Issue #6: Systemd Timers, Kubernetes Shift Down, K8s Agentic Framework and More

DevOps engineers will play a critical role in ensuring AI-driven workloads are scalable, secure, and observable.

This week's Newsletter Highlights:

• Systemd Timers
• Grafana Loki Architecture
• Kubernetes Shift Down Security
• AI Agents and Automation

?? Linux Foundation Offer (Few Days Only)

For a limited time, get up to 55% discount on CKA, CKAD, CKS, etc.

For Individual Certifications,

Use code MAR25CT at kube.promo/devops to get 40% flat discount

The following are the best bundles to save upto 55% with code MAR25CT.

1. KCNA + KCSA + CKA + CKAD + CKSa ($1685 Savings)
2. CKA + CKAD + CKS Exam bundle ($712 Savings)
3. CKA + CKS Bundle ($477 Savings)

? Act fast—this limited-time offer won’t be around much longer!

?? Detailed Guides & Insights

1. Systemd Timers (Better than Cron)

Systemd timers are more powerful and flexible than cron.

It is designed to work directly with systemd service units.

With systemd timers, you get:

• Fine-grained interval support (e.g., seconds).
• Logs are automatically managed by journalctl.
• Supports dependencies (e.g., jobs can wait for others to finish).
• Can automatically restart failed jobs.
• Enhanced security through systemd’s sandboxing features.

Overall, systemd timers address many issues and limitations of cron.

Here is how a systemd timer works

Unlike cron, which uses a single file to schedule tasks, systemd timers use two key files, service and timer files.

1. ?????????? ????????: Specifies when the task should run (e.g., daily, weekly, or at a specific time).
2. ?????????????? ????????: Defines what the task is (the script or command to execute).

For example, if you want to run a backup script daily:

• You create a service file to define the script that will be executed.
• You create a timer file to set the schedule for running that service. (e.g., every day at midnight).

The best part is, you can manage timers just like other services using commands like systemctl start, stop, enable etc.

We have published a comprehensive beginner’s guide with practical examples on systemd timers.

Hands On Guide: Systemd Timers Practical Example

2. Grafana Loki Architecture

In this blog,

You will learn the following.

1. What is Grafana Loki?
2. Loki Deployment Modes
3. Features of Grafana Loki
4. Grafana Loki Architecture
5. Grafana Loki Components
6. Difference Between Grafana Loki and EFK Stack
7. Example of How Chunks and Indexes are Stored in Loki

???????????????? ????????: Grafana Loki Architecture

3. Kubernetes Shift Down Security

As DevOps engineers, focusing on Kubernetes security is important.

You may already be familiar with Shift-Left Security, where developers scan their code for vulnerabilities and misconfigurations early in the development process.

However, Shift-Down Security takes a different approach—it moves security responsibilities from developers to platform engineering teams by embedding security directly into the Kubernetes platform itself.

This means the platform team provides hardened base images, enforces security policies, and ensures compliance, so developers can focus on building applications without worrying about security at every step.

The following are the key Principles of Shift-Down Security

1. Platform-Owned Security: Instead of making each development team handle security separately, the platform team takes responsibility for enforcing security across the organization.
2. Security as Code & Automation: Using Policy as Code (PaC) to automate security enforcement across clusters, CI/CD pipelines, and runtime environments.
3. Built-in Security: Instead of treating security as an add-on, it is integrated directly into the platform and applied consistently across all teams.

The Kubernetes Shift-Down Security approach ensures that security is a core part of the infrastructure, making it easier to manage at scale.

Official Paper: Read it Here

4. How Agentic AI is Reshaping Automation

AI is changing. It's not just about chatbots or creating text anymore.

Traditional Generative AI usually responds to single questions or does tasks one at a time. It might answer based on one instruction (zero-shot) or using a few examples (multi-shot).

Agentic AI takes AI further. It can do more than simply answer questions or generate text. It can think in multiple steps, learn new things on its own, collaborate with other AI systems, and complete tasks without constant human instructions.

Instead of just giving information, Agentic AI can plan actions, make decisions, and carry out tasks by itself to achieve specific goals.

This means AI can now:

• Analyze logs & monitor systems
• Trigger automated remediation
• Collaborate with other agents
• Optimize CI/CD pipelines dynamically

Also, AI Adoption is Growing fast.. GenAI adoption rate is 2x faster than PCs & the internet!

What does this mean for DevOps?

AI-powered automation is no longer a "nice-to-have", it's a competitive advantage.

Want to Dive Deeper?

Shruti Bhat discusses the evolution of AI strategy, highlighting how real-time AI search and retrieval systems are transforming enterprise data use. Read More

5. Goodbye SaaS, Hello AI Agents

Traditional SaaS applications (like CRM, ERP, and ticketing systems) require human intervention to complete tasks.

AI agents take it further, not only providing insights but also taking action autonomously.

For example, instead of manually sending follow-up emails to leads, an AI agent detects them, drafts an email in your tone, and sends it.

Why this Matters for DevOps Engineers?

• Enterprise software will transition from human-driven to AI-driven, requiring new orchestration frameworks (potentially Kubernetes for AI).
• Just like microservices needed Kubernetes, AI agents need orchestration, security, and monitoring.
• AI agents require both short-term and long-term memory to function.
• Unlike traditional stateless applications, agents need to store contextual history, increasing infrastructure demands.

DevOps engineers will play a critical role in ensuring AI-driven workloads are scalable, secure, and observable.

AI-powered agent frameworks are in their early stages (similar to Docker in the container revolution).

We will need orchestration tools for AI agents, similar to how Kubernetes orchestrates containers.

Want to learn more?

Watch the full discussion on how AI agents are replacing traditional SaaS.

??? Tool of the Week

Kagent: It?is a kubernetes native framework for building AI agents.?

The?kagent?framework is designed to be easy to understand and use, and to provide a flexible and powerful way to build and manage AI agents.