Python API integrated with Azure Service Bus Queue

Requirements

• Platform: x86-64, Linux/WSL
• Programming Language: Python 3
• Cloud Account: Azure
• Resource provisioning: Azure CLI

Allocate resources

The shell script provision_resources creates Azure resources by calling the Azure CLI, which in turn makes HTTP calls to the resource-specific API on Azure.

It will create the following hierarchy of resources:

graph TD
    A[Subscription]
    A --> B[Resource Group]
    B --> C[Service Bus Namespace]
    C --> E[Service Bus Queue]

    A -->|Contains| B
    B -->|Contains| C
    C -->|Contains| E

Loading

For this script to work it is necessary to have a configuration file named infra_config.env in your infra directory. It contains sensitive information such as tenant and subscription id as well as information used to reference resources. The file has been added to our .gitignore so that you don't accidentally commit it.

Structure of 'infra/infra_config.env'

TENANT_ID={TO_BE_SET_BY_YOU_MY_FRIEND}
SUBSCRIPTION_ID={TO_BE_SET_BY_YOU_MY_FRIEND}
LOCATION=northeurope
RESOURCE_GROUP_NAME=hvalfangstresourcegroup
SERVICE_BUS_NAMESPACE=hvalfangstservicebusnamespace
QUEUE_NAME=hvalfangstqueue

Deallocate resources

The shell script delete_resources deletes our Azure service bus queue, namespace and resource group.

Client config

Once you have provisioned the Azure resources, there is one last configuration file which has to be created. The file service_bus_config.env is expected to exist in the client directory. It is used by our service bus configuration class, which reads values from aforementioned files and maps them to it accordingly. It contains the means to reach our Service Bus Namespace.

Structure of 'client/service_bus_config.env'

CONNECTION_STRING={TO_BE_SET_BY_YOU_MY_FRIEND}
QUEUE_NAME=hvalfangstqueue

You can get hold of the namespace connection string either by peeking in your terminal logs as it was actually echoed (yes, I know) as part of our infrastructure provisioning OR you can go for ClickOps approach and copy it from the Shared access policies under your Service Bus Namespace on Azure - as the screenshot below illustrates.

Running API

The shell script run_client creates a new virtual environment based on our requirements file and serves our API on port 8000 using uvicorrn.

A Postman Collection has been provided, which contains example requests for interacting with our queue.