Preparing for CKA, CKAD, or CKS?

Here is an opportunity to save on your Kubernetes certification journey!

The Linux Foundation has a limited-time KubeCon sale.

Use code 40KUBE26CT at kube.promo/devops to get 40% off individual certifications.

If you are planning to cations,

You can save up to 47% by choosing the right Kubernetes certification bundle.

You can find all the certification bundle information in this repo 👇

𝗕𝘂𝗻𝗱𝗹𝗲 𝗗𝗶𝘀𝗰𝗼𝘂𝗻𝘁𝘀: github.com/techiescamp/li…

𝗡𝗼𝘁𝗲: It's a limited-time KubeCon sale.

So lock in the price before it expires.

♻️ PS: Repost and share it with the DevOps community.

#Kubernetes #cka #ckad #cks

GPU scheduling in Kubernetes is becoming an important skill now.

Here is why 👇

Most organizations are either experimenting or running ML workloads in Kubernetes.

As Devops engineers you should know how to set up and manage GPU resources.

We have created a simple guide on ng NVIDIA GPU Operator

By the end of this blog, you will have a clear understanding of:

- Need for GPU operator on Kubernetes
- Setting up NVIDIA GPU Operator on a Kubernetes cluster
- Verify if Kubernetes detects GPUs
- How to deploy a real GPU-based workload to validate the full stack.

To be honest, it is not as complicated as it sounds once you understand the basics.

Even if you dont have access to GPU's, just go through the content to have an understanding on how it works.

𝗥𝗲𝗮𝗱 𝗶𝘁 𝗵𝗲𝗿𝗲: devopscube.com/setup-gpu-oper…

#devops #mlops

🛠 DevOps Tool of the Week: Stern

When you want to check logs for your application in Kubernetes,

You run kubectl get pods, copy a Pod name, and then run kubectl logs.

But what happens when a Pod gets replaced during a deployment rollout?

You have to find the new Pod and That's where Stern helps.

Stern is an open-source CLI tool that lets you stream logs from multiple Kubernetes Pods and all their containers at the same time, directly in your terminal.

Here's what it can do 👇

- Shows logs from multiple pods and containers together.
- Automatically follows new matching Pods and stops following deleted ones.
- Filters pods using names, regular expressions, Kubernetes resources, or labels.
- Filters or highlights log messages.
- Color-codes log output so you can identify which pod produced each message.
- Supports timestamps, JSON output, and custom log formatting.
- Stern is useful when troubleshooting applications with multiple replicas or following logs during Kubernetes deployments.

𝗦𝘁𝗮𝗿𝘁 𝗵𝗲𝗿𝗲: github.com/stern/stern

#devopstools

📚 𝟓 𝐞𝐧𝐠𝐢𝐧𝐞𝐞𝐫𝐢𝐧𝐠 𝐛𝐥𝐨𝐠𝐬 𝐰𝐨𝐫𝐭𝐡 𝐫𝐞𝐚𝐝𝐢𝐧𝐠

Here’s the reading list 👇

If you work in DevOps, SRE, Platform Engineering, Cloud, Data Infrastructure, or MLOps, these are worth checking out.

1/ Netflow found why more CPU cores slowed container scaling - netflixtechblog.com/mount-mayhem-a…

2/ Github rebuilt Git storage for thousands of concurrent AI agents - github.blog/engineering/ar…

3/ AWS explains why one disaster recovery test isn't enough - aws.amazon.com/blogs/architec…

4/ Google explains how to safely run untrusted AI agents at scale - opensource.googleblog.com/2026/11/sandbo…

5/ Uber tested ways to make large Spark workloads faster - uber.com/no/en/blog/ube…

🎓 𝐏𝐫𝐞𝐩𝐚𝐫𝐢𝐧𝐠 𝐟𝐨𝐫 𝐂𝐊𝐀?

If you are preparing for the Certified Kubernetes Administrator exam, check out our CKA course by industry experts. You can also use 𝗗𝗘𝗩𝗢𝗣𝗦𝟯𝟬 at checkout for a discount:

👉 courses.devopscube.com/p/certified-ku…

Which one are you reading first?

New to Istio?

We have a simple beginner fiendly guide to help you set up Istio on Kubernetes

By the end, you will understand,

- Different Istio installation modes
- Setting up Istio using Helm
- How to validate the setup with a demo application & canary release strategy.
- Key concle and VirtualService

and more..

𝗗𝗲𝘁𝗮𝗶𝗹𝗲𝗱 𝗚𝘂𝗶𝗱𝗲: devopscube.com/setup-istio-on…

#Istio #Kubernetes #DevOps

In Kubernetes, you can get deeper node-level insights with NFD

Here is how 👇

By default, Kubernetes nodes have only basic labels from the kubelet.

On cloud setups like EKS, you will see a few extra ones for instance types or zones.

But these default labels dont tell you ware, kernel, or device capabilities.

A common use case where you need these features is in AI and GPU workload scheduling.

That is where Node Feature Discovery (NFD) helps.

