DEV Community: Richard Foster

Cisco CCNA in 2026: Is It Still Worth It When AI Is Changing Networking?

Richard Foster — Thu, 30 Apr 2026 16:33:15 +0000

Every few years, a technology shift arrives big enough to make people question whether established credentials still mean anything. We saw it when virtualization disrupted traditional data centers. We saw it again when cloud computing started eating enterprise infrastructure. Now, it's AI's turn to spark the debate and this time, the question is pointed squarely at the CCNA.
I've been around long enough to remember when people asked whether the CCNA was still relevant after SDN came along. Spoiler: it was. But the AI question deserves a more careful look, because what's happening right now inside both the networking industry and Cisco's certification program is genuinely significant.

What's Actually Changed in 2026

Let's start with facts, not opinions.
On February 3, 2026, Cisco executed its most substantial overhaul of the certification ecosystem in nearly a decade. The core CCNA exam still coded 200-301 kept its structure, but the syllabus was meaningfully upgraded to integrate AI, cloud, automation, and security content far more deeply than before.
Here's what actually changed inside the exam:
• AI and ML fundamentals were formally added, covering AI-driven network optimization, intelligent Wi-Fi channel adjustment, and AIOps basics like predictive maintenance and automated fault diagnosis
• Machine learning-based intrusion detection and AI-powered threat prevention concepts are now testable topics
• Automation frameworks including NETCONF/YANG, Ansible, and Terraform are part of the curriculum, along with basic Python scripting for network tasks
• REST API interaction with network devices has become a core competency, not an optional extra
• The old DevNet certification track was retired and fully rebranded as CCNA Automation, CCNP Automation, and CCIE Automation reflecting that automation is no longer a specialty skill but a baseline expectation
What Cisco is saying with these changes is pretty clear: the networking professional of 2026 is expected to understand not just how to configure a router, but how intelligent systems are increasingly making those decisions alongside them.

The AI Threat Narrative And Why It's Overstated

There's a version of this conversation that goes: AI is automating network configuration, troubleshooting, and log analysis, so network engineers will eventually be unnecessary. It makes for compelling headlines. It's also not an accurate picture of what's actually happening in enterprise environments.
AI in networking right now is excellent at pattern recognition and automation of repetitive, well-defined tasks. It can flag anomalies in traffic, suggest configuration changes, and optimize SD-WAN routing policies. What it doesn't do is replace the human judgment required when something genuinely unexpected happens a complex multi-vendor environment behaving erratically, a security incident unfolding in real time, or a major network redesign tied to a business acquisition.
The more honest framing is this: AI changes what network engineers spend their time on, not whether they're needed. And that actually strengthens the case for formal credentials, because knowing how these AI-assisted tools work and when not to trust them requires the kind of foundational knowledge that the CCNA still tests.
According to CompTIA's 2026 State of the Tech Workforce report, demand for certified network engineers in the United States grew 11% year-over-year in Q1 2026, with Cisco credentials ranking among the top five most requested certifications by enterprise employers. That's not the trajectory of a credential losing relevance.

What the Job Market Is Actually Saying

Numbers matter here more than theory.
The average salary for a CCNA-certified Network Engineer in the United States sits at $109,040 per year as of April 2026, according to ZipRecruiter data. The range is wide entry-level roles start around $50,000, while senior professionals with automation specializations are clearing $140,000 and above.
The Bureau of Labor Statistics projects 12% growth for network engineers between 2024 and 2034, significantly outpacing the average for all occupations. Cloud migration, hybrid infrastructure, and enterprise security demands are the primary drivers of that growth.
Perhaps the most telling figure: employer surveys consistently show that CCNA-certified candidates are hired 40-60% faster than non-certified peers with comparable experience. In a job market where speed of placement matters to both employers and candidates, that's a meaningful advantage.
Gartner has reported that 87% of IT leaders struggle to find qualified networking talent. The skills shortage isn't going away if anything, it's getting more acute as the technology becomes more complex. Certifications like the CCNA serve as one of the most reliable trust signals in a market where that trust is genuinely hard to establish.

