In early 2026, Meta reportedly created an internal token-spend leaderboard called “Claudeonomics,” ranking employees by how many AI tokens they consumed. The predictable result was enormous consumption. When token usage becomes the metric, people optimize for token usage. Meta may have been an extreme example, but the underlying idea was everywhere for a while. Across […]

Introduction Over the last several months I have been doing a bit of experimentation locally hosted LLMs, primarily with vLLM, llama.cpp, different quantization methods, GPU configurations, and models. My overall goal is a smart enough local agent that can code and troubleshoot, and ask smarter frontier models for help/review. Part of this journey has been […]

….and I bought an AMD GPU. It’s the first time I’ve gone non-Nvidia since before 2010 as I’ve been a Nvidia loyalist since the GeForce 256 back in 1999 (was 3dfx prior). But Nvidia GPU pricing on anything with decent VRAM has gotten absurd, even on the used market. I kept hearing that ROCm (Radeon […]

Fired up my RHEL9 server today, something that I have not done in a while due to the savage summer heat in the southern US. All my Dell servers have had the summer off for this reason. Upon login via ssh I saw a warning message, which I do not think I have seen before […]

The tool sprawl in AI engineering has officially outrun our collective memory, at least it has mine. Here is the state of the stack as I know it and use it.

The launch of ChatGPT on November 30, 2022, marked a significant shift in technology, prompting experimentation and learning among engineers. The article discusses evolving practices in agentic coding, emphasizing the importance of documentation and the roles of prompt, context, harness, and loop engineering in shaping AI's future development.

Troubleshooting Claude Code CLI Authentication Issues

The project review outlines the successful construction and configuration of a 3-node GPU cluster with InfiniBand networking. It details performance testing procedures and confirms healthy RDMA performance with minimal tuning.

Enterprise Servers are loud and hot, and while a couple of 13th generation Dell servers kept a room nice and toasty in the cold of winter, it a double whammy to your electric bill in the Spring/Summer. The noise can be problematic as well, especially since I no longer have a basement, where I relied … Continue reading

The document details the installation and configuration of three Mellanox ConnectX-4 Adapters across multiple servers. It covers verifying detection, driver loading, InfiniBand setup, subnet management, and IP over InfiniBand configuration for effective connectivity and testing in a lab environment.

Introduction InfiniBand is a mature interconnect technology known for high bandwidth and low latency. It has long been used in supercomputing and HPC environments, and has also been deployed in certain storage and clustered infrastructure designs as an alternative to Fibre Channel. More recently, InfiniBand has seen strong continued adoption in large-scale AI and GPU … Continue reading

This guide details installing NVIDIA drivers, CUDA Toolkit, and Container Toolkit on RHEL 10.1, utilizing simplified methods and new repositories for streamlined setup and verification processes.

CPU Overclocking on Ubuntu 24.04. In this post we will overclock the Intel Core i7-8086K Special Edition CPU - Intel's first CPU to hit 5.0Ghz out of the box.

Nvidia Datacenter GPU Manager (DCGM) facilitates GPU health monitoring, performance telemetry, and diagnostics for NVIDIA GPUs on servers. This post outlines installation and setup for RHEL 10, including driver validation and metric visualization.

Overview On some RHEL 10.1 installs users are running into this error, post-install, when attempting to install packages via dnf. Unsure if the issue is isolated to users attempting to install RHEL via the full DVD ISO, from the minimal boot ISO, and users deploying RHEL 10.1 via kickstart. More Details The issue is caused … Continue reading

This guide details setting up three LACP port-channels on a TP-Link SX3008F switch for RHEL 9/10 hosts, enabling 20Gbe connectivity for efficient NFS backups.

The shift from srvadmin (OMSA) to iSM (iDRAC Service Module) marks the end of bloated, “in-band” server management. This occurred between Dell 12th gen and Dell 13th gen server. If you have a 12th Gen Dell Server, you can still leverage Dell srvadmin (Idrac 7). I wrote a post on it here. While OMSA ran … Continue reading

Dell OpenManage Server Administrator (OMSA) is Dell’s on-host hardware management and monitoring framework for PowerEdge servers.  It runs inside the operating system and provides direct visibility into system hardware such as RAID controllers, physical and virtual disks, power supplies, fans, temperatures, memory, processors, and chassis health.  OMSA communicates with the server’s iDRAC and hardware controllers … Continue reading

I recently picked up a Netbotz 450 for almost nothing on Ebay. I have always wanted one of these things since I first deployed one in a SunGard facility in Philly about 12-15 years ago. These devices allow you to connect environmental sensors to monitor temperatures, rack intrusion, humidity, noise. You can also connect a … Continue reading

