Single Node OpenShift is Red Hat’s solution for edge use cases when we want to run OpenShift workloads on constrained and remote locations, where, typically the physical space for systems decreases and our network may be restricted by No or intermittent connectivity. SNO is capable of offering control and worker node capabilities in a single server, reducing the footprint and providing users with a consistent experience across the sites where OpenShift is present. The minimum system requirements for SNO will just be 4 cores and 16 GB of RAM. It can be deployed using standard installation methods on bare metal hosts and certified hypervisors. Finally, it's important to understand that when you use Single Node OpenShift, the most significant trade-off is the absence of high availability.
The purpose of this node will be to use it for the preparation, training and modeling of the algorithm that will later be used for detection. In this demo we will use the Assisted Installer provided with the Hybrid Cloud Console to deploy Single Node OpenShift on real bare metal hardware, equipped with a GPU, in my case, an NVIDIA Tesla M60 GPU Card.
As a starting point, let’s navigate to the OpenShift Red Hat Hybrid Cloud Console. There, we will need to log in using our Red Hat account credentials. Once on the landing page, we are going to click on "Create Cluster" so we can start using the assisted installation. In this case, we are going to use a real bare metal host, so we will need to select the "Datacenter" tab. Here, under the Assisted Installer section, let’s click on "Create Cluster". This will guide us to the cluster configuration wizard page.
As you can see there are plenty of fields that we can configure in our cluster:
- Cluster name:
pokedex(insert any preferred name). - Base domain:
pemlab.rdu2.redhat.com(to match our environment domain). - OpenShift version:
OpenShift 4.13.13(latest version at the time I'm writing this). - CPU architecture:
x86_64(our bare metal host uses Intel architecture). - Check the "Install single node OpenShift (SNO)" box.
Finally, let's leave the rest of the parameters as default and click on the blue “Next” button until you reach the Host Discovery section.
Note We're omitting the Operators section since the purpose of this demonstration is to illustrate the complete installation process for the different operators needed to build this demo. However, feel free to check the Logical Volume Manager Storage box in case you want to speed up the process.
In the Host discovery page, we can see that the installer is waiting for a host. Firstly we need to import it by clicking on "Add host". This will show us a pop-up window. The purpose of this step is to generate and install a discovery ISO in our node so it can be detected from the Installer. Here, we can see different options. In our case we want to boot using virtual media, so we need to select "Minimal image file". Note that if we are going to use this option, our node will require internet access to provision the image.
The next step is to provide our "SSH public key" so we can connect to the host machine. As you probably know, we can get it using our Terminal by running the following command. Then, copy and paste the result into the Public key section:
cat ~/.ssh/id_rsa.pub
At this point, we are good to "Generate the Discovery ISO". Instantly, we will be ready to download the generated image. Click on "Download Discovery ISO" to store it in your laptop. This will start the download process. The time it takes will depend on your internet connection.
Now, we will switch to our bare metal virtual console. Here we need to configure our node to boot using our discovery iso. Depending on the host, this procedure could be slightly different, but overall, the steps would be pretty similar to these ones:
In our case, we will click on "Connect Virtual Media". There, we will be able to "Choose the File" with the discovery iso we just downloaded. Now we are good to "Map the device". Also we will need to configure our host to boot using the virtual media. We will click on the "Boot" menu and select "Virtual CD/DVD/ISO" from the list.
Finally, all we need to do is reboot the system. Click on "Power Off System" and then "Power On the System" again. With this, the Discovery installation will begin.
Once finished and back to the OpenShift Red Hat Hybrid Cloud Console, we will be able to see our machine automatically recognised and listed in the Host Inventory section. Make sure to check the “Status” displayed is Ready:
At this point we can directly jump to the end of the process. Click "Next" to navigate to the Storage section, where we can list all the disks present in our node. Click "Next" again, and that will guide you to the Networking section. The environment where the node we are using has already configured all the networking infrastructure, so all we have to do is select the "Machine network" IP range we want to use. To configure your environment network, please refer to the OpenShift networking documentation.
Now we are ready to proceed with the SNO installation. Click "Next" to see a summary with your cluster details from the Review and create section. If everything is looking good for you, press the "Install cluster" button.
The installer will complete automatically all the steps needed to have our node up and running. We can track this process through this progress bar. It should take some time to finish the installation:
Once the installation has finished, we can access the OpenShift node using the credentials provided under the Installation bar. You should be able to spot the URL to the Web console, the kubeadmin user and the password that we will use to authenticate. Click on "Launch OpenShift Console". A new pop-up window will appear. To access the web console we need beforehand to configure our machine to resolve its hostname. We can do it by either adding the record to the DNS server or locally in our /etc/hosts configuration file. Copy and paste the entries there and click again "Launch OpenShift Console".
You will probably see a warning message complaining about the security of the page, but we can safely ignore it. Once on the logging page, we will use the credentials provided after the SNO installation. By this, finally we successfully logged into the Single Node OpenShift Web Console!
For this demo, we will also use the Command Line to run most of the commands. To connect to our SNO, once on the the Web Console, click on our current user "kube:admin" in the upper right corner and select "Copy logging command". This will open a new tab in our web browser. If we click on "Display token", we can copy the oc login command shown and paste it into our terminal:
Now run the following command to ensure everything is settled correctly:
oc get node
The output should look similar to this:
NAME STATUS ROLES AGE VERSION
r740.pemlab.rdu2.redhat.com Ready control-plane,master,worker 14d v1.26.7+c7ee51f
Instead of using the kubeadmin user, we can set up new users and give them the needed permissions. For this purpose we are going to configure an htpasswd identity provider. First of all, using the terminal, we will need to create the file that will contain the users and their passwords encrypted:
htpasswd -c -B -b ~/htpasswd user1 <password>
In case we need to add other users, the command will be quite similar to the previous one:
htpasswd -B -b ~/htpasswd user2 <password>
Now we can define the secret containing our users credentials:
oc create secret generic htpass-secret --from-file=htpasswd=~/htpasswd -n openshift-config
Finally, modify the oauth to add the identity provider:
oc edit oauth cluster
Add these lines:
...
spec:
identityProviders:
- htpasswd:
fileData:
name: htpass-secret
mappingMethod: claim
name: Pokedex
type: HTPasswd
In case you want to perform admin activities in the cluster with your user, you'll need to assign it admin permissions. Run this command for this purpose:
oc adm policy add-cluster-role-to-user cluster-admin user1
Now log out from the Web Console and log in again with your new user. And that's all!
In this first episode we have easily installed OpenShift in a single baremetal node, checked that everything is working correctly and enabled an Identity Provider to allow different users to access the node. In the next chapter we will focus on the Infrastructure configuration needed before starting working on the AI model.