We also talked briefly about NFD in our Kubernetes Swap newsletter

If you have used Ansible facts before, NFD works in a similar way.

NFD collects system details from Kubernetes nodes and exposes them as node labels.

In our newsletter, we covered,

- What is NFD in Kubernetes
- Key Use Cases
- NFD Architecture
- Hands on NFD example
- Avoiding avoid label explosion when using NFD.

𝗥𝗲𝗮𝗱 𝗶𝘁 𝗛𝗲𝗿𝗲: newsletter.devopscube.com/p/kubernetes-n…

Have you come across NFD before?

♻️ Repost if you think others will find it useful too

#kubernetes #mlops

[20% OFF] KubeCon + CloudNativeCon North America 🚀

This event has it all:

- Expert-led talks
- The latest innovations from leading vendors
- 11 CNCF-hosted co-located events
- Unforgettable parties, swag, and
- Unmatched networking opportunities with TeamCloudNative.

You ount on registration.

Use code COMTECHIES20 at events.linuxfoundation.org/kubecon-cloudn…

From KubeCon, you can:

- Learn directly from engineers solving real production problems
- Meet the cloud native community in person
- Discover what top companies are actually doing with Kubernetes and AI
- Build connections that can genuinely help your career

♻️ PS: Repost and share this with your network.

Are you planning to attend KubeCon this year? Comment below

#kubecon #CloudNativeCon

Planning to take the CKA certification?

Here is a complete CKA exam Guide with study notes.

You can use it for other certifications as well.

This repo covers key areas like Gateway API, Network Policies, Cluster administration, and more.

𝗚𝗶𝘁𝗛𝘂𝗯 𝗥𝗲𝗽𝗼: github.com/techiescamp/ck…

Bookmark the repo and start your CKA prep.

If you waiting for a CKA promo code,

Use code 40KUBE26CT to get a flat 40% off for the next few days!

Found it useful?

Star the repo ⭐ and share it with someone prepping for CKA.

Got questions on any CKA topic? Drop them in the comments 👇

#cka #kubernetes

Kubernetes Swap 🚀

When you deploy a Pod in Kubernetes, each node uses its physical RAM to run containers.

If a node runs out of memory, Kubernetes may start killing Pods to keep the node stable.

With Kubernetes swap, you can avoid this by allowing the node to use a part of al memory.

In our newsletter, we explained the following in detail.

- How Kubernetes Swap Works
- Key Use Cases for Swap
- Pod QoS Classes and Swap Behavior
- How Swap is Allocated for Pods
- Hands-on: Testing Swap on Worker Nodes
- Identifying Swap-Enabled Nodes using NFD.
- Swap Observability Using Metrics

𝗥𝗲𝗮𝗱 𝗶𝘁 𝗛𝗲𝗿𝗲: newsletter.devopscube.com/p/kubernetes-s…

Have you tried or used this feature?

What was your use case?

Let us know in the comments.

#kubernetes

This is the cleaner way to manage kubectl aliases

Before Kubernetes v1.33,

We had to manually add them inside .bashrc or .zshrc, mixing them with other shell configs.

It was not a native or structured approach.

Now, with kuberc, you can define all your kubectl aliases in a t is clean, simple, and you can even version control it.

We have created a complete hands-on guide to help you get started.

𝗗𝗲𝘁𝗮𝗶𝗹𝗲𝗱 𝗚𝘂𝗶𝗱𝗲: devopscube.com/kubectl-aliase…

Have you tried kuberc yet?

How are you managing your aliases? Share your setup or tips in the comments.

#kubernetes #devops

This is a must-read for DevOps Engineers 👇

A real-world incident where Netflix encountered significant network performance degradation across different regions.

The article provides an in-depth analysis of how the engineering teams at Netflix identified, investigated, and issue.

Key lesson?

When there's a network issue, after blaming network and application teams,

Don't forget to check the kernel upgrade 😁

𝗗𝗲𝘁𝗮𝗶𝗹𝗲𝗱 𝗕𝗹𝗼𝗴: netflixtechblog.com/investigation-…

A few concepts from the blog.

1. Window size in TCP - TCP mechanism that controls the flow of data between a sender and receiver. It determines how much data the receiver can accept before it must acknowledge receipt

2. /proc/sys/net/ipv4/ directory contains many TCP-related kernel parameters that can be tweaked to adjust the behavior of the TCP stack in Linux

3. You can check the TCP window scaling from the /proc/sys/net/ipv4/tcp_window_scaling file. You can temporarily enable and disable it using sysctl.

#DevOps #SRE #NetworkTroubleshooting

DevOps to MLOPS 🚀

How is MLOPS different from usual DevOps workflows?

MLOps is different from the usual DevOps workflows, just like machine learning is different from traditional software development.

For example,

In MLOps,

You need to track three components: code, data, vOps, you only track code.

Additionally, MLOps introduces the following.

