Blog post

How to Scale DevOps Workflows in Multicluster Kubernetes Environments

By Manjunath Bhat | August 02, 2022 | 0 Comments

This Gartner blog post is an excerpt from the detailed Gartner research published here – How to Scale DevOps Workflows in Multicluster Kubernetes Environments

Summary: Organizations are increasingly deploying multiple Kubernetes clusters across on-premises, cloud and edge environments. To streamline DevOps workflows, I&O leaders must build platform teams, automate cluster life cycle management, enhance developer self-service and adopt GitOps practices.

Strategic Planning Assumption(s)

By 2025, multicluster management and security will emerge as top challenges for organizations deploying Kubernetes applications.

Multicluster management presents a set of challenges for both application development teams and platform teams. First, the need to manage the life cycle of cluster software on multiple clusters — including updates and security fixes. This requires consistent visibility and manageability across a fleet of clusters. Second, the need to improve developer experience via self-service environment provisioning. Third, the need to secure and streamline the software delivery process in a consistent and automated manner. How can organizations enable frictionless DevOps workflows as they go from managing containers at scale to managing clusters at scale?

Step 1: Establish Platform Teams to Standardize Cluster Management Practices

Platform teams help alleviate the risks of DIY cluster management approaches by standardizing Kubernetes consumption models and the cluster life cycle management strategy. The complexity of management increases in proportion to the number of clusters. Platform teams must therefore enable self-service capabilities to provision and manage development and test environments (namespaces) within a cluster.  In addition, implement guardrails using policy-as-code to minimize risk exposure (i.e., credentials, secrets, code, etc.).  

The platform team builds an organization-specific container-native ecosystem which enhances developer agility and improves developer experience — not a wilderness of fragmented tools. Organizations typically end up with a “wilderness of tools” as more teams build containerized applications without a coherent way to address manageability, security, interoperability, scalability and architectural requirements.

The four steps to scale DevOps workflows in multicluster Kubernetes environments are to establish a platform engineering team, automate cluster life cycle management, enable self-service environment management, and adopt GitOps practices for declarative application and infrastructure delivery.

Step 2: Automate Cluster Life Cycle Management

Multicluster management responsibilities for platform teams include cluster life cycle management, cluster policy management, application management, data management, security and identity management and administrator interface.

Step 3: Improve Developer Experience by Enabling Self-Service Environment Management

Representative Set of Providers That Simplify Environment Management

In addition to cloud management and container management tools that provide some capabilities for developer self-service, Gartner sees a few providers focus on improving developer self-service in Kubernetes environments as one of their core value propositions. Here’s a representative list of providers:

Step 4: Adopt GitOps Practices to Achieve Scale

To manage multiple clusters at scale, the architecture involves a management cluster which is then used to manage a fleet of workload clusters. Workload clusters pull desired state and push the current status back to the management cluster.

Leave a Comment