DevOps & SRE notes
12.2K subscribers
45 photos
19 files
2.54K links
Helpful articles and tools for DevOps&SRE

WhatsApp: https://whatsapp.com/channel/0029Vb79nmmHVvTUnc4tfp2F

For paid consultation (RU/EN), contact: @tutunak


All ways to support https://telegra.ph/How-support-the-channel-02-19
Download Telegram
Understanding how many pods your infrastructure can actually support is crucial for reliability. This overview breaks down the nuances of Kubernetes cluster capacity and resource allocation.
https://dnastacio.medium.com/kubernetes-cluster-capacity-d96d0d82b380
πŸ‘2❀1
K8sQuest β€” A local, hands-on Kubernetes learning game with real-world troubleshooting challenges. Practice Pods, Deployments, Services, networking, storage, and debugging using kubectl on a local cluster (kind/k3d). No cloud required.

https://github.com/Manoj-engineer/k8squest
πŸ”₯14❀1
A good starting point for finding a Helm chart that is not officially provided by the vendor is the Community Helm Chart Repository.

https://github.com/trueforge-org/truecharts
πŸ‘4❀1
As announced November 2025, Kubernetes will retire Ingress-NGINX in March 2026. Despite its widespread usage, Ingress-NGINX is full of surprising defaults and side effects that are probably present in your cluster today. This blog highlights these behaviors so that you can migrate away safely and make a conscious decision about which behaviors to keep. This post also compares Ingress-NGINX with Gateway API and shows you how to preserve Ingress-NGINX behavior in Gateway API. The recurring risk pattern in every section is the same: a seemingly correct translation can still cause outages if it does not consider Ingress-NGINX's quirks.

https://kubernetes.io/blog/2026/02/27/ingress-nginx-before-you-migrate/
πŸ‘5❀3
Although Ingress-Nginx is still maintained and receiving security updates (e.g. controller-v1.15.0), it's time to start migrating to the Gateway API. ingress2gateway can help with that.
πŸ‘Ž3❀2πŸŽ‰1
Looking for a hosting platform to practice with Linux, Kubernetes, etc.? Register using my referral link on DigitalOcean and get $200 in credit for 60 days. By registering through my referral link, you also support this Telegram channel.

πŸ‘‰ Register
πŸ‘6❀1
Short-lived public TLS certificates are our future, with a 46-day maximum validity by 2029.

https://knowledge.digicert.com/alerts/public-tls-certificates-199-day-validity
😒6😱4πŸ’―2❀1
🚨 Trivy has been hacked, again.

---

What happened?

Attackers compromised the official aquasecurity/trivy-action GitHub Action β€” the one people use to run Trivy vulnerability scans in CI/CD pipelines. This was disclosed today (March 20, 2026). It's the *second* Trivy-related supply chain attack this month β€” the first one hit the Trivy VS Code extension on OpenVSX, where injected code tried to abuse local AI coding agents.

How did they do it?

The attacker force-pushed 75 out of 76 version tags in the aquasecurity/trivy-action repository. So if your workflow references this action by a version tag like @0.34.2, @0.33.0, or @0.18.0 β€” you're running malicious code. The only tag that wasn't touched is @0.35.0.

The tricky part: the malicious code runs *before* the real Trivy scan starts, so everything looks normal to the user.

What does the malware actually do?

It dumps the runner's process memory to grab secrets, harvests SSH keys, and steals credentials for AWS, GCP, Azure, and also Kubernetes service account tokens. Basically, it's an infostealer designed specifically for CI/CD environments.

How big is the blast radius?

Over 10,000 workflow files on GitHub reference this action, so potentially a lot of projects are affected. The compromised tags were still active at the time the article was written.

Key risks for you to think about:

Given your EKS and GitOps setup, here are the things I'd pay attention to:

1. K8s service account tokens leaked β€” if any of your CI pipelines use trivy-action and have access to your EKS clusters, those tokens could be compromised. Rotate them.

