Nov 28, Paris: AI Health Hack #2 with Alan, Qovery and AWS. Applications close Nov 21
Topic · Security

Infrastructure Security

Infrastructure security is the set of controls that protect the platform your applications run on, including patching, secrets management, policy enforcement, access control and audit logging.

Most infrastructure incidents start with something ordinary: an add-on nobody patched, a certificate authority that loses trust, a secret pasted into a variable, a token with more access than the job needed. The fix is usually known. The hard part is finding every place it applies across clusters and accounts, quickly.

Runbooks help with the response. Policy as code, scoped credentials, short-lived tokens and an audit log on every change help prevent the next one, and they matter more now that coding agents call infrastructure APIs directly.

These guides cover recent responses such as the ingress-nginx retirement, NVIDIA NIM security releases and the eMudhra CA key compromise, alongside secrets, SBOMs for the EU Cyber Resilience Act, OPA and Kyverno, RBAC and agent identity. Qovery keeps workloads and secrets in your own cloud account and applies RBAC, OPA-based policy tokens and audit logs to every operation.

In this topic
Start here

ingress-nginx Is Retired and a New CVE Just Landed: The 24-Hour Mitigation Runbook and What to Run Instead

ingress-nginx reached end of maintenance in 2026, so any CVE filed after that date - including CVE-2026-116472 - will never get an official upstream patch. Here is the four-step containment runbook you can apply in 24 hours, a fair comparison of the maintained ingress controllers and Gateway API implementations that should replace it, and a hostname-by-hostname migration plan with rollback.

Nov 8, 2026 · 13 min

NVIDIA NIM 2.1.5 Fixed 81 Vulnerabilities, and 2.1.6 Fixed 99 More: A Safe Upgrade Runbook for Self-Hosted Models on Kubernetes GPU Nodes

NIM 2.1.5 fixed 81 unique security issues (3 critical) across NVIDIA's vLLM and SGLang images, and 2.1.6 fixed 99 more a week later. Here is the eight-step runbook I use to upgrade self-hosted NIM containers on Kubernetes GPU nodes: inventory, bulletin triage, scanning, digest pinning, config diffs, cache warm-up, a one-pool canary and rollback.

Nov 18, 2026

After the eMudhra CA Key Compromise: How to Audit and Rotate Certificate Chains Across Every Kubernetes Cluster

Eight eMudhra intermediate CA keys were recoverable from their own certificates. Here is what was publicly reported, what it means for cert-manager, trust-manager and Crossplane, and a fleet-wide runbook with kubectl, cmctl and openssl commands to inventory, rotate and verify certificate chains.

Nov 14, 2026

A CVSS 10.0 in NVIDIA AIStore's PodSpec API: how to check if you are exposed in 15 minutes

NVIDIA disclosed a critical flaw (reported as CVE-2026-65162, CVSS 10.0) in AIStore's Kubernetes PodSpec API. Here is the bug class in plain terms, five kubectl commands that tell you whether you are exposed, and the remediation order that removes the attack path before the patch even lands.

Oct 27, 2026

What the EU Cyber Resilience Act's SBOM Requirement Actually Means If You Ship Containers to the Cloud

The EU Cyber Resilience Act requires an SBOM for every product with digital elements placed on the EU market, with full application from 11 December 2027. Here is what that means for teams shipping container images on AWS, GCP, Azure, Scaleway, or their own Kubernetes cluster: the exact legal text, the deadlines, the formats, and what the shared responsibility model leaves on your side.

Nov 9, 2026

How to Stop Secrets Leaking From Your Deployment Platform, and How Qovery Stores Yours

Railway's disclosure of leaked customer environment variables is a reminder that deployment platforms hold the keys to everything. Here is how secrets leak from platforms, the practices that prevent it or limit the damage, and exactly how Qovery stores secrets across its control plane and your data plane, with AWS Secrets Manager, Parameter Store, GCP Secret Manager, Doppler and Infisical.

Oct 22, 2026

Secret Manager Integration: One Source of Truth for Humans and Agents.

Production secrets should live in one place and stay there, whether your next deployment is triggered by a developer or an AI agent. The Secret Manager integration connects AWS Secrets Manager, AWS SSM, or GCP Secret Manager to Qovery so secrets are referenced, never copied, and enterprise governance holds regardless of who deploys.

