This solution introduces Pivotal Cloud Foundry (PCF) and the benefits of running PCF on Google Cloud Platform (GCP), such as:
Rapid VM provisioning for scaling the platform to help meet your developer and user needs.
Use Google's load balancer to help scale your apps to 1M+ requests in seconds without the need for pre-warming.
Save up to 30% with sustained use discounts for virtual machines that run for a full month period.
Save up to 80% with preemptible VMs for applications that can handle virtual machine restarts, compared to regular instances.
Fully tailor your PCF deployment’s VMs with Google Compute Engine custom machine types.
Leverage Google Cloud Storage’s low cost, durable and scalable object storage as a backing store for your Cloud Foundry deployment.
Use many of Google’s big data services through the Cloud Foundry service broker for storing your application data in a resilient and distributed manner.
About Pivotal Cloud Foundry
Pivotal Cloud Foundry is a cloud native platform for deploying next-generation applications. Based on open-source technology, PCF enables enterprises to rapidly deliver new experiences to their customers. PCF can be deployed on-premises and on many cloud providers to give enterprises a hybrid and multi-cloud platform. It provides a uniform way for you to launch and quickly iterate on ideas in the language of your choice, allowing you to focus on adding business value with applications rather than designing and implementing how the application will be deployed, scaled, and exposed to end users. PCF’s orchestration mechanisms enable developers to accelerate application development and reduce time to market.
The following diagram provides an overview of PCF’s components and functionality, including GCP components that you can integrate seamlessly with your deployment.
PCF assembles applications and their dependencies using buildpacks. When operating PCF on GCP, these artifacts are stored in Cloud Storage. A PCF deployment on GCP also utilizes Google's load balancer, and other software-defined networking features such as subnetworks and firewall rules. PCF allows administrators to register a service broker, which allows application developers to leverage GCP’s data services natively.
These platform-level integrations allow you to harness Google's infrastructure to reduce your operational overhead and provide a more resilient installation.
Running PCF on GCP
Running PCF on GCP provides developers with the state-of-the-art networking and computing infrastructure that runs services such as YouTube, Gmail and Google Maps. With GCP’s global HTTP load balancer, you can host your application code in a multi-region global environment with minimal effort and configuration. PCF allows you to provision Google’s data services such as Cloud SQL, Cloud Datastore and Google BigQuery from a service catalog, and easily discover and integrate them into your application regardless of which infrastructure provider PCF is deployed on. The PCF Elastic Runtime can be configured to provide high availability, self-healing and autoscaling of applications, with minimal configuration.
Scaling your PCF applications on GCP can lead to significant performance improvements compared to on-premises or other cloud providers, because of Google's sub-minute VM provisioning times. Cloud Foundry virtual machine images are called stemcells. Download Compute Engine stemcells at bosh.io.
The following diagram shows an example of a multi-region PCF deployment that uses the global HTTP load balancer’s single anycast IP to route user requests to the closest region.
Platform administrators can leverage OpsManager in PCF version 1.8 or higher to configure, deploy and scale PCF on GCP. You can rapidly provision Compute Engine instances in minutes to scale the underlying platform as your application traffic increases. PCF handles scaling down your applications and underlying infrastructure when traffic subsides, helping to ensure that you are only paying for computing resources that are actively used. With Compute Engine’s sustained use discounts, the baseline workload of your PCF installation can receive up to a 30% discount on its VM pricing.
The Google BOSH CPI is the mechanism that allows your PCF deployments on GCP to manage your Compute Engine VMs. If your workloads permit intermittent VM downtime, such as using Concourse CI for your application testing, you can combine Compute Engine’s preemptible VMs with BOSH's powerful self-healing capabilities to restore stopped VMs and save up to 80% off standard VM pricing.
PCF can be deployed across multiple providers, including on-premises data centers, to help give your developers a seamless experience between environments. Developers will not be able to differentiate which infrastructure provider their applications are running in due to the elastic runtime’s clean abstraction between applications and their underlying infrastructure. PCF’s platform abstraction allows deployment and site reliability engineers within your organization to focus on making sure that the platform is running reliably while developers can focus on ensuring that their application code is performing properly.