With uncertainty and disruption swirling through the economy, being agile has never been more critical for a business.

It’s vital to change everything from workflows and supply chains to service delivery and communication channels if a business is to survive and thrive in the coming months. Taking a ‘business as usual’ approach is no longer an option.

Increasingly, this agile strategy is also being applied to the task of software development. Organisations realise that having the ability to write and test code quickly can mean the difference between capitalising on a new opportunity and missing out.

The rise of DevOps
To help in this process, growing numbers of organisations are embracing the DevOps approach to code creation. DevOps enables a business to be agile enough to quickly adapt or be the first to market.

Many are also realising the need to add identity to their DevOps strategy. Previously, identity initiatives have tended to stretch organisations’ limits while they struggle to keep pace with application growth.

Now, finally, DevOps is starting to gain traction in the identity space, and for good reason — identity is becoming a crucial part of application onboarding. Without collaboration between identity teams and DevOps teams, many of the intended benefits of DevOps can quickly unravel.

The DevOps approach to software development breaks down traditional silos by eliminating manual tasks and replacing them with automation so that releases and updates can be pushed out faster and more frequently.

At its core, DevOps improves the efficiencies between application development and IT operations or infrastructure teams. Previously, developers would make manual requests to identity teams or IT ops and lose valuable time.

In DevOps, the identity teams enable the software to be consumable via APIs or infrastructure-as-code, which are the preferred methods for developers.

Also, the DevOps model has specific tools for implementation. The most common include Docker, which enables software in images or ‘containers,’ and Kubernetes, which is the orchestration engine that manages containers. These tools allow developers to easily spin up infrastructure and perform updates with minimal interruption.

DevOps and identity
When launching a new service or deploying a new application, it’s critical to do so within an organisation’s existing identity ecosystem. This is typically achieved by using a federation hub or centralised authentication services.

However, many identity teams struggle to keep up with the sheer volume of requests that come from the DevOps environment, which can significantly outnumber the release cycles of other application methodologies.

Also, in many cases, developers don’t care about the details of effective identity management. At the same time, the identity team doesn’t need to have DevOps experts to enable developers.

For identity to be consumed in a DevOps environment, identity teams need to prepare pre-configured profiles or images. They then need to make them readily available on the most popular tools used by DevOps teams. And most importantly, identity teams need the ability to centrally manage identity as the DevOps environment scales.

This can be achieved by using containers that allow an identity platform to be pre-packaged and consumed in any app environment. It’s also essential to have a centralised way of managing identity and understanding how it’s deployed in a DevOps environment.

The power of the container
When identity is deployed via containers, it provides operational flexibility that can have a significant impact on a business. Some of the key benefits include:

Consistent delivery: Deploying configurations as code eliminates the need for manual processes. This allows development teams to deploy identity with confidence in a repeatable manner.

Platform flexibility: Tools such as Docker and Kubernetes are platform agnostic and based on open standards, providing an organisation with an improved level of independence.

Better deployments: It’s also possible to leverage cloud scaling capabilities to meet demand, SLAs and other defined thresholds. Auto-scaling and auto-healing features optimise spending so that the organisation is using only the capacity that it requires.

Cloud flexibility
Building identity into DevOps processes does not limit teams to working in just one environment. Essentially, anywhere that you can put a Docker image, you can put identity. Docker images can be deployed in any cloud, including AWS, Azure and Google Cloud.

Docker images can also be deployed directly to an on-premises data centre for mission-critical use cases where uptime needs to remain under strict control. Whatever environment you choose, DevOps ensures that you easily move and re-apply your identity infrastructure when required.

Ensuring identity is included in any DevOps strategy can deliver significant advantages for an organisation. Rather than being seen as a necessary evil, it can become a crucial component in an agile and powerful infrastructure.

Google Cloud announced a wave of new capabilities for Anthos, its managed application platform, as part of the broader Google Cloud App Modernization Program (Google CAMP). The updates span hybrid AI, cluster management, developer tooling, identity federation, security blueprints, and workload migration tooling.