Jun 10, 2026

Northflank Security Incident (October 30, 2026): What Happened and How to Protect Your Secrets

On October 30, 2026, Northflank told all users to assume their environment variables were decrypted and to rotate every secret. Here is the verified timeline, what to rotate first, why shared platforms keep getting hit after Railway, and how Qovery limits the blast radius, including where we still hold secrets.

Oct 30, 2026

OPA Gatekeeper vs Kyverno vs a Managed Platform Layer: Where Should Deployment Policy Live?

OPA Gatekeeper and Kyverno enforce deployment policy at the Kubernetes admission controller, after a bad config has already been written. A managed platform layer enforces it before the config exists. Here is how the two layers compare, where Crossplane, Backstage, Port and Argo CD actually fit, and which one to start with.

Oct 9, 2026

The 8 Policy-as-Code Tools I'd Actually Use to Control What Gets Deployed to Your Cloud Accounts

A practical, side-by-side comparison of the policy-as-code tools that actually stop bad deployments - OPA/Gatekeeper, Conftest, Kyverno, Checkov, Trivy, Sentinel, Pulumi CrossGuard, Spacelift and Cloud Custodian - and exactly which enforcement point each one belongs at.

Sep 25, 2026

Full Control Over Permissions, Welcome API Policy Tokens

API Policy Tokens let you control what a Qovery token can do action by action, using OPA and Rego, so an agent or script never gets more access than the one job it was built for.

Aug 18, 2026

GitOps at Scale: How to Enforce Consistent Sync Policies, RBAC, and Audit Trails Across Multiple Kubernetes Clusters

A practical comparison of Argo CD, Flux CD, Harness, and internal developer platforms for multi-cluster GitOps - and how to get consistent sync policies, per-environment RBAC, and full deployment audit trails without hand-configuring every cluster.

Oct 2, 2026

Who Approved That Deploy? Audit Trails and RBAC Across Multi-Cluster ArgoCD

ArgoCD records what synced and when, not who approved it. Here is how to join Git PR reviews, ArgoCD RBAC and sync events, and Kubernetes API audit logs into one audit trail across every cluster - plus what each option costs you.

Oct 26, 2026

Kubernetes can't tell which AI agent took action: how to get per-agent attribution and audit trails in 2026

Kubernetes audit events record the credential that made the API call, never the AI agent behind it. Here is the exact audit schema, what it is missing, and five ways to get per-agent identity, scoped permissions and a reviewable audit trail with tools that exist today.

Nov 10, 2026

Google's Agent Gateway now reaches EKS, AKS and OpenShift: how to get per-agent identity and policy on your own Kubernetes today

Google put Agent Gateway for cross-cloud into Limited Preview on 11 November 2026. Here is what it enforces, and how to build the same per-agent identity, policy, egress control and audit trail on Kubernetes in your own cloud account with tools that are generally available now.

Nov 19, 2026

Inside Qovery’s security architecture: how we secure your cloud & Kubernetes infrastructure

Discover how Qovery bridges the gap between developers and infrastructure with a "security by design" approach. From federated identities and unique encryption keys to real-time audit logs and SOC2 Type 2 certification - see how we protect your data while eliminating vendor lock-in.

Jan 7, 2026

Strengthening the Security of Qovery's Access to Your AWS Account

Qovery now sends your organization ID as the sts:ExternalId when it assumes the IAM role in your AWS account, and the role's trust policy only accepts it on behalf of your organization. Your role is already updated. If you manage it with the qovery-role-creation CloudFormation stack, update the stack to the latest template.

Oct 23, 2026

The Secure PaaS Alternative: Keep the PaaS Experience, Run It on Your Own Cloud

Railway, Northflank and seven other developer platforms disclosed breaches in October 2026. Here is why a shared PaaS turns one breach into everyone's breach, how Qovery gives you git-push deploys while your apps and data run in your own AWS, GCP, Azure or Scaleway account, and what Qovery still holds.

