICS_CONFIG='{"domain":"triggermesh.io/sources", "component":"awssqssource", "configMap":{}}'

Then, run the event source with:

$ go run ./cmd/awssqssource

In a Docker container

Using one of TriggerMesh's release images:

$ docker run --rm \
  -e ARN=<arn_of_my_sqs_queue> \
  -e AWS_ACCESS_KEY_ID=<my_key_id> \
  -e AWS_SECRET_ACCESS_KEY=<my_secret_key> \
  -e NAME=my-awssqssource \
  -e NAMESPACE=default \
  -e K_SINK=http://<url_of_event_sink> \
  -e K_LOGGING_CONFIG='' \
  -e K_METRICS_CONFIG='{"domain":"triggermesh.io/sources", "component":"awssqssource", "configMap":{}}' \
  gcr.io/triggermesh/awssqssource:latest