ibmcloud_image_uploader

TMOS Image IBM Public Cloud COS Uploader

IBM VPC Generation 2 Cloud supports the integration of custom appliance virtual machine images sourced from IBM Public Cloud Cloud Object Storage (COS) SQL URLs. To make patched TMOS Images available to IBM VPC Generation 2 Cloud customers, the TMOS images must first be made available as IBM COS objects.

This python script and packaged Docker container preforms the task of taking local patched TMOS images and creating IBM COS buckets and objects, as well as creating public COS SQL URL catalogs.

The script functionality is driven by environment variables. This makes it simple to dockerize. The following environment variable are supported:

ENV Variable	Default	Required	Description
TMOS_IMAGE_DIR	None	Yes	Directory to look for patched images
COS_API_KEY	None	Yes	The COS resource service API key
COS_RESOURCE_CRN	None	Yes	The COS resource CRN (id)
COS_IMAGE_LOCATION	us-south	Yes	A single or comma-delimited list of regions to upload images
COS_AUTH_ENDPOINT	https://iam.cloud.ibm.com/identity/token	No	Set the IBM Cloud auth resource (use for testing)
COS_BUCKET_PREFIX	f5	Yes	The unique IBM COS bucket name prefix for you image catalog
IMAGE_MATCH	^[a-zA-Z]	No	The regex match for the TMOS images to include in the catalog
UPDATE_IMAGES	false	No	Delete and update COS object if they exist
DELETE_ALL	false	No	Force delete all found COS objects and buckets

Using the Python Script

Install the python dependencies into your environment (virtualenv suggested)

pip install -r requirements.txt

Set environment variables in your shell (bash used in example)

export TMOS_IMAGE_DIR='/data/F5Downloads/Latest'
export COS_API_KEY='xxxxxxxxxxxxx_xxxxxxxxxx_xxxxxxxxxxxxxxxxxxx'
export COS_RESOURCE_CRN='crn:v1:bluemix:public:cloud-object-storage:global:a/xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx:xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx::'
export COS_IMAGE_LOCATION='eu-de,eu-gb,us-east,us-south'
export COS_BUCKET_PREFIX='prod'

Run the python script

./ibmcloud_cos_image_uploader.py

Using the Docker Container

Build the docker container

docker build -t ibmcloud_image_uploader:latest .

Set environment variables in your shell (bash used in example)

export TMOS_IMAGE_DIR='/data/F5Downloads/Latest'
export COS_API_KEY='xxxxxxxxxxxxx_xxxxxxxxxx_xxxxxxxxxxxxxxxxxxx'
export COS_RESOURCE_CRN='crn:v1:bluemix:public:cloud-object-storage:global:a/xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx:xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx::'
export COS_IMAGE_LOCATION='eu-de,eu-gb,us-east,us-south'
export COS_BUCKET_PREFIX='prod'

Run the container with the supplying environment variables.

Note: by default the docker container will make the TMOS_IMAGE_DIR to the /TMOSImages directory inside the container. The example uses an volume to mount the host directory to the /TMOSImages directory in the container.

docker run --rm -it -v $TMOS_IMAGE_DIR:/TMOSImages -e COS_API_KEY="$COS_API_KEY" -e COS_RESOURCE_CRN="$COS_RESOURCE_CRN" -e COS_IMAGE_LOCATION="$COS_IMAGE_LOCATION" -e COS_BUCKET_PREFIX="$COS_BUCKET_PREFIX" ibmcloud_image_uploader:latest

After completion this process will create a f5-image-catalog.json image catalog file in each specified region. The URL an f5-image-catalog.json file will be required to create a VPC Gen2 custom image step.

Create a VPC Gen2 custom image

The URL takes the format:

https://[COS_BUCKET_PREFIX]-[region].s3.[region].cloud-object-storage.appdomain.cloud/f5-image-catalog.json

As an example, for the region us-south with an COS_BUCKET_PREFIX of f5test the URL would look like:

https://f5test-us-south.s3.us-south.cloud-object-storage.appdomain.cloud/f5-image-catalog.json