Kubernetes 1.33 | AI everything | Mental Health | Gitex Asia
We’ve entered May, it’s time to reflect on what you’ve done in the past four months and then redefine your goals. No need to take pressure, but no need to be too relaxed either. Let me explain this in my own way.
Let’s say you’ve set ambitious goals for 2025 and you’re working hard toward them. But due to unexpected changes in your life or in the lives of those close to you, things might slow down. That’s totally okay. Take a break, sort out your mental health first, and then bounce back. Just make sure you hold yourself together, your goals will still be there when you're ready.
Now, another scenario: maybe you do have time, but you’re wasting a lot of it. Maybe you're lacking motivation, or something feels off in life. In this case, take a short break and start checking in with yourself more often. I highly recommend talking to people → just talk, no need to explain everything. Talking really helps clear the clutter in your mind.
If, on this 1st of May 2025, you’re feeling clueless that’s fine too. Many people feel the same way. What matters is regrouping yourself and revisiting your goals. This might mean shifting to short-term goals before tackling long-term ones. Being mentally strong and reminding yourself that you are the best and you can do anything is very important.
A few things you can do:
Start increasing your knowledge and skills. Go on X (Twitter), see what’s new in your field, and upskill. This applies to any domain.
Spread happiness. Be empathetic. When you make others smile, you’ll feel real happiness too.
Remember: You yes, you reading this are important. Stay mentally strong, give back to the community, take care of your health, and aim to be at the top of whatever you’re doing. Let’s go!!
Alright, now that we’ve talked about mental health, let’s get back to Kubernetes.
Kubernetes 1.33 has been released with lots of amazing features. Here are a couple of highlights:
User namespaces enabled by default - With user namespaces now enabled by default in Kubernetes v1.33, workloads gain enhanced host isolation and a reduced risk of lateral movement without requiring any special configuration
Stable sidecar containers - Sidecar containers, now stable in Kubernetes v1.33, are auxiliary containers within a pod that enhance the main application with features like logging and monitoring. Implemented as always-restarting init containers, they start before the main application and operate throughout the pod's lifecycle, offering independent management and resource sharing. I created a video on sidecar containers in general and also the new alpha feature when it was released. Do check out and you will get good info and demo.
In-place resource resize feature graduates to beta, enabling the vertical scaling of Pods by updating their CPU and memory resources without requiring a restart. This enhancement allows for dynamic resource adjustments, optimizing utilization and minimizing downtime for stateful applications. I have covered this a year back when it was an alpha feature introduced so you can check its usage. The feature has evolved so do check the docs for more updates.
Overall there are 64 enhancements, of those enhancements, 18 have graduated to Stable, 20 are entering Beta, 24 have entered Alpha, and 2 are deprecated or withdrawn. Make sure to check out the official blog post and do thank the awesome release team whenever you get time on socials.
It’s AI-everything season everywhere, and Kubesimplify isn’t behind either! We’re publishing videos, courses, and discussions on all things AI. Make sure to check out the latest MCP course you don’t want to miss it along with Docker model runner and a guide on building AI agents from scratch.
In the coming days, I’m dropping a dope “All Things AI” episode with Logan, and I’m so hyped about it! Everything’s ready, so it could drop anytime—make sure to keep an eye on Kubesimplify and hit that bell icon.
Before we move on to the stuff I’ve been reading, a quick shoutout to my Gitex experience.
Gitex Singapore was an amazing showcase of tech, and I had the chance to deliver a workshop with Mumshad where we dived into GitOps with ArgoCD and virtual clusters for ephemeral PR environments. If you want to try the workshop yourself, I’ve created a Killercoda scenario for it—definitely check it out!
Enjoy some pictures form the event
Before you move on to the stuff I have been reading, make sure to subscribe to this free newsletter and share in your network as it helps and motivates me :)
Awesome Reads
Funding rounds - Two most important funding rounds in the cloud native space are series D announcement from Chainguard and Series C from Cast AI. Both of these have sponsored my time to create content in the past so huge congratulations to both the teams.
Kubernetes v1.33: Storage Capacity Scoring of Nodes for Dynamic Provisioning (alpha) - The new alpha feature in Kubernetes v1.33, StorageCapacityScoring, enhances the scheduler's VolumeBinding plugin by adding a scoring method based on node storage capacity during topology-aware volume provisioning. This allows users to prioritize nodes with either the most or least available storage for pod scheduling, optimizing resource utilization and future PV expansion.
Agents in your software factory: Introducing the LLM primitive in Dagger - Dagger now features a native Large Language Model (LLM) primitive, enabling developers to build AI-powered automation directly into their software delivery workflows as agents. These agents, configured with specific environments and prompts, can interact with other Dagger primitives to perform tasks like AI-driven testing and intelligent deployment gates, with real-time observability and portable execution.
Introducing OpenAI o3 and o4-mini - OpenAI has released its most advanced reasoning models yet o3 and o4-mini which combine deep, agentic thinking with full tool use in ChatGPT, enabling capabilities like web search, Python analysis, image understanding, and visual generation in a seamless workflow.
Kubectl Get Hacked - Kubeconfig files, essential for Kubernetes cluster access, can be exploited if an attacker gains access to a developer's machine, especially when using the
execdirective for authentication. This directive allows for command execution on the local machine during authentication, posing a security risk if a malicious kubeconfig is used.Model Context Protocol: A Primer for the Developers - The Model Context Protocol (MCP) is emerging as a standard for integrating external tools and data into AI agent workflows, offering a language-agnostic approach similar to REST. MCP utilizes servers to expose resources, tools, and prompts, and clients to connect AI models, facilitating communication through JSON-RPC over HTTP with built-in security and support from major AI platforms.
Awesome Git Repos
Deep Wiki - Up-to-date documentation you can talk to, for every repo in the world.
volsync - Asynchronous data replication for Kubernetes volumes
Like reading so go and share!