local-ci-cd-system/plans/benchmark-windows-host.md

Plan — Windows-host benchmarks (comparable to Linux-host §6–§10)

Scopo: la macchina è dual-boot (Windows ⇄ Linux, stesso hardware). Le baseline RUNBOOK §6–§10 sono quasi tutte sul Linux host. Questo plan raccoglie i numeri mancanti sul Windows host per chiudere la matrice host × guest × IP-mode e rendere il confronto Linux-vs-Windows completo e simmetrico.

Prerequisito: riavviare la macchina nel boot entry Windows. Tutto gira su Windows PowerShell 5.1 o pwsh 7; usare la stessa Measure-CIBenchmark.ps1 / Test-CapacityBurnIn.ps1 del repo (sono cross-platform, i default path passano a F:\CI\... quando $IsWindows).

---

1. Cosa esiste già e cosa manca

RUNBOOK §	Host	Guest	IP mode	Stato
§6	Win	WinBuild2025	DHCP	✓ 2026-05-17
§7	Linux	WinBuild2025	DHCP	✓
§8	Linux	WinBuild2025	static	✓
§9	Linux	Win + Linux	burn-in (static pool)	✓
§10	Linux	LinuxBuild2404	DHCP	✓

Mancanti sul Windows host (target di questo plan):

Nuovo §	Host	Guest	IP mode	Pair-with
§11	Win	WinBuild2025	static	§8
§12	Win	LinuxBuild2404	DHCP	§10
§13	Win	Win + Linux	burn-in 4×10	§9

§6 copre già Win×Win×DHCP. Opzionale: ri-eseguirlo nella stessa sessione per coerenza (stesso disco/fragmentazione del momento).

---

2. Prerequisiti sul Windows host

• Boot in Windows; verificare VMware Workstation + vmrun.exe in C:\Program Files (x86)\VMware\VMware Workstation\vmrun.exe.
• Template fully powered-off con snapshot integri: powershell & 'C:\Program Files (x86)\VMware\VMware Workstation\vmrun.exe' -T ws listSnapshots 'F:\CI\Templates\WinBuild2025\WinBuild2025.vmx' & 'C:\Program Files (x86)\VMware\VMware Workstation\vmrun.exe' -T ws listSnapshots 'F:\CI\Templates\LinuxBuild2404\LinuxBuild2404.vmx' Attesi: BaseClean e BaseClean-Linux. ✓ (anche PostInstall su Win)
• Credenziali in Windows Credential Manager (non keyring file): target BuildVMGuest (guest WinRM, user ci_build) e GiteaPAT (user Simone). Verificato.
• Chiave SSH guest Linux presente: F:\CI\keys\ci_linux (per §12/§13 lato LinuxBuild2404).
• Gitea raggiungibile (http://10.10.20.11:3100, HTTP 200) e repo Simone/burnin-dummy con build.ps1 + build.sh (verificato via API).
• F:\CI\BuildVMs\ privo di cloni orfani e F:\CI\ip-pool.json assente (pulito). Reset se serve: powershell Set-Content -Path 'F:\CI\ip-pool.json' -Value '{}' -NoNewline

---

3. §11 — WinBuild2025 static IP (pair §8)

.\scripts\Measure-CIBenchmark.ps1 `
  -TemplatePath 'F:\CI\Templates\WinBuild2025\WinBuild2025.vmx' `
  -SnapshotName BaseClean `
  -StaticIP 192.168.79.200 -Netmask 255.255.255.0 `
  -Iterations 10

• 10 iter completate, tabella per-fase stampata. Boot total avg 29.75 s.
• Confronto vs §8 (Linux host static): IP-floor confermato host-independent (22.81 s vs 21.77 s, +5 %). Penalità host: clone +58 % (NTFS), Ready +4.7 s (WinRM re-bind dopo NIC reconfig), destroy +82 % (bimodale), boot +24 %. Vedi RUNBOOK §11. Nota: primo tentativo fallito (timeout 300 s/iter) — template Windows privo dell'agent ci-static-ip; risolto copiando i template dal Linux host (parità template).

