Servers & virtualization

The choice of MikroTik model depends on the number of users, Wi-Fi requirements, number of ports, and desired internet throughput. For a small office (up to 20-30 users), compact hAP or RB series models are usually sufficient, while larger offices should consider models with hardware acceleration and more ports. The key is to account not only for current needs but also for future growth.

A basic setup includes configuring a DHCP server for automatic IP assignment, firewall rules to protect the network from external threats, VLAN segmentation (e.g., a separate guest Wi-Fi isolated from the work network), and a VPN server for secure remote access for employees. It's also worth setting up monitoring of the device's status.

Common mistakes include using default passwords and exposing management ports to the internet, lacking a default-deny firewall policy, ignoring VLAN separation between guest and work networks, and not keeping configuration backups. To avoid problems, change default credentials immediately, block management interface access from the internet, regularly update RouterOS, and keep configuration backups.

Yes, MikroTik supports several VPN protocols (such as IKEv2, WireGuard, L2TP/IPsec), allowing you to configure secure remote access to the office network for employees. This is especially relevant for hybrid teams or remote access to internal company services.

Key metrics to monitor include CPU load, RAM usage, disk free space and speed, network traffic and latency, and the status of critical services (web server, database, etc.). Comprehensive monitoring of all these parameters allows you to catch a problem at an early stage, before your clients or users even notice it.

The choice depends on infrastructure scale and needs. Zabbix is a versatile solution with ready-made monitoring templates for a wide range of systems, well-suited for classic server infrastructure. Prometheus+Grafana is a flexible stack popular in containerized and cloud environments, with powerful visualization. UptimeRobot is a simple service for basic external availability monitoring of sites and services, without deep detail. For full server monitoring, consider Zabbix or Prometheus+Grafana, and use UptimeRobot as an additional external availability check.

Most modern monitoring systems (Zabbix, Prometheus/Grafana) support integration with Telegram via bots and webhooks: when set thresholds are exceeded (e.g., CPU above 90% for several minutes), the system automatically sends a message to a designated chat or channel. It's important to properly configure alert thresholds to avoid noise from insignificant notifications while not missing truly critical events.

Thresholds should be set based on the system's real baseline load rather than arbitrary numbers — for example, by analyzing average CPU/RAM metrics over a week and adding a margin. It's also helpful to trigger alerts based on sustained duration rather than instant spikes (e.g., 'CPU above 90% for 5 minutes') to avoid false alarms from short-term load peaks.

The choice depends on budget, scaling needs, and the importance of latency to Ukraine. Hetzner is usually attractive due to lower pricing and European data centers, which provide low latency for Ukrainian users. AWS and Google Cloud offer greater flexibility, a wider range of managed services, and global infrastructure, but at a higher cost. For small and medium businesses with limited budgets, Hetzner is often the more optimal choice, while projects requiring complex cloud services or global scaling may benefit more from AWS or Google Cloud.

Latency depends on the location of the provider's data centers relative to Ukraine. European data centers (e.g., in Germany or Finland) typically provide lower latency for Ukrainian users compared to data centers in other regions of the world. Before choosing a provider, it's worth checking the specific location of available regions and, if possible, testing actual latency from Ukraine.

Paying for cloud services as a Ukrainian business has its own nuances — from available payment methods to currency and banking restrictions. Since terms can change, it's worth checking current payment options and invoicing availability for legal entities directly on the provider's website, or contacting us for a consultation before choosing a platform.

Docker is a containerization platform that allows you to package an application together with all its dependencies into a lightweight, isolated container. Unlike virtual machines, containers don't require a separate guest operating system — they use the host OS kernel, making them much lighter, faster to start, and more resource-efficient. This is especially useful for development, testing, and deploying applications across different environments with consistent results.

A typical approach is to describe all services (the Laravel application itself, the MySQL database, and Redis cache) in a docker-compose file, which defines how the containers interact within a single network. This allows you to spin up a complete development or production environment with a single command, identical on any machine. A detailed step-by-step example of such a configuration is covered in our article.

Docker isn't always justified — for very simple projects with a single application and no complex dependencies, or when the team lacks containerization experience and deadlines are tight, the overhead of adopting Docker may outweigh the benefits. It's also not worth introducing containerization 'just in case' if the infrastructure is already stable and doesn't require scaling or frequent deployments across different environments.

We approach every project individually — from analyzing current infrastructure and selecting the optimal solution (cloud, dedicated server, VMware or Proxmox virtualization) to implementing monitoring and configuration documentation. This means you get not just a 'configured server,' but a transparent, documented, and manageable system with ongoing support after project completion.