2. AWS credentials exposed β€” your IRSA roles, Secrets Manager access, anything the GitHub runner had in its environment could be stolen.

3. Tag pinning is not enough β€” this attack shows that even pinning to a specific version tag like @0.33.0 doesn't protect you. Tags in Git can be force-pushed. The safe approach is to pin to a full commit SHA not a tag.

4. Second attack in one month on the same tool β€” Trivy is popular, and attackers clearly see it as a high-value target. Worth thinking about whether your security scanning toolchain has a single point of failure.

What to do right now:

- Check if any of your GitHub Actions workflows reference aquasecurity/trivy-action by tag (not by SHA).
- If yes, treat your CI/CD secrets as compromised β€” rotate AWS keys, SSH keys, K8s tokens.
- Switch to referencing actions by commit SHA instead of version tag.
- Review your GitHub Actions workflow permissions β€” make sure you use least-privilege permissions: blocks.

This is a really good example of why "shift left security" needs to also include securing the security tools themselves. The scanner became the attack vector.

https://socket.dev/blog/trivy-under-attack-again-github-actions-compromise
Please open Telegram to view this post
VIEW IN TELEGRAM
πŸ‘9❀7😱2
⚑️ LocalStack archived its GitHub repo β€” what happened and what it means

On March 23, 2026, LocalStack archived localstack/localstack on GitHub (read-only) and consolidated everything into a single Docker image that requires an auth token β€” including in CI.
What changed:
- docker pull localstack/localstack:latest without LOCALSTACK_AUTH_TOKEN β†’ your pipeline breaks
- Free "Hobby" plan exists but requires account creation and is non-commercial only
- Paid plans start at $39/mo
- CI needs a dedicated CI Auth Token stored in secrets
Your options:
- Pin to an older tag (e.g. 4.12) β€” works short-term, but you accumulate parity drift and unpatched CVEs
- Create a free account β€” enough for individual non-commercial dev
- Pay β€” if LocalStack is embedded in team CI
For open-source projects: LocalStack launched a separate program offering free Ultimate tier licenses (100+ AWS services, Cloud Pods, IAM enforcement) to eligible OSS projects with OSI-approved licenses.

https://blog.localstack.cloud/introducing-localstack-for-open-source/
😒7😱3❀1🀯1
Terragrunt is now v1!

This means that Terragrunt will no longer have any breaking changes in minor releases, with all future breaking changes taking place in (infrequent) future major releases.

https://github.com/gruntwork-io/terragrunt/releases/tag/v1.0.0
πŸ”₯8❀2
A realistic exam simulator for CKAD, CKA, and CKS featuring timed sessions and hands-on labs with pre-configured clusters.

https://github.com/sailor-sh/CK-X
πŸ‘9
Forwarded from AWS Notes (Dzm Var)
πŸš€ Join our AWS Cloud Developer course and gain practical experience working with various AWS services and technologies during an intensive 2-month program!

🎯 Become familiar with over 15 AWS services and confidently apply your skills in real-world projects by the end of the course.

β–ͺ️Start date: April 13, 2026
β–ͺ️Free / Online
β–ͺ️In English

COURSE HIGHLIGHTS


☁️ Build a single-page application (SPA) using AWS S3 and CloudFront
☁️ Develop serverless APIs with AWS API Gateway and Lambda
☁️ Integrate with databases and S3
☁️ Master asynchronous microservices communication using SQS and SNS
☁️ Explore authorization with Lambda Authorizers and Cognito
☁️ Learn containerization with Docker and Elastic Beanstalk
☁️ Implement backend-for-frontend using API Gateway

Find out more details and enroll:
https://wearecommunity.io/events/aws-cloud-dev-2026q2
πŸ”₯4❀3πŸ‘1πŸ‘Ž1
In the modern world of services, the teams developing and working with them need to understand the concept of rate limiting. Being able to design these systems is crucial to avoid service denial and provide higher uptime. It is an important part not only for externally available services, but also for internal ones.

https://bytebytego.com/courses/system-design-interview/design-a-rate-limiter
πŸ‘5