 we will use Pixie's live view tool and the network flow graph script to get network stats for a list of IP addresses. This script was requested by a bunch of our early adopters, including some large streaming companies. Pixie has pretty broad visibility into the system, so you can actually get fairly detailed information about HTTP data, GRPC data, and a bunch of other protocols. But there's also a lot of traffic we don't yet support at a protocol capture level. So part of the idea with this script is to provide basic network stats, even for protocols we don't support, like CouchDB. So we've got the live view open, and this is our cluster, our prod cluster. Now it's a lot less interesting than some of our bigger customer clusters, but it'll work for this demo. So let's run the network flow graph script. You can select a pixel script with the drop-down menu here using the Pixie command button, and that can also be brought up with the command K keyboard shortcut. So we'll find the net flow graph. Now this script asks you to provide a list of IP addresses, and I happen to know that we have a database located at 10. So I selected that, and this will show me all the network information for that IP. So we can do things like group by pod, service, or namespace. So let's group by service. So here you can see all of the services that are talking to our database, which is this blue IP address right here. So PLC is the namespace for our cloud, so here's all of our services, and the thicker, bolder lines suggest more traffic, and when you hover over the edges you can see what all the traffic is. So one interesting thing that we discovered is that we actually have two shards of this Vizier manager service, and one of them is sending a lot more traffic to the DB than the other. So that's the nice thing about this graph, is that you can quickly get a high level of visibility about the whole network, and then use that to debug issues. And part of the Pixie approach is to have these open source scripts where teams and individuals can contribute new scripts or visualizations. You can see the source code for any script by opening the editor with this button or with the command control E keyboard shortcut. And we'll have a tutorial coming soon about how to modify these pixel scripts or write your own.