- 𝗗𝗮𝘁𝗮 𝘃𝗲𝗿𝘀𝗶𝗼𝗻𝗶𝗻𝗴: To track changes in datasets using tools like DVC (Data Version Control) or LakeFS

- 𝗘𝘅𝗽𝗲𝗿𝗶𝗺𝗲𝗻𝘁 𝘁𝗿𝗮𝗰𝗸𝗶𝗻𝗴: To log hyperparameters, metrics, and model versions for each training run

- 𝗠𝗼𝗱𝗲𝗹 𝗿𝗲𝗴𝗶𝘀𝘁𝗿𝘆: To version and manage models separately from application artifacts

- 𝗖𝗼𝗻𝘁𝗶𝗻𝘂𝗼𝘂𝘀 𝗧𝗿𝗮𝗶𝗻𝗶𝗻𝗴 (𝗖𝗧): Beyond CI/CD, models need regular retraining as data changes

In our newsletter edition,

We break down everything about what MLOps means for DevOps engineers. It covers,

- What is MLOPS
- Skills Required For DevOps Engineers
- Key MLOPS Tools
- Do you need GPU-based clusters to learn MLOps?

So if you are ready to go “all in” with MLOPS, this edition is for you.

𝗥𝗲𝗮𝗱 𝗶𝘁 𝗵𝗲𝗿𝗲: newsletter.devopscube.com/p/devops-to-ml…

Are you working on any MLOps projects right now?

Or are you planning to upskill in MLOps?

Either way, share your thoughts in the comments.

#devops #MLOPS

Deleting a pod ≠ stopping it instantly

Here is the hidden shutdown flow Kubernetes follows👇

- The API server sends a deletion event.
- Kubernetes executes the PreStop hook if any exists.
- kubelet sends a SIGTERM signal to the container's main process (PID 1).
- Applications ult to handle the signal and shut down gracefully.
- If the application process does not exit within the grace period, Kubernetes sends a SIGKILL signal to terminate the pod.

𝗡𝗼𝘁𝗲: Kubernetes does not give extra time for the PreStop hook. If your pod has a 30-second grace period, the PreStop hook and your app’s shutdown process must both finish within those same 30 seconds.

We covered the following in detail in our newsletter edition.

- What are SIGTERM and SIGKILL
- How Kubernetes handles SIGTERM and SIGKILL
- How applications can handle SIGTERM for graceful shutdown
- The role of PreStop hooks and why it matters
- The link between SIGTERM and container PID 1

→ 𝗥𝗲𝗮𝗱 𝗶𝘁 𝗛𝗲𝗿𝗲: newsletter.devopscube.com/p/kubernetes-g…

Over to you…

Have you written any custom logic to handle SIGTERM?

#devops #kubernetes

🛠 DevOps Tool of the Week: Kubeshark

Kubernetes logs can show which service failed, but

They don't always explain what happened between services.

To find the answer, you switch between pod logs and traces.

That's where Kubeshark helps.

Kubeshark is like Wireshark for cluster-wide network traffic using eBPF and displays it with Kubernetes context.

Here is what it does 👇

• Shows API requests and responses between services.
• Supports HTTP, gRPC, GraphQL, Redis, Kafka, DNS, and more.
• Decrypts TLS traffic using eBPF, without sidecars or proxies.
• Filters traffic by Pod, namespace, node, IP, or API details.
• Visualizes service communication and captures PCAP snapshots.
• Exposes traffic data through MCP for AI tools like Claude.

𝗦𝘁𝗮𝗿𝘁 𝗵𝗲𝗿𝗲: github.com/kubeshark/kube…

#devopstools

Weekend DevOps learning project 🚀

This guide aims to teach you the following practically.

- Client to Site VPN
- Client and Server Certificates Generation using EasyRSA
- Mutual TLS Authentication (mTLS)

So, why is it important to learn these?

VPN and mTLS are essential and DevOps engineers

Every organization uses VPN connectivity to securely connect to cloud resources.

The best way to understand VPN & mTLS is by setting up one.

𝗦𝗲𝘁𝘂𝗽 𝗚𝘂𝗶𝗱𝗲: devopscube.com/aws-client-vpn/

The following topics are covered in a theoretical format:

- SAML/Active Directory-based authentication.
- TCP vs UDP

Do further research to gain a detailed understanding of the topics above. They are fundamental for DevOps engineers

PS: ♻️ Repost if you find this useful. It helps the Learning community 🙂

Got any tips or info to share with the community?

Drop a comment :)

#devops #aws #devopsengineer #practicaldevops

Cloudflare cut AI inference costs by 68%.

Meta reduced GPU cluster interruptions by 99%.

Shopify found apps accidentally exposed to the internet.

Here are 5 engineering blogs worth reading if you work in DevOps, SRE, or Platform Engineering.

🧵👇

4/ Airbnb - Building Its Own Observability Platform

Airbnb shares what worked, what failed, and what it learned while building its own observability infrastructure.

🔗 airbnb.tech/infrastructure…

5/ Celonis - Migrating from Kubernetes v1.16 to v1.32

Celonis shares how it used automation and gradual workload migrations to move to managed Kubernetes across three cloud providers.

🔗 careers.celonis.com/blog/no-shame-…