← Back to inventory

Lab tool · Linux

Linux Endpoint Bench

Benchmark orchestrator for Ubuntu workstations that measures the impact of security agents (antivirus, EDR, DLP) on CPU, disk, memory, boot and applications.

Role
Sole author: orchestrator, test suites and operations documentation
Outcome
Approving a security agent now relies on repeatable measurements with real reboots and a controlled environment.

Coordinates 19 tests across 6 domains — including a boot battery with real reboots orchestrated via systemd — to compare a clean Ubuntu workstation against scenarios running corporate security agents.

Before running, a preflight check validates dependencies and power profiles, a doctor command diagnoses the environment and the orchestrator suspends automatic updates to prevent noise during collection. Alongside hyperfine microbenchmarks, it runs real office workloads by automating LibreOffice via UNO and browser flows with Playwright.

Built end-to-end with Spec-Driven Development (30+ specs shipped). The code is private; happy to show it in a conversation.

Illustration of the Linux Endpoint Bench terminal running preflight checks and the 6 test domains on Ubuntu
Illustration of the Linux Endpoint Bench terminal running preflight checks and the 6 test domains on Ubuntu