As we’ve explained in previous posts, the deployment of 5G Core (5GC) networks typically marks the point that cloud-native technology is introduced into the core of communication service providers’ mobile networks.

The benefits that cloud-native design principles offer, such as: software decomposition and improved lifecycle, greater interdependency between applications and underlying infrastructure, and higher levels of resilience and automation, will be crucial to achieve the expected efficiency and business innovation in 5G networks. However, it will also create new network operation and management challenges for service providers and will require an operational transformation like never seen before.
The operational challenges that come with cloud-native networks
Adopting cloud native principles will have a direct impact on business processes, OSS architecture and daily operations. Let’s break down them in some selected areas.

Virtualization: the migration from physical network functions (PNFs) to virtual network functions (VNFs) allows service providers to automate the testing and lifecycle management (LCM) of VNFs, enabling greater agility to launch applications in a virtualized environment. While legacy PNF networks involved physical locations and points for connectivity, virtualization introduces new interdependencies and an intermediary abstraction that disconnects resources from service features. This requires redesign of existing systems and processes, causing changes in network management tools and procedures as well as introducing a need for increased process orchestration.

New and legacy coexistence: the new cloud.native functions (CNFs) from the 5GC will need to coexist with the legacy systems in the network. Ensuring smooth and timely network migration, with reduced impact over the management layer, requires compatibility with the legacy architecture’s interfaces.

The figure below shows an example of a process to manage multiple domains in a hybrid environment with, for example, coexistence of PNFs, VNFs and cloud network functions (CNFs).

Managing the partner ecosystem: the evolving 5G business models, where service providers transition toward becoming digital service providers, suggest the partnering of traditional operators and various ecosystem players. It’s expected that service providers play the central role in driving 5G business and leveraging partnerships; for that reason, end-to-end governance of design, deployment and day-to-day operations is crucial. Service providers need new tools for full end-to-end visibility of the ecosystem.

Adopting DevOps for efficient software delivery: Cloud-native software and microservices increase software granularity, which provides service and business agility benefits, but also means frequent changes and updates. To manage the growth in software updates, a continuous pipeline with related capabilities around integration, testing and deployment must be enabled. Thus, a continuous integration and continuous delivery (CI/CD) approach is needed.

Troubleshooting, tapping and probing: 5GC’s higher complexity will increase the need for observability tools for different core network components. Current troubleshooting tools should be adapted to 5G technology to reduce time to resolution and manage the networks effectively. Different analytical data is essential to enable automation and improve network service quality. Probing the core network will become more complex in 5GC with the service-based interface (SBI) between NFs, along with the introduction of container-based cloud-native NFs, posing new challenges to traditional probing methods. As a result, a new approach is needed to maintain security while continuing to capture the desired data.

Network slicing and edge computing: Network slicing will enable the creation of separated, intent-driven logical networks over a shared infrastructure enabled by NFV, software-defined networking (SDN) and Management and Orchestration (MANO) technologies. Each slice addresses an application or user’s need, covering speed, capacity, latency, security, and so on.

Moreover, the distribution of network resources over a cloud infrastructure enabled by edge computing will allow workloads to be executed in different parts of the network, meeting different use cases’ requirements.

Mastering these new technologies involves much more than running a virtualized network element. Virtual machines and/or containers need to be portable between data centers; the network orchestrator needs to have complete control of resources; and all components need to work together to allow a network to become cloud native. Service providers also need to address enhanced network automation capabilities to be able to handle the dynamic nature of the architecture.

It’s time to transform network operations
To achieve operational efficiency in cloud-native 5G core networks and business innovation, operational transformation is required. With the move toward cloud native, service providersface a scenario made up of new challenges and inherited pain points from earlier technology shifts but also great opportunities.

Based on our experience working with leading service. providers on their journey to NFV and new 5GC, we have found that operators often continue with their existing approach to operations when they move to a cloud environment, resulting in critical pain points being carried forward. With the evolution of hybrid networks consisting of physical and virtual elements, and considering multiple domains across core, radio and transport, it meant limited end-to-end (E2E) governance for design and operations across domains.

