AWS IoT Rules have predefined rules for sending a message to an SQS Queue, but for some reason not to retrieve a message from a queue using an IoT rule (or does it? if you know how, leave me a comment). You can easily retrieve a message using a Lambda function using the AWS SDK apis though, and you can call a Lambda function using an IoT Rule, so let’s set that up.
To create an IoT Rule to call the Lambda function, which we’ll trigger with incoming MQTT messages on a topic called topic/checkForMessage:
Next, select an action from the long list of available actions, chose ‘Invoke a Lambda function’:
Select the Lambda function we want to call, in this case it’s one created earlier (it has to exist to show in the list, if not press ‘Create a new Resource’ to create one):
On the next summary screen press ‘Create Rule’ and you’re done:
To allow the IoT Rule to call the function, we need to grant the lambda:invokeFunction rule.
Per the docs, we can use the AWS CLI to add the permission like this:
aws lambda add-permission
To apply this to our function and rule, replace:
“function_name” : “LightsOnReceiveMessageFromQueue”
source-arn arn: aarn:aws:iot:full-arn-for-the-rule – see below
I’m not sure the AWS Console for IoT shows the ARN for IoT Rules anywhere in it’s pages, but you can easily list it with the AWS CLI, using:
$ aws iot list-topic-rules
Ok, plugging in my values into the aws cli statement I have a permission added.
This is it for the IoT Rule. To summarize, this allows us to:
- respond to incoming messages from an AWS IoT device publishing a message to an MQTT topic called topic/checkForMessages
- when a message arrives from the device on the topic, it triggers the IoT Rule we just created
- the rule invokes an AWS Lambda to interact with an AWS SQS Queue to pull a message from a queue.
I’ll share more details on the implementation of the Lambda to interact with the SQS queue and the implementation of the node.js app on a Raspberry Pi in upcoming posts. You’re probably wondering what this is that I’m building? Check back for my followup posts to find out!
This is the second post in a series on AWS and IoT, the first is here: