Skip to content

Load Testing Tools ​

Category: Tools

The Azure Local Load Testing Framework includes five standalone load testing tools. Each tool targets a specific performance domain — storage, network, database, or system stress — and can be run independently from a management workstation with no CI/CD pipeline required.

Choosing a Tool ​

mermaid
flowchart TD
    A[What do you need to test?] --> B{Storage I/O}
    A --> C{Network}
    A --> D{Database}
    A --> E{CPU / Memory / Burn-in}
    B -->|Windows HCI host\nFleet-scale| F[VMFleet]
    B -->|Linux VMs\nFine-grained control| G[fio]
    C --> H[iPerf3]
    D --> I[HammerDB]
    E --> J[stress-ng]

Tool Summary ​

The Status column is the same for every tool today. The per-tool detail (implemented, unit-tested, proven on hardware) is in the implementation status.

ToolWhat It TestsTarget OSRun FromStatus
VMFleetStorage IOPS, throughput, latency via DiskSpd VM fleetWindows (HCI host)PowerShellImplemented, unit-tested, not proven on hardware
fioStorage I/O with fine-grained control (block size, queue depth, I/O engine)Linux VMsPowerShell + AnsibleImplemented, unit-tested, not proven on hardware
iPerf3Network bandwidth, jitter, packet loss between nodesLinux / WindowsPowerShellImplemented, unit-tested, not proven on hardware
HammerDBSQL Server OLTP (TPC-C) and OLAP (TPC-H) benchmarkingWindows VMsPowerShellImplemented, unit-tested, not proven on hardware
stress-ngCPU, memory, cache, and I/O stress for burn-in and capacity planningLinux VMsPowerShell + AnsibleImplemented, unit-tested, not proven on hardware

Running a Test (Standalone) ​

Every tool follows the same three-step pattern. No pipeline or CI/CD setup is needed — just a PowerShell terminal on your management workstation.

Step 1 — Install the tool on target nodes ​

powershell
.\tools\<tool>\scripts\Install-<Tool>.ps1 `
    -ClusterName "hci01.corp.infiniteimprobability.com" `
    -Nodes @("hci01-node1", "hci01-node2")

Step 2 — Run a test with a workload profile ​

powershell
.\tools\<tool>\scripts\Start-<Tool>Test.ps1 `
    -ClusterName "hci01.corp.infiniteimprobability.com" `
    -Nodes @("hci01-node1", "hci01-node2") `
    -Profile "general"

Step 3 — Collect results ​

powershell
.\tools\<tool>\scripts\Collect-<Tool>Results.ps1 `
    -ClusterName "hci01.corp.infiniteimprobability.com" `
    -Nodes @("hci01-node1", "hci01-node2") `
    -RunId "<run-id-from-previous-step>"

Each tool page below has the exact commands with real parameters.

Common Lifecycle ​

All tools share the same eight-phase lifecycle. When running standalone, you execute each phase manually. The CI/CD pipelines automate this same sequence if you want to schedule or trigger runs automatically.

PhaseWhat HappensManual Command
Pre-CheckValidate cluster connectivity and prerequisitesInitialize-Environment.ps1
InstallDeploy tool binaries to target nodesInstall-<Tool>.ps1
DeployConfigure test environment (VMs, volumes, etc.)Deploy-<Tool>.ps1 (VMFleet only)
TestExecute workload profileStart-<Tool>Test.ps1
MonitorCollect PerfMon / Azure Monitor metrics (parallel)Watch-<Tool>Monitor.ps1
CollectParse and aggregate raw resultsCollect-<Tool>Results.ps1
ReportGenerate PDF / DOCX / XLSXNew-TestReport (ReportGenerator module)
CleanupRemove test artefacts (optional)Remove-<Tool>.ps1

Automation (Optional) ​

If you want to automate test runs on a schedule or as part of a CI/CD workflow, see CI/CD Pipelines. The pipelines call the same scripts listed above — they are not a separate system.

Diagrams ​

Architecture and workflow diagrams are maintained as draw.io files in docs/assets/diagrams/ and exported as PNG images to docs/assets/images/ for documentation use.

DiagramSourceDescription
Solution Architecturesolution-architecture.drawioEnd-to-end framework overview
VMFleet Workflowvmfleet-workflow.drawioVMFleet pipeline phases
Config Flowconfig-flow.drawioYAML → JSON configuration pipeline
CI/CD Pipelineci-cd-pipeline.drawioGitHub Actions / Azure DevOps flow
Network Topologynetwork-topology.drawioCluster network paths for testing

NOTE

Exporting Diagrams To export updated PNGs from draw.io source files, the MkDocs drawio plugin handles this automatically during site builds. You can also export manually from draw.io desktop or VS Code extension.

Released under the MIT License.