Siloed stacks within and across domains lead to disparate and siloed operations, causing multiple touch points for E2E service flow. Manual, time-consuming and inconsistent processes plus a lack of seamlessly integrated operating models increase the complexity of managing the network. Patchy and manual procedures for deploying disjointed workloads limit the ability to smoothly create and manage an E2E service.

Without the necessary changes to the operating model, the 5GC evolution will only increase the overall cost and complexity of managing the network as well as time-to-market (TTM) for new services – the opposite of what is expected with 5G.

These pain points transcend pure operational issues and affect business metrics in several ways:

A lack of clearly defined E2E business processes results in a long service sales creation and fulfillment cycle, which will severely limit enterprises’ business success.
Disjointed and multiple touch points caused by non-standard service design models extend the sales cycle and impact revenue and customer experience.
Absent onboarding standards and guidelines impact TTM and increase the complexity of monetizing new technologies (such as products/services like SD-WAN).
Increased errors in customer solution design and service delivery create frustration for both customers and service providers as well as revenue losses, higher expenses and even churn.
A lack of E2E network and process visibility makes it impossible to achieve zero-touch automation from selling to billing.
Insufficient closed-loop automation and E2E integration across the ecosystem negatively impacts network availability and time to resolve issues, resulting in degraded user experience.
Limited visibility and visualization of E2E topology increases time to repair.
A lack of comprehensive competence planning means organizations will lag behind in terms of technical transformation and knowledge, unable to extract the business value from new technologies.

Transforming for efficiency
A good transformation strategy is imperative to address the opportunities and challenges cloud native adoption brings. This strategy should be supported by the C-level and align business and technology needs with an organization’s current maturity level.

Your transformation strategy should tackle three main areas:

Processes: to manage, automate and orchestrate this hybrid core network in a cost-efficient way. The figure below shows some guiding principles to this operational process transformation.

The time is now
The deployment of cloud-native 5G core networks will reinforce the need to look at revamped modern solutions to address the operational challenges while also scoping an architecture with legacy and virtualized components at the core. It is necessary to invest in technology that can move operations closer to a zero-touch network, for quick creation and delivery of innovative services that also leverage edge computing and network slicing.

We are ready to support service providers through this operational transformation journey with our concise portfolio that includes all necessary solutions, as well as services and expertise, to tackle the challenges along the journey to a cloud-native 5G core network. Our solutions offer a smooth evolution alternative while efficiently managing total cost of ownership (TCO) during the migration to 5G.

Do you want to understand more about these operational challenges? Why do they prevent you from reaching your new 5G network’s full potential? What should you consider when defining your operational transformation strategy – and how can Ericsson’s solutions and expertise help you through this process? All of this and more is explained in our latest guide: Transforming operations on the way to 5G.

What Docker Brought to Enterprise Deployment

Docker, introduced in 2013, fundamentally changed how developers and operations teams think about packaging and shipping software. Before containers, applications were tightly coupled to the environment they were built in, making portability across cloud, on-premises, and hybrid environments a constant source of friction.

Gahl Levy, Product Manager at DataStax, responsible for the company’s container and Kubernetes product development, described the shift this way:

“Docker in itself was a game-changer when it was released — it made it simpler and easier to deploy applications in a platform-agnostic way across an enterprise’s infrastructure. Rather than tending to individual server needs, DevOps personnel can focus on containerising their applications. They then know their applications will work in the same manner across cloud, on-premise and hybrid cloud environments.”

Gahl Levy, Product Manager, DataStax

A Docker container bundles all runtime assets — application code, system tools, libraries, and configuration — into a single portable unit. Unlike virtual machines, containers share the host operating system kernel, making them significantly lighter and faster to start. This portability makes them a natural fit for hybrid cloud architectures where workloads span multiple platforms.

