For our walkthrough, we will be using the Jump host as our host to download the content bundle since it has access to the internet. This isn't required and these steps could be done from your laptop or any other internet connected host. We are doing using the jump host in this case because it has a fast connection to the internet and is avaliable as part of our simulated disconnected environment in the nxext step.
Start by connecting to the Jump host via SSH. You can grab the public IP of the Jump host from the outputs captured during the environment prep stage or you can use the command below as a convenience.
export JUMP_HOST_PUBLIC_IP=$(aws cloudformation describe-stacks --stack-name ocp4-disconnected --query 'Stacks[0].Outputs[?OutputKey==`JumpInstancePublicIp`].OutputValue' --output text)
ssh-add ~/.ssh/ocp4-disconnected
ssh ec2-user@${JUMP_HOST_PUBLIC_IP}In order to pull the container images required to install OpenShift, you will need a pull secret that allows you to authenticate to the Red Hat registry.
If you do not already have a Red Hat account, register for a FREE Red Hat Developer account that will give you access to Red Hat software, including the ability to get the pull secret needed for this tool.
You can find your pull secret on the Red Hat OpenShift Console
Keep this page handy as you will be prompted for the value of your pull secret when you get ready to run the tool.
We will now run this tool the download and package the content we'll need in order to perform an install of OpenShift in a disconnected environment in the next step This .tool accepts a few parameters and will prompt for any that are not passed in. For this walkthrough, we will use the tool in that method and have it prompt us for the required input.
Warning
If you choose to specify parameters via the CLI instead of being prompted, it
is NOT recommended to override --output-dir as it is already
configured as part of the container to output to /mnt/data. You should map
a volume in your podman run (as seen below) to /mnt/data to ensure your
data persists.
The information that we need to give the tool is:
- The OpenShift version to download content for, and
- Our pull secret from above. The pull secret only needsto be specified on the first run as it will be saved and used for future runs.
If you do not specify the OpenShift version when prompted, it will default to
the latest stable version. You can also specify the version in the format of
x.y (e.g. 4.12) and it will find the latest stable of that version. If you
want a specific z-stream release, you can also specify the version as x.y.z
(e.g. 4.12.30).
podman pull ghcr.io/redhatgov/ocp4-disconnected:latest
podman run -it --rm --name ocp4-disconnected --platform linux/amd64 --volume /mnt/ocp4_data:/mnt/data:z ghcr.io/redhatgov/ocp4-disconnected:latestAfter the tool finishes, all of the content we need is in a single tar file at
/mnt/ocp4_data/ocp4_bundle.tar.
If you are doing this in a different environment than this simulated
environment and have passed a different path as your --volume mapping in the
command above, the ocp4_bundle.tar will be in whichever directory you
specified.
Now that we have our ocp4_bundle.tar, we are ready to transfer to the
disconnected environment. As mentioned earlier, in a real environment we would
follow the approved process for transferring content to our disconnected
environment, but for this walkthrough we will be using an S3 bucket. You can
grab the name of the S3 bucket from the outputs captured during the environment
prep stage or you can use the command below as a convenience.
export S3_TRANSFER_BUCKET=$(aws cloudformation describe-stacks --stack-name ocp4-disconnected --query 'Stacks[0].Outputs[?OutputKey==`S3TransferBucket`].OutputValue' --output text)
aws s3 cp /mnt/ocp4_data/ocp4_bundle.tar s3://${S3_TRANSFER_BUCKET}At this point, we are done with the work we need to do in the connected environment. In the next steps, we will be working in the disconnected environment.