Fit Servers

 Install Portainer CE on an Ubuntu 24.04 dedicated server with Docker. Secure the UI, manage containers, back up and update. Running Docker on a dedicated server is powerful, but once you have several containers, images, volumes and networks, the command line becomes hard to keep track of. Portainer adds a web dashboard on top of Docker, so you can see and manage everything in one place. What You Will Learn • Installing Docker Engine on Ubuntu 24.04 from the official repository • Creating a persistent volume and deploying Portainer CE • Restricting access with UFW or an SSH tunnel (and why Docker ports can bypass UFW) • Testing container management from the Portainer web interface • Backing up and updating Portainer • Troubleshooting common installation problems Why Use a Dedicated Server? Container workloads benefit from bare-metal performance: no noisy neighbors, full CPU and disk throughput, and complete control over the firewall and network. Quick Preview Once Docker is install...

  Running a dedicated server means you are responsible for its uptime and performance. CPU saturation, low memory, a full disk or unexpected network traffic can all hurt your applications before you notice. That is why monitoring should be part of your setup from day one. Prometheus is an open-source monitoring and alerting toolkit that collects time-series metrics from configured targets, stores them locally and lets you query them with PromQL . Paired with Node Exporter , it gives you a clear picture of your Linux server’s health. What our tutorial covers Installing Prometheus and Node Exporter from official release archives, with checksum verification Creating dedicated, unprivileged service accounts Configuring prometheus.yml and validating it with promtool Setting up systemd services so both tools start at boot and restart on failure Verifying health with the readiness endpoint, the targets API and the up metric Accessing the web interface securely through an SSH tunn...

 Learn what IPMI is, how it enables remote dedicated server management, and why businesses use it for hardware monitoring, troubleshooting, and recovery. Imagine your dedicated server stops responding at 2:00 AM. The operating system is locked up, SSH is dead, and the data center is hundreds of miles away. Before modern remote management, you would have had to wait for a technician to walk to the rack. Today you can fix it from your laptop, thanks to IPMI . What is IPMI? IPMI (Intelligent Platform Management Interface) is a standardized interface for monitoring and controlling server hardware independently of the operating system. It runs on the Baseboard Management Controller (BMC) , a dedicated chip on the motherboard with its own standby power and network connection. Why dedicated servers use IPMI Remote power management: power on, power off, or hard reset a frozen server. Hardware health monitoring: track temperatures, fan speeds, and voltages before failures happen. K...

Learn how to deploy Google’s Gemma 3 on a dedicated Ubuntu GPU server with vLLM, Docker and systemd. Step-by-step guide.   Running a large language model on your own server gives you full control over your data, your latency and your costs. Google’s Gemma 3 is a strong open model family for this, and vLLM makes serving it simple. What you will build A Gemma 3 endpoint on a single Ubuntu GPU server, bound to localhost, managed by systemd and tested with both text and image prompts. vLLM exposes an OpenAI-compatible API, and Docker keeps the CUDA and Python dependencies isolated from the host. Choosing a model size Gemma 3 comes in 1B, 4B, 12B and 27B sizes. The 4B, 12B and 27B models accept images as well as text and support a 128K-token context window. As a rough guide: 4B: one 16 GB GPU 12B: one 40–48 GB GPU, or two 24 GB GPUs 27B: one 80 GB GPU, or multiple GPUs bfloat16 needs an NVIDIA GPU with compute capability 8.0 or higher (Ampere or newer). The steps in short Instal...

  Learn how to install and configure HAProxy as a high-performance load balancer on Ubuntu 24.04, covering setup, routing and SSL. If you are scaling your infrastructure, routing traffic efficiently is just as critical as the compute resources behind it. HAProxy is a standard choice for high-performance TCP/HTTP load balancing, and it is lightweight enough to run on a modest server. What is covered in this tutorial This guide sets up HAProxy as a reverse proxy and load balancer on Ubuntu 24.04 LTS with a simple three-node layout: one HAProxy node and two Nginx backend servers. Step-by-step overview Prepare the backends: install Nginx, create a distinct page on each server so you can see the balancing work, and restrict port 80 with UFW so only the load balancer can reach it. Install HAProxy: install it from the Ubuntu repositories (version 2.8 or newer) and verify the version. Configure HAProxy: back up the default file, then use a consolidated config with round-robin bal...

Compare Dense vs. MoE AI models and learn how active parameters and expert routing affect GPU server memory, compute and hardware needs. Large language models are no longer defined simply by how many parameters they contain. Two models can have very different hardware requirements even when their parameter counts look similar, because the architecture decides how much computation happens per token and how the weights are spread across GPUs. Dense models vs. Mixture-of-Experts (MoE) A dense model uses essentially the same parameters for every token. An MoE model contains multiple expert networks, and a learned router activates only a subset for each token. The key concept: total vs. active parameters DeepSeek-V3 has 671B total parameters but activates 37B per token. The compute per token is lower than a dense 671B model, but the full set of weights generally still needs to sit in GPU memory. What this means for your GPU server VRAM: MoE does not automatically use less VRAM. Inactiv...