CIO Awareness of Docker: Where Things Stand

To understand how CIOs are approaching Docker and container architectures, Silicon spoke with Marco Palladino, Co-founder and CTO of Kong, described as a widely adopted open-source API platform.

“Docker, a technology born in 2013, is now popular enough to be recognised by the large majority of CIOs, especially within the technology, banking and financial sectors. What varies is the depth of expertise and understanding of the technology. However, after more than half a decade, it’s safe to say that the awareness of Docker has been established.”

Marco Palladino, Co-founder and CTO, Kong

Awareness, however, does not always translate into deep adoption. Many organisations are still evaluating how to integrate containerised workloads alongside their existing infrastructure.

How Kubernetes Enables Hybrid Cloud Deployments

Kubernetes extends what Docker started by providing orchestration — automated scheduling, scaling, self-healing, and management of containerised workloads across clusters of machines. Crucially, it abstracts away vendor-specific cloud APIs, making it a strong enabler of hybrid and multi-cloud strategies.

“Kubernetes breaks down the dependency on vendor-specific APIs — like the AWS API, the Azure API and so on — and replaces it with a portable, platform-agnostic API, therefore making the technology a perfect fit in hybrid environments. Kubernetes is by no means a simple technology, and it shouldn’t be seen as a panacea for existing infrastructural or architectural problems. The organisation must invest resources to build a capable R&D team to benefit the most from new, modern tooling.”

Marco Palladino, Co-founder and CTO, Kong

This vendor-neutrality is one of Kubernetes’ most strategically important properties. Organisations that build on Kubernetes gain the flexibility to move workloads between AWS, Azure, Google Cloud, and on-premises data centres without rewriting their deployment configuration.

Key Pressure Points for CIOs Managing Docker at Scale

Adopting containers does not eliminate operational complexity — it shifts it. Palladino identified fragmentation as the primary challenge organisations face when containerising at scale:

“While some workloads may be running well in Docker and modern container orchestration platforms like Kubernetes, it may still be hard to run legacy applications in these environments, effectively creating a two-speed highway within the organisation that can cause fragmentation over time. It’s important that there is a plan in place to reconcile this fragmentation by having a strategy that allows CTOs to connect, secure, monitor and deploy both modern applications and legacy applications across traditional environments and simultaneously on containers.”

Marco Palladino, Co-founder and CTO, Kong

Common pressure points organisations report when scaling Docker deployments include:

• Legacy application compatibility: Not every existing workload is straightforward to containerise, particularly stateful applications and those with complex dependencies.
• Security and secrets management: Containers introduce new attack surfaces that require dedicated tooling and practices.
• Networking complexity: Service-to-service communication across container clusters adds latency and demands careful architecture.
• Vendor lock-in risk: Relying heavily on a specific cloud provider’s managed Kubernetes service can reintroduce the lock-in that containers were meant to avoid.
• Team capability gaps: Kubernetes has a steep learning curve. Organisations that underinvest in training and tooling often struggle to extract its full value.

Are Virtual Machines Obsolete?

A common misconception is that containers replace virtual machines. Palladino clarified the relationship:

“Virtual machines are still required to run Docker containers because that’s where container orchestration platforms like Kubernetes ultimately run. I wouldn’t consider Docker a replacement for VMs. Docker is an additional abstraction layer that sits above virtual machines and allows them to more efficiently manage resources while providing a standardised distribution for any workload across any cloud and data centre. Because of this, I see Docker as the future of the hybrid cloud.”

Marco Palladino, Co-founder and CTO, Kong

In practice, Kubernetes cluster nodes are typically VMs (or bare-metal servers). Docker adds a layer on top, enabling fine-grained workload isolation and portability without replacing the underlying compute infrastructure. The two technologies are complementary, not competing.

Docker, Kubernetes, and Microservices Architecture

One of the most significant architectural impacts of Docker and Kubernetes is the acceleration of microservices adoption. By making it straightforward to package, deploy, and independently scale individual services, containers lower the operational barrier to breaking apart monolithic applications.

