In this video, run through a demo of Migrate for Anthos for hybrid compute scenarios.
- [Instructor] Anthos is a set of services that allows you to run containerized applications anywhere on Google Cloud, another vendor or on prem. It has a rich set of interfaces for management, monitoring and more. I highly recommend that you take a look at the technical overview shown in Google documentation here, starting with the architecture. You can see that there are a number of layers that are supported, and a key aspect of Anthos is policies. So you can see this middle layer here has policy enforcement around Anthos config management, Anthos enterprise data protection and a policy controller. Now, there's more to Anthos than that. So again, do take the time to read through the technical documentation. Probably the best way to get some understanding of the capabilities is to explore some of the samples. So we'll briefly look at that in this movie. So the example that Google has is a banking application. It's a microservices application and again, I built it out here. It does take at least 15 minutes, literally just to click the button to set up but if you read it, it probably takes a couple of hours to understand. So again, if this is something you're interested in, read the tutorial and then a best practice is to create a project 'cause it creates so many different resources. That way when you're done, you can just delete the whole project. So I've done that and I have run that and let's take a look at it. So the first thing is out of the Google Cloud Platform Marketplace, I selected the Anthos Sample Deployment and executed that as a deployment and of course, that's a set of application files that creates a number of artifacts. So at a high level, this is creating a cluster of Google Container Engine instances, managed by Kubernetes and Anthos services, running on GCP. So you can see in the successful deployment that I have a number of services here and if I want to examine any of the individual files, I'm just going to go ahead and look at one of them. So let's say I want to see the CPU quota. I can see what that information is set by default here in this particular deployment. Now, once the deployment succeeds, which like I said, took about 15 minutes in this project container that I created, then you have in the web UI an Anthos section. Now, you can see some of the Anthos capabilities in the section and you can see that if I go to the dashboard here that I'll have a billing report, and I just turned this one yesterday, so I don't really have any charges. You have the service mesh, which is observability. The configuration management, health and the cluster status. So if I go into the service mesh, and I select show the system services, you can see that I've got a number of services. I can see them viewed as a table but interestingly, I can also see them viewed as a topology. Now, if I drill into this, you can see I have observability over time and I can adjust this here. So if I go to a certain section, I can see latency. It's a little bit hard to read 'cause it's kind of crunched together. I can see the QPS there. So I can see for, and these are the various microservices, how my cluster is functioning, given the resource limits that it's been set up to work with. Now, of course, this is running on top of on the Google Cloud Kubernetes. So if you're using the web UI in addition to using the Anthos web UI to understand the overhead on your cluster, you're going to want to take a look at GKE web UI, which here, if I take a look, I can see that inside of here, I have my cluster with workloads and services and the boa is the banking application. So if I set for my cluster for the namespace of boa and istio, which is the service that is providing the ingress and egress. So if I go here to the front end, I can actually put some load. This is going to redirect. And this is the application, the banking application. So I can go ahead and make a deposit. We can see that will show up in a minute here. And then I can make a payment. And there's my payment and my deposit. So now if I go back into my cluster, I can take a look at the level of the microservice. So I'm not sure actually which one will be used here. Maybe transactions. On the workload and I can look at the level of the cluster to see, again I don't have much data 'cause I've just been playing around with this for this movie, so I haven't done a lot with it, but the idea is very granular management of your microservice pieces. Again, Anthos has several capabilities. It can also be used for migration of VM-based workloads to containers. It's a very rich ecosystem. In addition to exploring the sample, if you are interested in learning about this, I would recommend reading this article. This talks about some scenarios of control plane revisions. So again, some really powerful things you can do in terms of high availability with Anthos. For example, one scenario is auto-injection. So Anthos uses Kubernetes admission control. So a mutating admission webhook is registered to watch for newly created Pods and then the webhook is configured with namespace selector so it matches the Pods that are being deployed. And when a Pod matches the webhook, consults an injection service, provided by Istio OD to obtain a new mutated configuration for the Pod. So again, it's a dynamic compute cluster so really what it's going on here. And this is just one scenario. There's multiple. Now, in addition to this sample, there is a dev sandbox that is at anthos-sandbox.dev, which is another alternative way to play around in a more interactive fashion with some of the features of Anthos. So I recommend to get started, you either set up the sample that I've shown here and/or you work in the Google Cloud shell with the developer sandbox so that you can start to understand the capabilities and how that would match your scenarios.
- Enterprise concerns
- Enterprise scenarios
- Setting up your organization’s account
- Managing billing
- Enterprise compute services
- Enterprise storage and database services
- Enterprise data pipelines
- GCP developer and DevOps tools