Oct 31, 2026

Best Practices for Container Security on AWS

Containers are superior to Virtual Machines (VM) in many aspects, including the benefits of being lightweight, easily scalable, highly portable, etc. AWS provides two primary container services i.e., Amazon Elastic Container Service (Amazon ECS) and Amazon Elastic Kubernetes Service (EKS). Although both ECS and EKS are growing in popularity but achieving security while using these services is relatively a new topic.

Apr 13, 2022

6 Best Practices to Automate DevSecOps in Days, Not Months

Six DevSecOps practices that work: shifting security left, SAST and DAST in the pipeline, securing CI/CD itself, and where scanning belongs.

Dec 3, 2025

Zero-friction DevSecOps: get instant compliance and security in your PaaS pipeline

Shifting security left shouldn't slow you down. Discover how to achieve "Zero-Friction DevSecOps" by automating secrets, compliance, and governance directly within your PaaS pipeline.

Dec 22, 2025

Security Considerations for Your Internal Developer Platform

In today's world, where cloud resources and data management tools play an increasingly critical role, the concept of an Internal Developer Platform (IDP) is gaining momentum. Imagine a platform where developers seamlessly design, build, and deploy applications. That's precisely the promise of IDPs. But here's the highlight: with great power comes greater responsibility. Security within IDPs isn't just an optional add-on; it's the core essence. This article is a guide for teams eager to strengthen their developer platforms. It journeys through the various aspects of IDP security, delving deep into its diverse world. From the fundamental principles of access control and Kubernetes management to the elevated considerations of building a security-centric organization culture, we've got it covered. Whether you're new to the IDP environment or a seasoned developer, our objective remains unchanged: to arm you with the knowledge to transform your developer platform into an impenetrable fortress.

Oct 11, 2023

Kubernetes network isolation and NetworkPolicy at fleet scale

Kubernetes NetworkPolicy acts as a native firewall to control pod-to-pod communication. Because Kubernetes allows all lateral traffic by default, configuring network isolation manually creates massive configuration drift. Agentic control planes automate zero-trust compliance, deploying and reconciling network policies simultaneously across thousands of global clusters.

Feb 7, 2026

How We Use an AI Agent to Handle Vanta Vulnerabilities Faster

We have a commitment to fix critical vulnerabilities within 10 days. Here is the scheduled agent we built on Qovery Agent Tasks to investigate Vanta findings, carry fixes through code changes and checks, and hand every decision back to an engineer.

Oct 7, 2026

Frequently asked

What is infrastructure security?
Infrastructure security protects the platform applications run on: cloud accounts, Kubernetes clusters, networks, CI/CD pipelines and the credentials that connect them. It covers patching, secrets management, access control, policy enforcement, network isolation and audit logging.
How do you respond to a Kubernetes CVE across many clusters?
Inventory first: find every cluster, namespace and image running the affected component and version. Apply a mitigation you can roll out in hours, such as a config change or a network policy, then schedule the real upgrade with a canary and a rollback plan. Keep the inventory automated so the next CVE starts from a known list.
What is policy as code, and should I use OPA or Kyverno?
Policy as code expresses rules such as allowed registries, required labels or resource limits as versioned code that is checked automatically. OPA Gatekeeper uses the Rego language and also works outside Kubernetes; Kyverno writes policies as Kubernetes YAML and is easier to start with. Both run as admission controllers.
How should secrets be managed on Kubernetes?
Keep the source of truth in a secrets manager such as AWS Secrets Manager, Vault or Doppler, sync values into the cluster at deploy time, enable encryption at rest for Kubernetes secrets, scope access per environment, and rotate on a schedule. Never commit secrets to Git or print them in build logs.
Does the EU Cyber Resilience Act require an SBOM?
Yes. The Cyber Resilience Act requires manufacturers of products with digital elements sold in the EU to draw up a software bill of materials, with full application from 11 December 2027 and vulnerability reporting obligations starting earlier, in September 2026.
How do you give AI agents infrastructure access safely?
Give each agent its own identity and a scoped token instead of a human credential, check every request against policy before it runs, limit which environments it can touch, and log every action with the agent and the person behind it. Qovery does this with OPA-based policy tokens and audit logs.

Every operation scoped, checked and logged.

Qovery runs your workloads in your own cloud account, with RBAC, OPA-based policy tokens for agents, secrets manager integrations and an audit trail on every change.