Skip to content

Latest commit

 

History

History
100 lines (67 loc) · 3.14 KB

README.md

File metadata and controls

100 lines (67 loc) · 3.14 KB
Raw

Terraform, triggers and serverless containers

Setup

Create a file secrets.auto.tfvars file holding your project ID, access key and secret key:

project_id = "your-project-id"
access_key = "your-access-key"
secret_key = "your-secret-key"

The API key you use only needs the following permission sets:

  • ContainerRegistryFullAccess
  • ContainersFullAccess
  • MessagingAndQueuingFullAccess

Optional: you can also define in this file the language used for the container. By default, when not specified, it will use the python container.

container_language = "go"

The values for container_language can be either python or go (see inside docker/ folder).

Deploy

The deployment will do the following:

  1. Create a Scaleway registry namespace
  2. Build and deploy a container image with a Python/Go HTTP server
  3. Deploy a public and private Serverless Container using the built image
  4. Create Scaleway MnQ SQS queues and NATS subjects
  5. Configure triggers from these SQS queues and NATS subjects to each container
  6. Print the endpoints of each SQS queue, NATS subject and container

To run the deployment:

terraform init

terraform apply

Running

You can test your triggers by sending messages to the SQS queues and NATS subjects configured with Terraform.

Below there is an example showing how to do this using a Python script in the tests/ folder of this repo.

Setup

First you need to export your queue credentials and URLs:

export AWS_ACCESS_KEY_ID=$(terraform output -raw sqs_admin_access_key)
export AWS_SECRET_ACCESS_KEY=$(terraform output -raw sqs_admin_secret_key)
export PUBLIC_QUEUE_URL=$(terraform output -raw public_queue)
export PRIVATE_QUEUE_URL=$(terraform output -raw private_queue)
export PUBLIC_SUBJECT=$(terraform output -raw public_subject)
export PRIVATE_SUBJECT=$(terraform output -raw private_subject)
export NATS_CREDS_FILE=$(terraform output -raw nats_creds_file)

You can then set up a Python environment in the tests directory:

cd tests
python3 -m venv venv
source venv/bin/activate
pip3 install -r requirements.txt

Sending messages

You can now run the script to send messages to both queues with:

python3 send_messages.py

Viewing function logs

In your Cockpit, you can access the logs from your queues and functions.

Navigate from your Cockpit to Grafana, and find the Serverless Containers Logs dashboard. There you should see the "Hello World!" messages for the functions created by this example.

To get direct deep links to the function logs in Cockpit, you can also run the following from the root of the project:

echo "Go to $(terraform output --raw cockpit_logs_public_container) to see public container logs in cockpit"
echo "Go to $(terraform output --raw cockpit_logs_private_container) to see private container logs in cockpit"

Finally, you can modify docker/python/server.py (or docker/go/server.go) as you wish and run terraform apply to redeploy the new version of the container.

Cleanup

Run terraform destroy to remove all resources created by this example.