Why Build a Cloud Foundation
The cloud foundation is an organizational approach to managing cloud adoption and consumption as a single organizational capability. As an operations model, the cloud foundation provides application teams a single point of contact for provisioning, accessing and managing all cloud services they need to get their job done. Like a cloud provider, the cloud foundation offers its services as a platform using a unified control plane integrating tenant management, identity and access management and billing. The cloud foundation frees application teams from organizational burden and empowers their focus on software delivery.
From IT Services to Cloud Foundation
Most organizations today manage IT as a portfolio of "IT Services". An IT Service provides a distinctive functionality to its customers. For example, an "IAM Group" service may allow a customer to create a group in the organization's central directory and manage its members through the service. The owner of the "IAM Group" IT service gets to define its own processes for how customers can procure, manage and get charged for the service.
With IT services as the established management paradigm in mind, most organizations trying to move to the cloud instinctively reach towards IT services for providing the required organizational capabilities to procure, manage and charge back cloud services (cloud governance). As the following section will demonstrate, this corrupts the advantages that cloud computing has to offer. Whereas cloud computing emphasises customer self-service, consistent API-driven automation and pay-per use, most IT services build on semi-manual processes, with inconsistent provisioning workflow and charge back back to a shared cost pool.
💡 To take full advantage of cloud computing, organisations need to adopt a cloud-native organizational approach for providing capabilities to provision, manage and charge back cloud services.
Functional Pillars of Cloud Governance
No matter how an organization decides to implement its cloud governance, it always needs to cover the following functional pillars:
🗂 Tenant Management: Manage provisioning, configuration and lifecycle of cloud environments (e.g. AWS Accounts). This is often an original responsibility of cloud foundation teams that requires internal alignment between platform specialists.
🔐 IAM: Provide Identity and Access management capabilities for all available cloud platforms and services. Needs alignment with Enterprise IAM stakeholders.
🔖 Security & Compliance: Provide capabilities for governing workloads and enforcing security guidelines across all available cloud platforms and services. Interfaces with IT Security and Compliance stakeholders.
💵 Cost Management: Provide cost management and chargeback capabilities for all available cloud platforms and services. Interfaces with Finance/Controlling stakeholders.
🛠 Service Ecosystem: Provide managed services that help teams build and operate application on the cloud faster and more efficiently (e.g. on-prem connectivity). Interfaces with internal or external teams providing these services.
While representing independent functions, organizations need to integrate these functional pillars to implement an effective and wholistic cloud governance.
As an illustration, let's look at the dependencies of the Tenant Management pillar. For an effective cloud governance, each of the pillars needs to maintain different metadata associated with a cloud tenant (e.g. an AWS Account).
Manually maintaining this associated metadata in different functional silos is challenging and burdensome. Data inconsistencies are likely and make applying good governance difficult. Imagine detecting a critical security vulnerability in one of your AWS Accounts, only to figure out that the responsible security contact is no longer working at your organization...
This example is only scratching the surface of the complex inter-dependencies that can exist between pillars. Managed services for example will require integration with IAM (who can order, configure a service?), Compliance (is the service allowed for a particular cloud environment?) and Chargeback (who's going to pay for the service?) etc.
A Team's Cloud Journey - with IT Services
Let's look at the cloud journey for an application team that wants to build a typical cloud application. The team wants to take advantage of cloud-native services for machine learning, however it also needs access to on-premise systems that hold required data. The team therefore needs a cloud tenant and on-premise connectivity. When an organisation manages its cloud capabilities as IT Services, the team's cloud journey will look like this:
As the figure above illustrates, the team needs to "shop around" different services and integrate their capabilities. Provisioning each service involves finding the right contact point, following its procurement process and then composing it with other services. This is a challenge as teams may need to learn about the hidden interdependencies between IT services and how they can make them "click" together. For example, when a Team provisions a new AWS Account, they may need to set up an IAM group with a conventional name that contains the AWS Account number.
A Team's Cloud Journey - with a Cloud Foundation
The Cloud Foundation approach frees application teams from integrating different functional pillars. Instead, the cloud foundation team assumes responsibility for integrating the different functional pillars. Like the service offering of a cloud provider, the cloud foundation enables application teams to provision and manage cloud services from a single organisational capability and leveraging a consistent control plane offering API-driven self-service, IAM and billing.
To achieve integration between the different pillars, the Cloud Foundation team implements capabilities like a Self-Service Multi-Cloud Tenant Database and Federated Identity and Authentication. To provide On-Premise Network Connection, the team provides an API-driven service that orchestrates provisioning an IP range from IPAM and then setting it up on AWS Transit Gateway.
Of course, building a cloud foundation is a challenging task. It can involve interesting technical challenges that allow an organization to achieve a high degree of automation that compares favorably to public services. However, achieving full service automation is not required for adopting the cloud foundation as an organizational approach. The key to its success is that it absorbs responsibility for service integration and allows an organization to exercise wholistic and effective cloud governance.
💡 Most organizations find the most difficult challenge in building a cloud foundation is the required interdisciplinary collaboration and alignment of different stakeholders.
Here is how the Cloud Foundation approach compare to the IT Services approach.
|Criteria||IT Services||Cloud Foundation|
|Function Alignment||Stakeholder-to-Stakeholder or Service-to-Service. Lack of global coordination. Missing visibility of governance capabilities across functional pillars leads leading to blind spots.||The Cloud Foundation team acts as a single point of contact to all stakeholders and creates alignment and visibility about governance capabilities.|
|Points of Contact||Application teams need to contact each IT service owner individually. Service owners need to maintain individual customer databases with metadata and billing info.||Application teams use the Cloud Foundation as a single point of contact to maintain project metadata and billing info.|
|Prevalent Automation Pardigm||Semi-manual workflows.||Declarative and API-driven automation.|
|Service Integration||Application Teams need to individually connect and integrate services. They need to learn about potential hidden interdependencies between services they want to consume and resolve resulting integration challenges.||The Cloud Foundation assumes responsibility for integrating required services and makes the result available as a consistent, integrated offering.|