Hybrid AI for On-Premises and Edge Environments

Google Cloud introduced hybrid AI capabilities for Anthos, enabling organisations to run AI workloads wherever their data lives — whether on-premises, at the edge, or in the cloud. The motivation is twofold: keeping AI inference close to data sources for latency and compliance reasons, and reducing the complexity of building AI-enabled applications across distributed environments.

Speech-to-Text On-Prem reached general availability on Anthos via the Google Cloud Marketplace. It gives organisations full control over speech data subject to data residency and regulatory compliance requirements. At the time of the announcement, five languages were supported, with more planned.

Cluster Management: Attached Clusters and Bare Metal

Two significant infrastructure additions expand where and how Anthos can be deployed:

• Anthos Attached Clusters: This feature allows existing Kubernetes clusters — regardless of where they run — to be managed through the Anthos control plane. It provides centralised configuration management and Anthos Service Mesh capabilities without requiring clusters to be provisioned by Google.
• Anthos for Bare Metal (Beta): Anthos can now run directly on physical hardware on-premises or at edge locations, eliminating the need for a hypervisor layer. Google noted it was itself an early adopter, working toward using Anthos for bare metal as an internal production container platform.

Developer Tooling: Cloud Code and Cloud Run for Anthos

Google unified Cloud Code IDE plugins with Cloud Run for Anthos, enabling developers to build and deploy serverless applications without leaving their editor. Supported IDEs include VS Code and IntelliJ. Supported languages at launch include Java, Node.js, Python, and Go.

Key developer experience improvements include:

• Local Cloud Run Emulator: Developers can validate code changes locally with automated re-deploys triggered on every save, closing the feedback loop without cloud round-trips.
• Local Debugging: The emulator supports local debugging of Cloud Run applications directly from the IDE.
• Remote Push: Once code is ready, changes can be pushed from the IDE directly to a remote development environment in Google Cloud.
• Cluster Provisioning from the IDE: Developers can create Kubernetes clusters with Cloud Run for Anthos enabled from within Cloud Code, with the IDE pre-populating fields such as project ID, zone/region, and node count.

Identity and Security Enhancements

Anthos Identity Service allows organisations to extend their existing identity infrastructure to Anthos workloads rather than maintaining separate authentication systems per environment.

• OpenID Connect (OIDC) support was generally available for on-premises deployments and in beta for Anthos on AWS, with additional protocol support planned.
• Anthos Security Blueprints provide templated best-practice configurations covering auditing and monitoring, policy enforcement, and locality restrictions (for data sovereignty scenarios).
• Blueprints also offer purpose-built compliance automation for regulated industries including financial services, retail, and the public sector.

Workload Migration to Anthos

Migrate for Anthos received new capabilities to ease the transition of existing workloads — even those for which source code is unavailable:

• CRD-based API for build migration automation: Enables integration of the migration pipeline with custom tools and processes.
• On-premises VM conversion: Support for Anthos deployed on-prem to convert VMs running locally to containerised workloads.
• Windows container support (Beta): Broadens Anthos migration coverage beyond Linux workloads.
• Google Cloud Console integration: Migration workflows are accessible through the standard Cloud Console web UI, reducing the need for command-line tooling.
• Cloud Foundry migration via Kf on Anthos: A new migration path for Cloud Foundry applications using Kf, which presents developers with a familiar Cloud Foundry-like interface running on top of Anthos, easing the transition without requiring full application rewrites.

Summary

These announcements collectively push Anthos further toward its goal of being a consistent application platform across any infrastructure — Google Cloud, other clouds, on-premises, and bare-metal edge. For DevOps and platform engineering teams, the most immediately actionable additions are the attached cluster management, the bare-metal beta, the Cloud Code developer workflow improvements, and the CRD-driven migration automation for Migrate for Anthos.

For up-to-date information, consult the official Anthos product page and Google Cloud release notes, as feature availability will have evolved since this August 2020 announcement.