“By allowing applications to be quickly packaged, distributed and deployed across hybrid and multi-cloud environments, Docker and Kubernetes make it easier to build decoupled and distributed applications that can run anywhere. However, as we decouple our systems into microservices, we enable more API communication over a network across our applications. Performance and security become even more critical to encrypt, secure and accelerate the flow of information across our services.”

Marco Palladino, Co-founder and CTO, Kong

Palladino also warned against depending too heavily on cloud-native vendor tooling for security and traffic management within containerised environments:

“While Docker and Kubernetes allow us to deploy our workloads across multiple clouds and platforms, it’s important that we do not rely too much on cloud vendor solutions to secure and monitor our traffic. In the long term, that creates inconsistencies and fragmentation. Instead, we want to adopt platform-agnostic technology that can be deployed within our Kubernetes cluster that consolidates how we manage and operate our services across every environment, across every cloud.”

Marco Palladino, Co-founder and CTO, Kong

A New DevOps Development Model

For DevOps teams, Docker and Kubernetes are not just deployment technologies — they represent a shift in how software is conceived and built. The container model encourages platform-agnostic, distributed thinking from the earliest stages of development.

“Docker and its ecosystem provide a platform-agnostic, distributed way of thinking about our applications. By doing so, they make it easier to build decoupled systems that move away from legacy monolithic architectures. While adopting modern architectures will increase team productivity by lowering the friction to create, package and distribute software, it’s essential also to address new challenges — including data security, reducing performance latency, and being comfortable running large-scale, decoupled and distributed systems across multiple platforms.”

Marco Palladino, Co-founder and CTO, Kong

The Future of Containerised Applications

Looking ahead, containers are expected to become smaller, more standardised, and more deeply integrated with emerging paradigms such as serverless computing and Functions-as-a-Service (FaaS).

“Containers will get smaller and nimbler. There will be more of a focus on standardised base images. These can make containers lighter weight and allow underlying operating systems to focus on optimisations for those base image types.”

Lee Atchison, Senior Director, Cloud Architecture, New Relic

Palladino offered a longer-term perspective on where containerisation is heading:

“We will be able to take full advantage of containerised applications when we also provide an enabling path for legacy application and brownfield use cases to transition to these modern architectures. Containerised applications will become the foundation — the primitive — that we work with to build more complex, self-healing systems over time. Serverless and FaaS are also interesting technology trends because they paint a vision of the future where we will not need containers at all.”

Marco Palladino, Co-founder and CTO, Kong

Scott McCarty, Principal Product Manager for Containers at Red Hat, noted the trajectory of enterprise adoption:

“We’ve seen a real expansion in awareness around containers since 2014. The adoption of containers was a bottom-up movement, similar to Linux. It started with pioneering developers and systems administrators looking for better management of deployments. In the last 18–24 months, it has become the strategic direction for large institutions, and many more CIOs are paying attention.”

Scott McCarty, Principal Product Manager, Containers, Red Hat

Key Takeaways for DevOps and Engineering Leaders

• Containers complement VMs — they do not replace them. Kubernetes runs on top of virtual machines or bare-metal nodes. Plan your infrastructure accordingly.
• Kubernetes abstracts cloud vendor APIs, making it a strong foundation for genuine hybrid and multi-cloud portability.
• Fragmentation between legacy and containerised workloads is the most common operational pain point at enterprise scale. A deliberate migration strategy is essential.
• Microservices bring new API and security challenges. Invest in platform-agnostic tooling for traffic management, observability, and security rather than relying solely on cloud-provider native solutions.
• Team investment is non-negotiable. Kubernetes is complex. Organisations that build strong internal R&D capability extract significantly more value from the ecosystem.
• Serverless and FaaS are the next frontier, potentially abstracting containers away from developers entirely in specific workload categories.

Docker and Kubernetes are no longer emerging technologies — they are the operational backbone of modern application deployment. For CIOs and CTOs navigating hybrid cloud strategy, understanding how to leverage them effectively, while managing the complexity they introduce, is now a core competency.