Daily Reading List – February 4, 2025 (#486)

It’s been a while since I’ve shipped a reading list with this little AI content. It wasn’t intentional; other topics captured my attention today. Read below for content on security, Kubernetes, software deployment, and more.

[blog] Locking Down the Cloud with Cloud Armor – Easier Than You Think. Ideally you can basically “check a box” and protect your cloud environment from DDoS attacks, SQL injection, and more. Oh wait, you can.

[article] The End of Search, The Beginning of Research. Performing a web search to get a specific answer still matters. But the rise of focused “research agents” (like the new one from Open AI, or Gemini Deep Research) are a real breakthrough.

[blog] It’s OK to hardcode feature flags. Controversial take? Sure, but I like the argument. See if you agree.

[blog] Detection as Code in Google SecOps with Terraform. Cool idea which makes sense to me. Security teams can use IaC products like Terraform to define threat detection rules and deploy them as they would with software.

[blog] Infrastructure as Code is too generic. Continuing the thought from the last piece, is IaC too generic? Brian says there are problems with standard tools that don’t understand the purpose of resources or how they relate to each other.

[article] The Startup Drake Equation. Extensive, interesting piece from Jason here. What are all the activities and outcomes that need to line up for a startup to succeed? How do you reduce some key risks?

[blog] Get Started with n8n on Google Cloud for AI Workflow Automation. I hadn’t heard of n8n until today. Karl does a great hob showing us how to visually model (AI) workflows using an open platform deployed to a Kubernetes cluster.

[blog] A Generative AI Agent with a real declarative workflow. Let’s talk more about workflows. Guillaume shows how to take a code-based AI app and re-create it with a declarative workflow.

[article] How to Encourage the Right Kind of Conflict on Your Team. Very good advice here. It’s ok to have some positive tensions and the freedom to disagree. But done in the right way.

[blog] Kubernetes History Inspector. William gushes over this “information dense” view of Kubernetes logs. Try out this new open source tool.

[blog] How we improved GKE volume attachments for stateful applications by up to 80%. I wouldn’t have thought of this problem, but then again, I wouldn’t have personally experienced it. If you have lots of Kubernetes clusters with stateful workloads, you’ll care about this.

Want to get this update sent to you every day? Subscribe to my RSS feed or subscribe via email below: