Free Cloud & DevOps course from zero: terminal to Kubernetes

Shanraq has released a new free course, Cloud & DevOps from Zero. It is designed for someone who may be opening a terminal for the first time and does not yet know how DNS differs from an IP address, why containers exist, or what happens between git push and a working website.

The course is open in Kazakh, Russian, and English. It contains 24 ordered lessons, one ready-to-run CloudLab learning project, and a practical path from a local process to a secured service with automated delivery, observability, backup, and recovery.

Open the free Cloud & DevOps course

Why we released this course now

The word “cloud” can make computing sound weightless. Every cloud service actually runs on physical processors, disks, networks, and power systems inside data centres. As the world uses more AI, online services, and data processing, that infrastructure grows. So does the need for people who can release, observe, and recover software systems safely.

The International Energy Agency reports that global data centre electricity consumption grew by 17% in 2025. Its central outlook sees consumption almost doubling from 485 TWh in 2025 to 950 TWh in 2030, while consumption by AI-focused data centres triples over the same period. This does not promise growth in every job title, but it is a strong physical signal of rapidly expanding computing infrastructure. Source: IEA, Key Questions on Energy and AI.

The software layer has also moved beyond experimentation. A Cloud Native Computing Foundation survey found that 82% of surveyed container users run Kubernetes in production. The sample mainly reflects the cloud native ecosystem, but it clearly shows that containers and orchestration have become working foundations for modern services. Source: CNCF Annual Cloud Native Survey.

Why this matters for Kazakhstan

Kazakhstan is connecting digital development with domestic computing capacity as well as software products. The first 50 MW facility of the Data Center Valley project is under construction in the Pavlodar region. According to the Government, commissioning is planned for June 2027, while the site’s long-term energy potential allows staged growth to 1 GW. Plans include data centres, AI factories, GPU clusters, and cloud platforms. Source: the official website of the Prime Minister of Kazakhstan.

Projects of this kind need many different specialists: power, cooling, and telecommunications engineers; physical security and network specialists; cloud and platform engineers; SREs; DevOps engineers; and cybersecurity specialists. Our course focuses on software and platform operations. It does not train electrical or cooling designers for physical facilities, but it provides a foundation for roles that must move code reliably from a repository to a running system.

If DNS, TCP, and TLS are new to you

That is not a gap to hide. It is where the course begins.

The network first appears as a familiar office building. A domain name is the organisation’s familiar name. DNS is the directory that finds the building address. An IP address is the building’s network address. A port is the number of the right reception desk. TCP establishes a conversation with that desk. TLS checks the other party’s identity and creates a protected corridor. HTTP is the language in which the visitor asks for a particular document.

After the image, the course always gives the precise technical definition and explains where the analogy stops. The learner then predicts the result of a command, runs it, compares evidence with the prediction, and rebuilds a short support map from memory. This rhythm uses principles from Viktor Shatalov’s teaching method: start with the whole, then use a support signal, a familiar image, an exact concept, action, and active recall.

New terms do not arrive in a burst. Each one gets a place in the larger picture and an immediate practical check.

What you will build in 24 lessons

One Go service called CloudLab runs through the entire course. It accepts requests, stores data, reports its state, and exposes metrics. You first run it as an ordinary process, then gradually turn it into a system that can be released and recovered.

The path is concrete:

1. open a terminal, navigate directories, and record changes with Git;
2. understand Linux users, permissions, processes, and logs;
3. trace a request through DNS → IP → TCP port → TLS → HTTP;
4. package CloudLab as a Docker image and separate persistent data from a container;
5. connect services with Docker Compose and add readiness checks and resource limits;
6. create a CI/CD pipeline that checks the code, builds an image, and prepares a repeatable release;
7. handle secrets, deploy a new version, and roll back safely;
8. prepare a cloud server, SSH access, a firewall, a domain, and HTTPS;
9. describe infrastructure with OpenTofu and server configuration with Ansible;
10. run the application on Kubernetes, observe its signals, create a backup, and perform a recovery drill after a learning incident.

The course is estimated at 106 minutes of reading and 480 minutes of practice. Those figures are a guide. For a beginner, completing one experiment with understanding is more valuable than quickly scrolling through three lessons.

Why the learning service is written in Go

Modern infrastructure uses many languages, but much of the cloud native control layer is built in Go. Docker Engine, Kubernetes, Prometheus, Caddy, and OpenTofu are all developed in Go.

Go suits infrastructure tools for practical reasons: a project can produce one binary, the standard library has strong networking and HTTP support, and built-in concurrency helps with many simultaneous tasks. You do not need to know Go before starting; the application is ready, and the learner operates its life cycle. Anyone who wants to understand the code more deeply can take our separate free course, Go: From Zero to Your Own Blog.

Verifiable work instead of a useless certificate

Shanraq does not issue a decorative diploma for opening a lesson tab. A stamped PDF cannot restart a failed service, close an exposed secret, or prove that a backup can actually be restored.

The outcome is work that can be shown and repeated: a repository with a clear history, a container image, automated checks, a secured server, HTTPS, infrastructure as code, Kubernetes manifests, metrics, a backup, and a learning incident report. These are portfolio materials and evidence for a concrete technical interview conversation.

The course does not promise a job or income after the final lesson. The best evidence of skill is the desire to understand and sustained effort turned into a working result. You keep a tool that can solve real problems and support further professional growth.

How to begin

Open the course introduction, prepare Git, Docker, and Go through their official instructions, and take your time. Use two passes for every lesson: first understand the story and map; then run the experiment, verify the result, and reconstruct the support chain without looking.

If a command produces a different result, that is not defeat. It points to the exact place to check the directory, port, permissions, process state, or preceding step. Finding that cause is the daily work of an infrastructure engineer.

The course is free and open: Cloud & DevOps from Zero — 24 practical lessons.