Create docker for Azure SDK and code repository mount

In this scenario Docker is used for running scripts that manages Azure resources. All files in local repository are mounted in local folder in docker. The code can be run locally in docker or directly from docker shell.

The advantages of this solution is to separate the environment and it can be run anywhere where docker is supported (Linux, Windows).

Steps:

Prepare Dockerfile file
Get Azure credentials and setup environment variables
Build a docker image
Run script in doker

1. Dockerfile preparation

Create repository folder named “azure”. Then in that folder create the file Dockerfile. Then update the file with all required settings in our example:

Python version

First decide what Python versions are available in docker official images. In this example Python 3.7 is used:

FROM python:3.7

Code repository

Define the current working directory and copy all code from local repository. Working repository is “azure” and all files will be copied from local folder to “azure” folder in docker:

WORKDIR /azure
COPY . /azure

Microsoft Azure SDK for Python

Azure SDK for Python provides set of packages that is used to manage Azure resources like networks, storage, virtual machines and so on. It is developed in Gighub and available here: https://github.com/Azure/azure-sdk-for-python

To install Azure SDK add in Dockerfile the following:

pip install azure

Define script that will be executed

Create the Python file list_regions.py in azure folder that will be doing some tasks on Azure account (list regions in this example):

from azure.common.credentials import ServicePrincipalCredentials
from azure.mgmt.resource import SubscriptionClient
import os


AZ_TENANT_ID = os.getenv('AZ_TENANT_ID')
AZ_CLIENT_ID = os.getenv('AZ_CLIENT_ID')
AZ_CLIENT_SECRET = os.getenv('AZ_CLIENT_SECRET')
AZ_SUBSCRIPTION = os.getenv('AZ_SUBSCRIPTION')


credentials = ServicePrincipalCredentials(
    client_id=AZ_CLIENT_ID,
    secret=AZ_CLIENT_SECRET,
    tenant=AZ_TENANT_ID
)

client = SubscriptionClient(credentials)

regions = client.subscriptions.list_locations(AZ_SUBSCRIPTION)
for location in regions:
    print(location.name)

In Dockerfile define what script will be executed during docker run:

CMD ["python"] ["./list_regions.py"]

Final Dockerfile

This is the final Dockerfile file:

FROM python:3.7
WORKDIR /azure
COPY . /azure
pip install azure
CMD ["python"] ["./list_regions.py"]

All files in azure folder:

$ ls -a
.		..		.idea		Dockerfile	list_regions.py
$

2. Azure Credentials and environment variables

For the Azure management there is needed Azure credentials like SUBSCRIPTION, TENANT, CLIENT and KEY. To get these items visit the Azure portal. More information how to get these credentials visit here.

Add these credentials to your shell profile like .bash_profile and export:

# Azure Credentials
export AZ_SUBSCRIPTION="XXXXXXXX-XXXX-XXXX-XXXX-XXXXXXXXXXXX"
export AZ_CLIENT_ID="f844c3e6-0502-4685-8acb-2477e01fd1d7"
export AZ_CLIENT_SECRET="Pyf8.zKBs54NT*3M?Fa?iq92Rx_lDL69"
export AZ_TENANT_ID="a50ba86d-ee84-42a6-9fb3-811fb481daeb"

Once running the docker these credentials will be copied to docker and exported.

3. Build docker image

To build a docker run docker command in the same directory where there is Dockerfile and python file:

$ docker build -t azure1 .

This command creates a docker container. The option “-t” adds a tag and this tag is a name of a docker image.

4. Run

Run command after docker start

The script list_regions.py will be executed on docker and exit. If there is needed to do some update in script the docker needs to be re-created. To run the script add environment options to pass Azure credential variables:

$ docker run -e AZ_SUBSCRIPTION=${AZ_SUBSCRIPTION} -e AZ_TENANT_ID=${AZ_TENANT_ID} -e AZ_CLIENT_ID=${AZ_CLIENT_ID} -e AZ_CLIENT_SECRET=${AZ_CLIENT_SECRET} azure1:latest
eastasia
southeastasia
centralus
eastus
eastus2
westus
northcentralus
southcentralus
northeurope
westeurope
japanwest
japaneast
brazilsouth
australiaeast
australiasoutheast
southindia
centralindia
westindia
canadacentral
canadaeast
uksouth
ukwest
westcentralus
westus2
koreacentral
koreasouth
francecentral
francesouth
australiacentral
australiacentral2
southafricanorth
southafricawest
$

Run command from docker and mount local repository

To have option to update the script and automatically upload to docker run this command:

$ docker run -it --entrypoint /bin/bash -v /Users/dragan/PYPROJECTS/azure:/azure -e AZ_SUBSCRIPTION=${AZ_SUBSCRIPTION} -e AZ_TENANT_ID=${AZ_TENANT_ID} -e AZ_CLIENT_ID=${AZ_CLIENT_ID} -e AZ_CLIENT_SECRET=${AZ_CLIENT_SECRET} azure1:latest
root@a5e41a5bae7f:/azure# 
root@a5e41a5bae7f:/azure# python list_regions.py 
eastasia
southeastasia
centralus
eastus
eastus2
westus
northcentralus
southcentralus
northeurope
westeurope
...
root@a5e41a5bae7f:/azure#

Options:

-i - Keep STDIN open even if not attached
-t - Allocate a pseudo-TTY
--entrypoint - Overwrite the default ENTRYPOINT of the image (there will be run /bin/bash)
-v - Bind mount a volume (mount local volume to volume in docker)
-e - Set environment variables

If you wan to make script modifications all changes will be instantly done in docker as well.

IT Blog