This Week in DevOpsCube ☕
What Latest in DevOps?
DevOps ChangeLog: Releases & Breaking Changes
Zero-Downtime Kubernetes Ingress Controllers on GCP
Learn about SLSA, a must-know security concept for DevOps Engineers
A Look into Cloudflare’s MLOps Platform
DevOps Remote Jobs
and more..
☕ Grab a coffee and catch up on this week’s DevOps, MLOps, and AI insights and resources.
📢 Latest in DevOps
AWS launched FinOps Agent in public preview: An AI agent that auto-investigates cost irregularities and answers cost questions in Slack/Jira, and turns optimization recommendations into tickets.
GitHub Actions Steps Can Now Run in Parallel: GitHub Actions now supports running independent steps in parallel, helping reduce workflow execution time and speed up CI/CD pipelines.
Grafana Assistant Investigations: Grafana introduced an AI-powered assistant that automatically investigates incidents, identifies root causes, and helps teams remediate issues faster.
AWS Lambda Introduces MicroVMs: AWS launched Lambda MicroVMs, enabling developers to run AI-generated or untrusted code in isolated, stateful Firecracker-powered sandboxes without managing infrastructure.
Chainguard Launches Athena: A tool that helps fix open-source security vulnerabilities before attackers can exploit them.
🛡️ SLSA Explained For DevOps Engineers
In this guide, you will learn about.
- What is SLSA provenance & SLSA trust levels
- How to implement SLSA in your CI/CD pipeline
- Verifying provenance in Kubernetes with admission policies
- Three supply chain incidents every DevOps engineer should know.
👉 Read it Here: SLSA Explained
📖 Worth Reading This Week
GitOps with 15,000 Kubernetes Clusters: Learn the scalability limits of GitOps by testing Argo CD and vCluster across 15,000 virtual Kubernetes clusters, uncovering key bottlenecks and architectural trade-offs.
Zero-Downtime Kubernetes Ingress on GKE: Learn why Kubernetes Ingress can still experience downtime during pod termination on GKE and how to configure graceful shutdowns to achieve true zero-downtime deployments.
Cloud-Native is the Foundation for Agentic AI: Learn how Kubernetes and cloud-native technologies provide the security, scalability, and observability needed to run production-ready AI agents.
Zero-Trust AI Agents on Kubernetes: Red Hat explains how to securely run AI agents on Kubernetes using workload identities, mutual TLS (mTLS), and Zero Trust principles instead of static API keys.
🤖 MLOps Engineers Generally Earn More Than Traditional DevOps Engineers
We looked at 200+ MLOps job listings from platforms like LinkedIn, Indeed, and others.
Here is what we found:
DevOps fundamentals like Docker, Kubernetes, and CI/CD tools are must-have skills.
90% of the jobs mentioned Python as a key skill.
25% of the jobs mentioned Go, mainly for platform-building roles.
MLflow, Kubeflow, and Airflow were the most common MLOps tools highlighted for model tracking, pipelines, and orchestration.
Monitoring and observability tools like Prometheus and Grafana were also mentioned in several postings.
There were also references to Vector Databases, RAG, LLMs, and other GenAI-related technologies.
We have a dedicated MLOps repo where you can learn every MLOps concept from scratch using a project.
👉Start Here: MLOps for DevOps Engineers
⚙️ SLSA Provenance Generation Using GitHub Actions
In this guide, you will learn about:
- Generating SLSA provenance with GitHub Actions
- Verifying a provenance file
- What happens when verification fails
👉 Read it Here: SLSA Provenance Creation using GitHub Actions
📋 Changelog: Releases & Breaking Changes
Helm 3 Reaches End of Life: Helm announced that Helm 3 will reach end of life, with the final feature release in September 2026 and security patches ending in February 2027, encouraging users to migrate to Helm 4.
GitHub Enterprise Server 3.21 is Generally Available: GitHub Enterprise Server with improved GitHub Actions, stronger security controls, better policy management, and enhanced monitoring for self-hosted environments.
