How to Get Started with Monitoring on OpenShift Using Randoli | Agent Installation Guide

If you're running production workloads in Open Shift, this particular video is for you. Hi, Kunal here, devil engineer at Randoli. In this particular demo, I'm quickly going to walk you through how you can install the Randoli agent in your Open Shift cluster so you can get started with monitoring your infrastructure and your applications running on Open Shift using Randoli. Let's get started. To keep this demo short, I've already gone ahead and connected to an Open Shift cluster. If I do cubectl get nodes, you can see we have a very basic open shift cluster where we have three control plane nodes and three worker nodes which are running. The next step would be to adding a new open shift cluster in the randol UI and getting the relevant credentials. So you can log to the randol dashboard and head over to the clusters page. Here we can click on connect cluster. We can give the cluster a relevant name. In this particular case I can say open shift demo. From the Kubernetes distribution dropdown, we can select open shift. You have the option to give any relevant tags if you wish to, but in this particular case, let's leave it as empty. We'll click on create cluster. Now once the cluster is created on the UI, we'll follow the direct instructions which are given on screen. So we can download the credentials file and then we can create a new name space called as randomly agents and create the secrets using the credentials file. So I'll just directly copy and paste this particular command. So the name space has been created. We can say cubectl create - f. I will give the location of the credentials file which is in my downloads folder. So the secrets has been created. To do a quick verification we can say cubectl get ns. And here you can see the randoly agents name space has been created. With this we can quickly verify the secrets as well. So we can do cubectl get secrets minus n randoly agents and we can see the two secrets have been created. Now till this point we have correctly configured the authentication which the operator and the agent will use to communicate with the randoli platform. Now in order to deploy and manage the randoly agent in our cluster, we'll install the official randoli agent operator from the open shift operator hub. So this is the dashboard for my open shift cluster. I can head over to the operators. Click on operator hub. In the search bar, we can type randoli and here you'll find the certified randoli agent operator which is purposefully built for open shift clusters. Now this particular operator will install the agent with customized defaults and permissions which are required to run properly on an open shift cluster. Make sure that you're selecting the latest version. Currently, as of this recording, the latest version is 1.17.2-1 and we can go ahead and click on install. Here we leave all the default configurations and we'll just click on install. So this will install the randomly agent operator for our open shift cluster. Right? So the agent operator has been installed. We can quickly verify this by clicking on view operator. And here we can find more details about this particular operator. But the important part right now is that the operator has been successfully installed. Now because randoly is open telemetry native, the agent relies on open telemetry for a number of features. If open telemetry is not installed in your cluster, we recommend to install the red hat build of open telemetry which is already provided by Red Hat. So coming back to the operators hub page, we can search for open telemetry and here you'll find the Red Hat build of open telemetry which basically will configure all the open telemetry related permissions in your cluster which later the randomly agent can seamlessly use to collect the relevant telemetry data. So here we can click on the install button. Again we'll keep all the default settings. Make sure that you are on the latest version. Currently, as of this recording, the latest version is 0.127.0-2. So, I'll just hit the install button. This will install the open telemetry operator as well. One of the last dependencies that we need to install in order to properly configure the randomly agent is network observability. The agent relies on this to capture network level data from your cluster. So, let's go ahead and set this up using the operator provided by Red Hat. So in the operators hub page we can go ahead and search network observability and you can see we have the network observability operator that deploys a monitoring pipeline which consists of an ebpf agent flows pipeline and a console plug-in for flows visualization with powerful filtering options. Once this is configured you'll be able to take the full advantage of all of these features in the randoli UI. Keeping all the default options, we can click on install. Again, we'll keep all the default options and make sure you are installing the latest version. At the time of recording, it is 1.9.1. And we'll click on install. Now, there is one last dependency that the net observe operator needs to run, which is the flow collector. This is basically a custom resource which is required to capture the traffic flow. To keep things simple, you can go ahead to our documentation page for Red Hat Open Shift and find the flow collector custom resource yaml which you can directly apply in your cluster in the net observe name space. So let's do that. So first we can create a new name space for net observe. I can write cubectl create ns net observe. I have all the configuration for the flow collector in flowcolctor.yaml as you can see right here. One important thing to note here is that if you're using the recommended custom open telemetry collector by randoli, make sure the open telemetry exporter in the flow collector yaml points towards the randoli agents where randoli's open telemetry collector has been deployed. If everything looks good, we can go ahead and write cubectl create d-f flow collector.l which will create the flow collectors custom resource in the net observe name space. Now we can quickly verify the status of the network observability operator and the flow collector by heading over to the installed operators page. And here we can select network observability. Here we can click on flow collector. And as you can see the status of the flow collector is ready. That means the collector has been successfully deployed. With all the prerequisite operators installed we can now go ahead and create a new randoly agents instance. So we can go ahead in the randomly agent operator and here we can head over to the randomly agent tab and click on create randomly agent. Now here based on the solutions that you are trying out you can go ahead and select the relevant option from the particular list. In this case I'm only selecting the observability which will enable and install all the observability related components with my agent. Once the option is selected we'll switch to the YAML view. And here instead of open shift net observe operator name space we'll go ahead and change this to randomly agents which is the default name space for all the randomly agents related components. Once this is changed we can click on the create button and this will create the randoly agent instance inside our cluster which when successfully deployed will install all the necessary observability components in order to visualize and get the insights on the randoly UI. Now the agent installation typically will take around 2 to 5 minutes depending on the product that you have selected during the configuration. We can go ahead and check the status in our cluster as well. So we can do ubectl get pods minus n randoly agents and as you can see that all the components plus the randoly agent is running and that means everything has been correctly configured. And now we can go ahead on the Randoli UI. Click on the clusters page. And as you can see our Open Shift demo cluster has been connected via the Randoli agent. Now if you also wish to try out the cost solution alongside observability, there are two extra steps that you need to do. The first one is to enable the cost management tag inside our agent configuration. So you can go ahead to the randoly agent instance. And here in the select product section we can enable the cost management tag and click on confirm change. This will enable and install all the cost related components inside the randoly agents name space. The second component you need to install is the vertical pod autoscaler which is required for a number of cost related features. If the VPA is not already installed in your cluster, you can head over to the operators hub on the console and type VPA. And here you'll find the official vertical pod autoscaler operator which is provided by Red Hat. You can go ahead and install in your cluster and you'll be able to use the cost solution alongside observability as well. And that's the end. Your open shift cluster is now connected to Randoli with full observability enabled. And now you can go ahead and try out the different features to build a robust monitoring and troubleshooting workflow for your infrastructure and your applications. For more details and a written version of this, you can go ahead and check out our documentation. You'll find the link somewhere below. Thank you for watching.