stress-ng — System Stress Testing¶

This page has moved

The main stress-ng landing page with run instructions is now at stress-ng. This page contains additional technical details.

Purpose¶

CPU, memory, cache, and system stress testing for capacity planning and burn-in validation.

When to Use¶

CPU stress testing for thermal and power validation
Memory stress testing for stability under pressure
Burn-in testing of new cluster nodes
Capacity planning — determine maximum safe utilization levels

Key Capabilities¶

370+ stress tests across CPU, memory, filesystem, network categories
YAML job files for reproducible configurations
Metrics: bogo ops/s, thermal zones, power measurement
Stressor class system for targeted testing

Prerequisites¶

In addition to the common prerequisites:

Linux VMs deployed on the Azure Local cluster
Ansible 2.14+ for deployment
stress-ng package installed on target VMs (automated via Ansible role)

Quick Start¶

# Run all-core CPU stress on every node
.\tools\stress-ng\scripts\Start-StressNgTest.ps1 `
    -ClusterName "hci01.corp.infiniteimprobability.com" `
    -Nodes @("hci01-node1", "hci01-node2") `
    -Profile "cpu-stress"

# Collect bogo-ops metrics
.\tools\stress-ng\scripts\Collect-StressNgResults.ps1 `
    -ClusterName "hci01.corp.infiniteimprobability.com" `
    -Nodes @("hci01-node1", "hci01-node2") `
    -RunId "<run-id-from-previous-step>"

File Locations¶

Component	Path
Scripts	`tools/stress-ng/scripts/`
Profiles	`tools/stress-ng/config/profiles/`
Alert Rules	`tools/stress-ng/monitoring/alerts/alert-rules.yml`
Report Templates	`tools/stress-ng/reports/templates/`
Tests	`tools/stress-ng/tests/`

stress-ng Documentation¶

Document	Description
Installation	Install stress-ng on Linux VMs
Workload Profiles	CPU, memory, and I/O profile configuration
Monitoring	Alert rules active during stress-ng runs
Reporting	Result collection and report generation
Troubleshooting	Common issues and resolutions