What CCNA Actually Covers Now A Realistic Breakdown

For anyone who hasn't looked at the exam blueprint recently, it's worth understanding where the weight sits across the six core domains:
• Network Fundamentals (20%) OSI model, TCP/IP, IPv4/IPv6, VLANs, and subnetting
• Network Access (20%) Switching, STP protection mechanisms (Root Guard, Loop Guard, BPDU Guard), and wireless configurations
• IP Connectivity (25%) Routing protocols and IP services. This carries the most weight and is where most candidates pass or fail.
• IP Services (10%) DHCP, NAT, NTP, QoS fundamentals
• Security Fundamentals (15%) Access control, VPN concepts, threat defense basics
• Automation and Programmability (10%) Python basics, REST APIs, Ansible, and now the AI/ML networking concepts
The automation and AI section at 10% is meaningful but not overwhelming. Cisco's own guidance on the exam is explicit: candidates don't need to understand AI algorithms in depth, but they do need to understand where and why AI is applied in modern network environments. That's a reasonable bar.

Career Paths That Open With CCNA in 2026

One of the strongest arguments for the CCNA right now and this has shifted in the last two years is how naturally it connects into the fastest-growing adjacent fields.
The credential is increasingly viewed as a bridge, not just an endpoint. After CCNA, the most common and lucrative progressions are:
• CCNP Enterprise for senior network engineering and architecture roles
• CCNP Security / CCNA Cybersecurity for those moving into security operations, now a renamed track under Cisco's 2026 restructure
• CCNA/CCNP Automation the newly launched track that replaced DevNet, focusing on AI-integrated automation infrastructure
• Cloud certifications (AWS/Azure) for network engineers pivoting into cloud networking roles
The fact that Cisco explicitly built the new Automation track on top of the CCNA framework isn't accidental. It signals that the associate-level credential is meant to be the foundation, not the ceiling and that the ceiling now extends clearly into AI-driven infrastructure territory.

Preparing Smart: What Actually Works in 2026

The exam has evolved, and preparation strategies need to reflect that. The days of pure memorization-based CCNA prep are over. Cisco's updated blueprint explicitly de-emphasizes theoretical recall and emphasizes hands-on configuration and practical troubleshooting.
Tools that matter for preparation right now:
• Cisco Packet Tracer still the most accessible simulation tool, and free
• EVE-NG or GNS3 for more complex multi-device lab scenarios
• Cisco Modeling Labs (CML) the premium option for realistic enterprise simulations
• Postman for practicing REST API interactions, which are now testable
• Python basic scripting knowledge, focused on network automation use cases
Many candidates also use structured question banks and practice exams to identify gaps before sitting the actual test. If you're looking for a starting point to benchmark your readiness, reviewing Cisco exam practice questions and study guides from a dedicated certification resource is a practical first step before investing in full training programs.
The exam runs 120 minutes, covers approximately 90-100 questions, and requires a passing score of 825 out of 1000. The question types have expanded to include simulation scenarios not just multiple choice which is another reason hands-on lab practice matters more than it used to.

Reviewing the Supervisor / Namespace / Workload split while preparing for VMware 3V0-24.25

Richard Foster — Tue, 07 Apr 2026 17:19:20 +0000

TL;DR
• The hardest part of 3V0-24.25, at least for me, was not memorizing commands. It was understanding which layer owns the decision.
• Once I stopped treating VMware vSphere Kubernetes Service like “just Kubernetes on top of vSphere,” the blueprint started to make more sense.
• The mental model that helped me most was to separate things into Supervisor, namespace, and workload cluster responsibilities.
I’ve been spending more time around VMware Cloud Foundation and Kubernetes lately, and although I’m comfortable enough with vSphere, I didn’t want to pretend I already had deep intuition for the cloud-native side of the stack.
So I set myself a target: get properly prepared for Advanced VMware Cloud Foundation 9.0 vSphere Kubernetes Service (3V0-24.25).
Broadcom describes this certification as validating professionals to deploy, operate, and secure vSphere Kubernetes Service environments inside VMware Cloud Foundation. The official exam guide also says the exam is 60 items, 135 minutes, delivered through Pearson VUE, with a scaled passing score of 300, and that the minimally qualified candidate typically has 6–12 months of hands-on experience with VKS components.
To avoid studying blindly, I started with the official 3V0-24.25 exam guide instead of jumping straight into random questions. That was the point where the scope of the exam became much clearer to me.
The blueprint is much broader than “know Kubernetes basics.” It explicitly touches things like differentiating VMs vs containers, using Kubernetes architecture, networking, storage, service mesh, and Helm, understanding Supervisor capabilities and topologies, configuring networking with VDS / NSX Segments / VPCs, working with storage policies and persistent volumes, managing identity and access, handling Kubernetes releases and content libraries, provisioning and scaling clusters, upgrading them, managing snapshots, and even implementing backup and restore with Velero.
At some point during prep, I started checking practice material too. One of the pages I came across was the CertsWarrior page for 3V0-24.25 exam dumps. I didn’t really use it as a replacement for the official blueprint. What it helped me do was spot which topics I was still mentally mixing together. CertsWarrior’s page currently lists 61 questions, 135 minutes, a 60% passing mark, and an update date of April 6, 2026, so it worked for me as a quick checkpoint while reviewing objectives.