4. §12 — LinuxBuild2404 DHCP (pair §10)

.\scripts\Measure-CIBenchmark.ps1 `
  -TemplatePath 'F:\CI\Templates\LinuxBuild2404\LinuxBuild2404.vmx' `
  -SnapshotName BaseClean-Linux `
  -GuestOS Linux `
  -Iterations 10

• 10 iter completate. Boot total avg 13.51 s.
• Confronto vs §10 (Linux host, 8.82 s): +53 % host-side — clone +191 % (NTFS), IP-acquire +52 % (NAT vmnet8 Windows + guestinfo). Guest Linux resta deterministico (IP σ ≈ 0.05 s). Vedi RUNBOOK §12.

5. §13 — Concurrent burn-in 4×10 (pair §9)

# WinBuild2025
.\scripts\Test-CapacityBurnIn.ps1 `
  -Parallelism 4 -Rounds 10 `
  -TemplatePath 'F:\CI\Templates\WinBuild2025\WinBuild2025.vmx' `
  -RepoUrl 'http://10.10.20.11:3100/Simone/burnin-dummy.git' `
  -CloneBaseDir 'F:\CI\BuildVMs\' `
  -ArtifactBaseDir 'F:\CI\Artifacts\' `
  -VmrunPath 'C:\Program Files (x86)\VMware\VMware Workstation\vmrun.exe' `
  -BuildCommand 'powershell -ExecutionPolicy Bypass -File .\build.ps1'

# LinuxBuild2404  (snapshot OBBLIGATORIO, default è BaseClean=Windows)
.\scripts\Test-CapacityBurnIn.ps1 `
  -Parallelism 4 -Rounds 10 `
  -TemplatePath 'F:\CI\Templates\LinuxBuild2404\LinuxBuild2404.vmx' `
  -SnapshotName BaseClean-Linux `
  -RepoUrl 'http://10.10.20.11:3100/Simone/burnin-dummy.git' `
  -CloneBaseDir 'F:\CI\BuildVMs\' `
  -ArtifactBaseDir 'F:\CI\Artifacts\' `
  -VmrunPath 'C:\Program Files (x86)\VMware\VMware Workstation\vmrun.exe' `
  -GuestOS Linux `
  -BuildCommand 'bash build.sh'

• [~] Linux 40/40 PASS (10/10 round). Windows 36/40 (7/10 round). I 4 fail Windows = fault WinRM transitori (round 6–8: connect-timeout 5986 + WSManFault "shell not found") sotto 4× concorrenza; auto-recuperati (slot IP rilasciati, 0 cloni orfani, round 9–10 di nuovo puliti). Linux/SSH 0 fault.
• Round avg vs §9 (Win ~78.6 s, Lin ~70.2 s): Windows host Win ~91.6 s (round PASS) +16 % + instabilità WinRM; Lin 96.4 s +37 %. Vedi RUNBOOK §13. Pre-req scoperti: (1) parità template, (2) prod venv F:\CI\python\venv stale (pre-ip_pool) → reinstallato pip install ., (3) CI_VENV_PYTHON da settare (bug shim $IsWindows su PS 5.1+StrictMode).

---

6. Output e documentazione

• Tutti i run appendono a F:\CI\Logs\benchmark.jsonl (single-job) e stampano summary (burn-in).
• Risultati aggiunti a docs/RUNBOOK.md §11–§13 (+ §14 matrice completa), con confronto host Linux-vs-Windows per ciascuna coppia.
• Pulizia post-run: F:\CI\BuildVMs\ vuoto, artifact burnin-* rimossi, F:\CI\ip-pool.json reset.

---

7. Matrice finale attesa (dopo questo plan)

Host \ (Guest, mode)	Win/DHCP	Win/static	Lin/DHCP	burn-in 4×10
Linux	§7	§8	§10	§9
Windows	§6	§11	§12	§13

Confronto colonna-per-colonna = effetto host a guest+mode costanti. Confronto riga-per-riga = effetto guest/IP-mode a host costante.