Summary

Google Kubernetes Engine is a powerful combination of platform and tooling to support your containerized workloads in the cloud. While GKE is a great choice for just about any type of workload, as we mentioned early in this chapter, it's especially well-suited for microservices.

GKE provides much more than just a platform for running your containerized workloads. GKE also provides the tooling and supporting system components to deploy, operate, and manage workloads at enterprise scale. Other native GCP-managed services, such as Cloud SQL and Cloud Datastore, are also easily made available to connect to from your GKE-managed container clusters. Some of the most common integrations include:

• Subscribing or publishing to pub/sub topics
• Utilizing Cloud SQL and/or Cloud Datastore for a transactional database
• Storing files in Cloud Storage

How you integrate with these services will differ based on your specific needs and the services you use. For instance, you may want to create a proxy when integrating with Cloud SQL so that your cluster as a whole is communicating with your database instead of individual nodes. On the other hand, when subscribing to pub/sub topics you may want to maximize parallelism and throughput by allowing each instance of a workload to pull messages off a topic.

Of course, there's much more you can, and should, learn about Kubernetes and GKE. There are entire books dedicated to Kubernetes that cover the nuances of getting the most out of the platform and integrating it with other systems and platforms. I would also encourage you to look deeper into infrastructure as code and CI pipelines that allow for unprecedented levels of testing and deployment automation.