 In this video, we're going to use the Pixie data source plugin and Grafana's node graph panel to visualize the HTTP requests flowing through my cluster. So first, what is Pixie? Pixie is an open source observability tool for Kubernetes applications. Pixie uses EBPF to automatically collect a lot of data out of the box, such as infrastructure and network metrics, application profiles, and full body requests. Since Pixie automatically traces all HTTP traffic in your cluster, once you install Pixie, you can immediately get this service map showing the flow of traffic. So here you can see services and pods talking to other services, pods, and remote IP addresses. So Pixie represents a pod. It puts the namespace in front of it, so you can see we've got the stock shop namespace and this is our front end pod, and it's talking to a variety of pods. Here we have an external IP address that's making HTTP requests to our orders pod. As long as one endpoint is within the cluster, Pixie is able to trace that request. If you hover over an edge, you can see the error rate on the top and the latency in milliseconds on the bottom. Okay, so let's build this panel. So I already have Pixie installed on my cluster, and I have the Pixie Grafana data source plugin installed, and then I've also added the data source, and to configure this, check out the directions in the description below for our other YouTube videos. So let's go back to dashboards, and let's add a panel. Okay, so we're going to use the nude node graph visualization, and we'll call this HTTP for this map. So I'm going to copy in a script, and you can find that script in the description below of the video. I'm going to use the Pixie data source plugin, and this may take a second. Let's actually reduce it to the last 30 minutes, and usually you have to refresh this once or twice. Okay, for some reason it requires refreshing a few times. So here's our node graph, and the way this works is that the pixel script we have down here returns two tables, and you can switch to look at the table view up here. It returns an edges table, which has the source and the destination, and then the stats you want to associate it with. So in our case, we have error rate, latency in milliseconds, and the nodes table, which just lists all of the nodes that exist in our node graph. So we can go back here, and we can quickly see with this graph which pods and services and remote IP addresses are talking to each other within our cluster. To see more of what you can do with Pixie's Grafana plugin, be sure to check out the other videos in the Grafana playlist.