OpenTofu v1.12.0 is out: OpenTofu added a feature that stops resources from being deleted by accident and now lets teams control this with a simple setting instead of hardcoding it.
Kubernetes 1.33 Is No Longer Supported: As of June 28, 2026, Kubernetes 1.33 stopped getting security fixes. If your cluster is still on this version, it's time to upgrade to 1.34 or newer.
Docker Content Trust Retiring: Docker's image signing service is going to be stopped on December 8, 2026. If you are using this service, its time to move on.
🛠️ DevOps Tool of the Week (Kagent)
Running agents like any other Kubernetes workload is the hard part. That is where kagent helps.
It is an open-source framework that lets you build and run AI agents as native Kubernetes resources.
Here is what it does 👇
- Define agents as Kubernetes custom resources and manage them with GitOps.
- Connect to OpenAI, Anthropic, Vertex AI, Azure OpenAI, or local Ollama models.
- Give agents access to tools through MCP servers.
- Chain multiple agents together to investigate and resolve problems.
- Track prompts, tool calls, and token usage with OpenTelemetry and Prometheus.
👉 Start Here: Kagent
🛠️ A Look into Cloudflare’s MLOps Platform
One of the best ways to learn MLOps is by studying how companies build real-world ML platforms.
Cloudflare shared how they built their internal MLOps platform using Kubernetes, GitOps, and open-source tools like Kubeflow, deployKF, Airflow, and MLflow.
The interesting part for DevOps engineers is that the foundation is built using familiar tools:
Kubernetes for running workloads.
Argo CD for GitOps-based deployments.
Helm and Kustomize for managing configurations.
JupyterHub for providing development environments for ML teams.
MLOps introduces new workflows, but the infrastructure foundation is still cloud-native.
👉 Read it Here: Inside Cloudflare's MLOps Platform
🎓 Complete Kubernetes & CKA Course
10,000+ engineers have learned through DevOpsCube courses.
From container fundamentals to CKA preparation, every course is self-paced and built around real scenarios.
This is not a long video lecture series.
The CKA course is text-based, illustration-rich, and designed for faster learning and quick revision whenever you need it.
👉 Coupon: Use code FLASH40 to get 40% OFF today.
Note: This is a community only offer! Make use of it!
🛡️ 15 Factor Apps
Security is usually the last thing discussed in DevOps, but in cloud-native, authentication has to be a core design principle.
Kubernetes is a perfect real-world example for this.
- Every Pod gets a JWT (service account token).
- When the Pod talks to the Kubernetes API, it attaches this token.
- The API server checks the token and applies RBAC rules to decide what that Pod is allowed to do.
This is 15-Factor security in practice. If you want to know more about 15-Factor Apps,
👉 Read it Here: 15 Factor Apps
💼 DevOps Remote Jobs
Techvantage.ai - DevOps Engineer - Azure (3-4 Yrs)
EverestEngineering - Senior DevOps Engineer (5-9 Yrs)
Branch International - DevOps Engineer (7 Yrs)
InOrg Global - DevOps Architect
NTT DATA - Senior DevOps Engineer
Zenara Health - DevOps & Security Engineer (5-10 Yrs)
Red Hat - Site Reliability Engineer
Multinational Cloud & AI Firm - GCP SRE / DevOps Engineer (5-10 Yrs)
Drivetrain - Site Reliability Engineer (5+ Yrs)
Jinendra Infotech - AWS DevOps - SRE (5-7 Yrs)
Join 20K+ DevOps & MLOps Community
20,000+ trusted Engineers read our newsletter, where we simplify DevOps, MLOps and Kubernetes into practical, easy-to-follow guides.
DevOpsCube team:
Lead Editor: Bibin Wilson / Designer: Aljin / Tech Researcher: Aswin
Interested in advertising with us? Get in touch
Need DevOps consulting? Contact us
Want to learn from courses? Check our courses
Forwarded this email? Join our mailing list