By default the Cisco ASAv has weak ssh config that many modern OS’s do not support. This error is from the client side ssh client refusing to use any of the host key algorithms that are offered by the ASAv. To get around this you can try the following. However as you can see from … Continue reading

Introduction The TP-Link TL-SX3008F is a 8-Port 10GE SFP+ L2+ Managed Switch (with some Layer 3 capabilities.) Its not too expensive, not too cheap, has decent reviews, and its rackmountable and fanless which is nice for a homelab. Details below Unboxing Arrived in a simple box and included the following Serial Connection In order to … Continue reading

This post details the setup of an NTP/Chrony server on Ubuntu 22.04 using a GPS receiver. It covers hardware requirements, package installations, GPSD configuration, firewall settings, and Chrony installation and configuration. Key functionalities include time synchronization and location sharing, all while ensuring secure incoming and outgoing network traffic.

This guide outlines the installation and configuration of Network UPS Tools (NUT) for monitoring a Tripp Lite USB PDU, enabling Zabbix integration. It covers prerequisites, file configuration, USB permissions, firewall settings, service management, and setting up Zabbix agent parameters, concluding with adjustments for SELinux to facilitate data collection.

Quick start guide on installing RHEL command-line assistant powered by RHEL Lightspeed

Chrony is a flexible and modern implementation of the Network Time Protocol (NTP). It is designed to synchronize the system clock with NTP servers, reference clocks, or manual input with high accuracy. It excels in environments where the connection to the time source is intermittent or when the system clock is frequently interrupted (such as … Continue reading

This guide will walk you through a factory reset and initial config as well as upgrade from OS8 to OS9 on the Dell Force10 S4810 switch

Step-by-Step Guide to install Nvidia Drivers, Nvidia CUDA Toolkit, & Container Toolkit on Red Hat Enterprise Linux 9

In this post we will walk through the process of upgrading the IOS image on a Cisco Catalyst WS-C4500X switch

Need to reset your Cisco Catalyst 4500X  back to the factory default settings. Follow this quick and easy guide

In this post we will install and configure dnsmasq with a webUI frontend to simplify manaing dns management

I recently got my hands on a couple of gigabyte servers. These machines came preinstalled with Ubuntu 20.04. Credentials for a OS local user account were on a sticker on the machines. However there was no indication of what the BMC credentials were. According to this document, there should be default credentials that we can … <a href="" class="more-link">Continue reading <span class="screen-reader-text"></span></a>

Prerequisites I have already installed Nvidia proprietary drivers and the Nvidia Cuda Toolkit. I documented the install of the Cuda toolkit in an older post which can be found here. Since I have Nvidia GPUs in my host system, and I intend to run some services in containers, I want to make sure that I … <a href="" class="more-link">Continue reading <span class="screen-reader-text"></span></a>

This guide details the installation process for Instructlab + VLLM on Ubuntu 22.04 with an Nvidia CUDA Support.

First determine what driver is installed. Below we can see that we are running the “nouveau” driver and not the Nvidia proprietary driver. Add rpm fusion repos. Dnf update. Install kernel-devel which provides kernel headers and makefiles required to build kernel modules. Install akmod-nividia, which checks if there are any missing kernel mods at boot … <a href="" class="more-link">Continue reading <span class="screen-reader-text"></span></a>

The NVIDIA CUDA Toolkit is essential for developing GPU-accelerated applications and includes various tools. Users must uninstall older versions from Ubuntu before installing the latest toolkit from NVIDIA’s website for better performance and compatibility. The summary highlights differences between installation methods regarding version availability, ease, and update management.

“The NVIDIA CUDA Toolkit is a software development kit that helps users create GPU-accelerated applications. It includes libraries, compilers, debuggers, and optimization tools”. Since we have an Nvidia GPU we will install it as shown below. Note: If you installed the nvidia-cuda-toolkit from default ubuntu noble repo, uninstall it first as that version is probably too old. … <a href="" class="more-link">Continue reading <span class="screen-reader-text"></span></a>

Introduction to RamaLama RamaLama is an open-source project developed to simplify AI model deployment and management using OCI (Open Container Initiative) containers. Ramalama enables seamless execution of AI workloads across different hardware configurations, supporting both GPU-accelerated and CPU-based environments. By leveraging container engines like Podman and Docker, RamaLama includes all necessary dependencies, eliminating complex installation … <a href="" class="more-link">Continue reading <span class="screen-reader-text"></span></a>