OTUS Publishes Extensive Digest Covering Kubernetes, GitLab CI, DevSecOps and Infrastructure Reliability
OTUS has published a comprehensive digest designed to help IT specialists tackle complex infrastructure tasks that rarely occur in isolation. The material connects issues such as unstable deployments with CI/CD pipelines, container management, secrets handling, performance problems with monitoring and load testing, and network failures with routing, high availability, and Linux configuration.
The digest is structured into several thematic blocks. The first block offers free lessons on Kubernetes, containers and automation. Scheduled sessions include “K8S + Vault — how to obtain secrets?”, “Using GitLab CI for Ansible”, “Docker in production”, “Kubernetes + CI/CD + GitOps”, and “DRY principle in GitLab CI”.
A separate section covers monitoring, logging and performance with lessons on OpenTelemetry, first load testing practices, and comparison of ELK, EFK and Graylog systems.
Network, storage and resilience topics feature lessons on VLAN configuration, RAID5 recovery in Linux, MPLS for corporate networks, and building fault-tolerant clusters using VRRP and HAProxy.
DevSecOps and infrastructure security lessons address business outcomes of DevSecOps projects, practical secure release processes involving SAST, SCA, container scanning and security gates, as well as Linux kernel protection mechanisms.
Additional lessons focus on Linux kernel modules and expert-level PostgreSQL administration for high-load environments.
The article section contains practical guides such as “Your docker-compose.yml will break: 5 settings everyone forgets”, migration from Fail2Ban to CrowdSec for Netbird and Caddy, IP forwarding at L2 and L3 layers, self-service deployment strategies, Kubernetes architecture overview, container security with capabilities, seccomp and AppArmor, L2 loop and broadcast storm detection, CI automation anti-patterns, and a case study of an Intel X710 multicast storm.
For deeper study, OTUS presents advanced courses with entrance testing: “DevSecOps Implementation”, “Observability”, “High-Load Systems Infrastructure”, “Linux Kernel Development”, “PostgreSQL Administration — Expert Level”, “Network Engineer — Advanced”, “Kubernetes Infrastructure Platform”, and “CI/CD with GitLab”.
Related articles
Samara University Study Shows Social Ties Maintain Order in Minecraft Virtual Cities
Researchers from Samara University examined how player-run cities in Minecraft sustain order without heavy reliance on formal rules or enforcement mechanisms. Olga Udalova conducted eight interviews with virtual mayors and surveyed 246 players across multiple communities. The study found that dense social connections, personal reputation, and informal authority of mayors play a far greater role than written regulations. In some cities detailed norms exist while others operate with minimal rules, simply requiring players to act reasonably. A notable example is Boostergrad, a semi-anarchic community of 117 residents where griefing is formally permitted yet rarely occurs because reputational damage outweighs any short-term gain. Residents resolve conflicts through negotiation and voluntarily compensate for damage, as illustrated by one player who paid 2000 units of in-game currency after accidentally destroying another’s loot. Overall the research demonstrates that trust and social networks, rather than fear of punishment, keep these virtual societies stable.
Why Automation Alone Fails to Improve SOC Efficiency: The Case for Managed Operational Models
Mature security infrastructure does not guarantee protection against cyber threats, as attacks can still cause downtime, crisis recovery, and financial damage even when key controls are deployed. Automation of routine tasks helps speed up responses but cannot raise SOC effectiveness without clear decision criteria, defined responsibilities, and repeatable incident-handling processes. Organizations using an average of 83 security tools from 29 vendors face fragmentation that automation alone cannot resolve. A managed SecOps model covering the full incident lifecycle—from detection and prioritization through investigation, response, closure, and post-incident improvement—is required. Positive Technologies highlights that SecOps platforms such as MaxPatrol 360 provide the unified environment needed to link context, processes, coordination, automation, metrics, and continuous improvement.
Building a Minimalist UI Test Framework with Playwright and Pytest
Vladislav Timashenkov from InfoWatch presents a practical approach to constructing a concise UI automation testing framework using Playwright. The article focuses on leveraging native Playwright features instead of creating additional abstraction layers. It covers launching a Chromium browser once per session, handling authentication via API calls to obtain cookies, and managing isolated BrowserContext instances for test independence. Readers learn how to implement Page Object models that rely on Playwright's built-in locators and auto-waiting mechanisms. The tutorial includes real code examples for fixtures, context factories, and a sample test for tag management functionality. This method reduces maintenance overhead while ensuring tests remain scalable across different environments.
Russia Develops Domestic ERA-GLONASS System for Public Transport Tracking During GPS and GLONASS Disruptions
AO GLONASS has created a national system that maintains real-time monitoring of buses and trolleybuses even when satellite navigation signals are jammed or lost. The solution relies on the state-run ERA-GLONASS platform, placing autonomous markers at stops and identifiers on vehicles to record exact positions without depending on GPS or GLONASS. Data is transmitted over a protected ERA-GLONASS communication channel that remains available during mobile internet restrictions thanks to inclusion in official white lists. The first deployment begins in Kaluga on 1 September, covering three trolleybus routes with live tracking for passengers, dispatchers, carriers, and mapping services. The battery-powered markers are designed for five years of maintenance-free operation in any weather and are built entirely with Russian components and encryption algorithms. Future plans include extending the technology to municipal vehicles and emergency services fleets.