Book Image

Google Cloud Platform for Developers

By : Ted Hunter, Steven Porter
Book Image

Google Cloud Platform for Developers

By: Ted Hunter, Steven Porter

Overview of this book

Google Cloud Platform (GCP) provides autoscaling compute power and distributed in-memory cache, task queues, and datastores to write, build, and deploy Cloud-hosted applications. With Google Cloud Platform for Developers, you will be able to develop and deploy scalable applications from scratch and make them globally available in almost any language. This book will guide you in designing, deploying, and managing applications running on Google Cloud. You’ll start with App Engine and move on to work with Container Engine, compute engine, and cloud functions. You’ll learn how to integrate your new applications with the various data solutions on GCP, including Cloud SQL, Bigtable, and Cloud Storage. This book will teach you how to streamline your workflow with tools such as Source Repositories, Container Builder, and StackDriver. Along the way, you’ll see how to deploy and debug services with IntelliJ, implement continuous delivery pipelines, and configure robust monitoring and alerting for your production systems. By the end of this book, you’ll be well-versed with all the development tools of Google Cloud Platform, and you’ll develop, deploy, and manage highly scalable and reliable applications.
Table of Contents (17 chapters)

IAM and billing

One of the major selling points of Cloud Functions is the potential for major cost savings. Because functions only use compute resources during invocation, they tend to be much cheaper than maintaining dedicated services that tie up resources throughout their lifetime.

Cloud Functions are billed based on a few operational metrics:

  • Number of invocations
  • Provisioned compute resources and duration of execution
  • Network resources

For number of invocations, functions are charged in units of millions. The current price for this is $0.40 per million invocations per month, with the first two million invocations per month being free. Compute resources are calculated as units of 100 milliseconds execution time multiplied by the amount of compute resources allocated during deployment with the --memory flag. For network resources, all inbound traffic is free, while outbound...