Scaling in Google Cloud is conceptually similar to AWS. You have a load balancer, instances, and policies that can scale your instances up or down. Google Cloud also, just like AWS, offers more than only one way to manage services and autoscaling: Kubernetes can be directly managed within Google Cloud (just like in the AWS Elastic Kubernetes Service (EKS); see Chapter 14, Docker and the Cloud, for details).

Let's assume, though, that we have set up our app through App Engine like we did in Chapter 15, Deploying Microservices in the Cloud. In that case, autoscaling is already enabled by default! There is nothing more we need to do. Google is automatically checking the CPU, RAM, and other metrics, and scales up and down depending on your needs.

If, however, you have a more sophisticated setup, in which you have set up a few instances of your own, you can have autoscaling set up when setting up managed instance groups.

The following screenshot shows the setup when setting up a new managed instance group. Notice how you can add autoscaling policies here—very similar to AWS, as shown previously: 

And that is it; App Engine is now doing the rest for you. Notice how the settings are, in principle, the same between Google Cloud and AWS? The reason is that scaling has become universally similar across different providers and solutions. Everyone is looking for the same things to scale up or down.

Now that we have had a brief look at Google Cloud, let's see how DigitalOcean does this in comparison.