The mistake I kept making

Early on, I kept treating every problem like it lived inside a Kubernetes cluster.
If I saw a question about RBAC, storage, scaling, registries, YAML, or deployment, my brain immediately went to standard Kubernetes objects and commands. That wasn’t always wrong, but it was incomplete.
The official blueprint kept pulling me back up a level.
Some objectives are clearly about the platform layer: enabling and designing the Supervisor, choosing the right networking model, thinking about load balancer sizing, configuring storage policies, or understanding namespace architecture and ingress / egress options. Other objectives are more about the cluster lifecycle layer: releases, content libraries, rolling updates, autoscalers, snapshots, packages, private registries, and backup / restore. Then there’s the actual application workload layer, where ordinary Kubernetes thinking becomes more relevant.
That was the shift that helped me most.

The mental model that finally clicked

I started organizing my notes into three buckets:

1) Supervisor layer

This is where I put things like:
• enabling platform capabilities
• architecture and topology decisions
• networking choices
• storage policy integration
• zones and namespaces
• load balancing implications
That bucket came directly from the exam objectives around Supervisor capabilities, networking, storage integration across zones, namespace design, and reference architecture.

2) VKS cluster lifecycle layer

This is where I grouped:
• cluster creation
• release management
• content libraries
• upgrades
• scaling
• autoscalers
• add-on services
• snapshots
• backup / restore
• registry configuration
This bucket made sense because Broadcom’s own VKS material describes the service as providing the components for provisioning and managing the lifecycle of Kubernetes clusters in the vSphere Supervisor environment.

3) Workload layer

This is where the more familiar Kubernetes concepts came back into focus:
• Pods and Deployments
• persistent storage
• ingress behavior
• RBAC
• autoscaling
• YAML changes
The official Kubernetes docs were useful here because they helped me separate the generic Kubernetes concept from the VMware-specific implementation detail. Kubernetes’ RBAC docs describe RBAC as role-based control over access to API resources; the storage docs explain PersistentVolumes and StorageClasses; and the autoscaling docs explain how HorizontalPodAutoscaler adjusts workload capacity based on observed utilization. Those concepts absolutely matter, but in this exam they are wrapped inside the VKS / Supervisor operating model.

What changed once I studied this way

Once I started asking myself, “Which layer owns this?”, I stopped getting trapped by surface familiarity.
Before that, a question about storage would make me think only about PVCs and StorageClasses. A question about access would make me think only about Kubernetes RBAC. A question about scaling would make me think only about HPA.
But 3V0-24.25 keeps pushing you to think more like an operator and architect inside the VMware Cloud Foundation stack. That means understanding where the Supervisor ends, where namespaces come into play, and where standard cluster-level Kubernetes administration begins. That framing is all over the official exam objectives.

The resources I found most useful

I kept coming back to four kinds of resources:
• the official Broadcom certification page
• the official 3V0-24.25 exam guide PDF
• the official vSphere Kubernetes Service architecture docs and release notes
• and, for quick self-checking, the CertsWarrior 3V0-24.25 dumps
Broadcom’s public materials define the certification scope and exam structure, the TechDocs material reinforces how VKS fits with Supervisor architecture and lifecycle management, and the CertsWarrior page was useful to me as a lightweight way to pressure-test weak spots after reviewing the official objectives.

How I prepared for Oracle Project Management Cloud 2025 Implementation Professional - Delta

Richard Foster — Tue, 07 Apr 2026 16:19:26 +0000

TL;DR
Already having some Oracle Cloud implementation exposure helps, but what mattered most for me was reviewing the release-specific changes, going back through Oracle’s own implementation flow, and then using a couple of practice resources to pressure-test weak areas. What finally made the difference was not memorizing screens, but understanding how Oracle expects implementation projects, setup tasks, and quarterly-update knowledge to connect. Oracle’s official exam page says this Delta exam is 60 minutes, has a 68% passing score, and retires on May 29, 2026.

