| |
| ℹ️ News, Updates & Announcements |
| |
|
| |
| Microsoft Defender Now Blocks Pod Privilege Escalation |
| |
| |
Microsoft Defender for Cloud just dropped a preview of Restrict Pod Access, a new move to box in lateral movement and privilege creeps inside Kubernetes. Containers get chattier, it gets nosier.
Container image re-scans now hit every 30 days, but only for images currently in use. Less drift, more signal.
Support for scanning images from JFrog Artifactory and Docker Hub is out of preview and into GA. |
|
| |
|
| |
|
| |
| 2025's Cloud Native Reality Check: Who's In, Who's Lagging |
| |
| |
56% of backend devs now count as cloud native. That rise tracks with heavy use of API gateways (50%) and microservices (46%). Only 30% touch Kubernetes directly, but hybrid (30%) and multi-cloud (23%) setups are gaining ground. The shift? Tighter security and chunkier, modular infra.
System shift: Cloud native isn’t all about Kubernetes anymore. It's leaning into internal platforms and MLaaS layers that spare developers from wrestling with bare-metal config. |
|
| |
|
| |
|
| |
| Streamline Complex AI Inference on Kubernetes with NVIDIA Grove |
| |
| |
NVIDIA released Grove, a Kubernetes API baked into Dynamo, to wrangle the chaos of modern AI inference. It pulls apart your big, messy model into clean, discrete chunks - prefill, decode, routing - and runs them like a single, orchestrated act.
The trick? Custom hierarchical resources. They let Grove handle startup order, gang scheduling, topology-aware placement, and multilevel autoscaling without breaking a sweat.
Why this matters: Grove turns AI inference into something Kubernetes can actually understand, declarative and dependency-aware. This is scheduling for large, multi-role models that live in the real world. |
|
| |
|
| |
|
| |
| KServe becomes a CNCF incubating project |
| |
| |
KServe is upgrading. The CNCF pulled it into incubation, backing it as the Kubernetes-native way to serve both generative and predictive AI. Translation: it’s not a side project anymore - it’s core infra.
Version 0.15 steps up with tighter integrations across the stack: vLLM, Envoy Gateway, llm-d, Knative, and Istio. The goal? Fully abstracted, scalable LLM inference. No hand-wiring. No mess.
Big picture: KServe’s new badge signals a shift toward standardized, production-ready orchestration in Kubernetes-first AI pipelines. |
|
| |
|
| |
|
| |
| Amazon Elastic Kubernetes Service gets independent affirmation of its zero operator access design |
| |
| |
Amazon EKS just went full Fort Knox. It now runs on a zero operator access model - meaning even AWS can’t peek inside your Kubernetes control or data plane.
The setup leans on the Nitro System’s confidential compute, guarded APIs, and multi-party approval pipelines. NCC Group also kicked the tires and gave it the all-clear: no hidden backdoors for AWS staff. |
|
| |
|
| |
|
| |
| CNCF Launches Certified Kubernetes AI Conformance Program at KubeCon |
| |
| |
CNCF just kicked off the Certified Kubernetes AI Conformance Program (beta). Think of it as a litmus test for running AI workloads on Kubernetes without duct tape and hope.
The spec lays down a reference architecture, GPU and networking test criteria, and an annual renewal loop. Full automation is on deck by v2.0 in 2026.
Big picture: Kubernetes is moving from "it runs AI if you squint hard enough" to a legit standard for portable, production-grade AI/ML workloads. Less chaos. More click-and-go. |
|
| |
|
| |
|
| |
| AWS Backup now supports Amazon EKS |
| |
| |
AWS Backup just added support for Amazon EKS. Now you can back up cluster state and persistent volumes, no agents, no third-party hacks.
It handles scheduling, retention, and immutability out of the box. Restore full clusters or drill down to specific components, even across Regions and accounts. |
|
| |
|
| |
| 👉 Enjoyed this?Read more news on FAUN.dev/news |