Story time

I’ve worked around Oracle Cloud long enough to know that Delta exams can be deceptively tricky.
On paper, they look smaller than a full implementation exam, so it’s easy to think, “I already know this product, I’ll just skim a bit and be fine.” That was exactly the trap I wanted to avoid with Oracle Project Management Cloud 2025 Implementation Professional - Delta (1D0-1057-25-D).
This certification is positioned by Oracle for ERP Project Management Cloud Implementation Consultants, IT Admins, and Functional Leads, and the free Delta path is meant for people who already hold the earlier Oracle Project Management Cloud Implementation credential.
That framing actually helped me a lot. It reminded me that this is not really a beginner exam. It is more like Oracle checking whether you stayed current with the quarterly changes and still understand the implementation side properly. Oracle’s own exam page says the exam has been validated against 25A, 25B, 25C, and 25D, which immediately told me my prep needed to be update-aware, not just product-aware.
The first thing I did was stop guessing and go straight to Oracle’s official exam page. That gave me the practical details I wanted up front: 60 minutes, 68% passing score, and a retirement date of May 29, 2026. Once I had that, I knew this was the kind of exam where I needed focused revision, not endless wandering through docs.
After that, I went back into Oracle documentation, especially the implementation flow. One Oracle Help page on implementation projects was surprisingly useful because it reminded me how Oracle wants setup to be approached in a structured way: configure the offerings, opt into the needed functional areas and features, then generate the task list for the implementation project. That sounds basic, but revisiting that workflow helped re-center my thinking around implementation logic rather than just feature trivia.
I also spent time checking Oracle’s Project Management release materials. Oracle’s readiness center shows separate “What’s New” pages for 25A, 25B, 25C, and 25D, which matched the way I was trying to study anyway: not as one giant product recap, but as a year’s worth of changes and refinements that could realistically show up in a Delta exam.

What I focused on

The mistake I nearly made was treating this like a random multiple-choice exam.
What helped me more was splitting prep into three layers:

1. Core implementation flow

I reviewed how Oracle structures implementation projects, offerings, and functional areas, because that is the backbone for making sense of setup questions.

2. Product implementation knowledge

I spent time in Oracle Project Management implementation documentation and the getting-started material to refresh the bigger picture. Oracle’s Project Management implementation guide is extensive, but even scanning the structure helped me remember how interconnected setup decisions are across project financial management and related areas.

3. Quarterly-update awareness

Because this is a Delta exam, I paid attention to release-specific changes across 25A–25D instead of assuming my older notes were enough. Oracle’s readiness pages made that part much easier to organize.

Practice resources

I did not want to rely on only one source, so I used three different types of prep material.

1. CertsWarrior

I reviewed the CertsWarrior page for 1D0-1057-25-D as one of my practice checkpoints. On that page, CertsWarrior lists the exam as 60 questions, 68% passing marks, 60 minutes, with content updated on April 3, 2026, so I used it mostly to test whether I was still rusty on certain topics after going through Oracle material.

2. Practice Test Software

I also looked at the Practice Test Software page for 1D0-1057-25-D as another practice source. I liked using a second prep page simply because it forced me to compare how I was answering, instead of becoming too comfortable with one format.

3. Oracle official resource

My anchor resource was still Oracle’s official exam page for Oracle Project Management Cloud 2025 Implementation Professional - Delta. That was the page I trusted for the actual exam facts and scope.

What actually helped me most

Oddly enough, the biggest improvement in my prep came when I stopped asking, “Do I remember this screen?” and started asking, “Why would Oracle want this configured that way?”
That one shift made a lot of questions easier.
Instead of trying to remember isolated details, I started thinking in terms of implementation decisions:
• what gets enabled first
• what belongs inside the implementation project flow
• how functional areas and setup tasks are organized
• what changed recently enough to matter for a Delta exam
That approach felt much closer to real implementation work anyway.

The exam side of things

Because it’s a Delta exam, I personally found the time pressure more noticeable than the content volume. Sixty minutes is enough if the material is fresh, but not enough if you are still hesitating between two similar answers. That is why the practice phase mattered for me. Oracle’s published details confirm the time limit is 60 minutes and the passing score is 68%.
Another thing worth noting is eligibility. Oracle says the free Delta route is for people who have already earned the earlier Oracle Project Management Cloud Certified Implementation Professional or Specialist credential. That made me treat the exam like an update-and-validation exercise, not a first exposure to the product.