Skip to content
FSS Knowledge Base

System-Maintenance — Status

Section titled “System-Maintenance — Status”

Type: Ongoing project (KB-native — no repo) Owner: Core/IT dept (Director, A1) Home: Core/IT/Projects/System-Maintenance/ — this folder is the SSOT + portal (DASHBOARD.md)

Current State

Section titled “Current State”

Kickoff fully closed out (2026-07-21 → confirmed 2026-07-22). All findings resolved and Talbot’s signed off every Next Step: ESU enrolled, WSL vhdx compacted, .wslconfig left uncapped (deliberate), Azure Data Studio/DbSchema uninstalled. Project now sits in Ongoing cadence (Roadmap phase 5) — no open items. Next trigger: AI-workstation build scoping (new task when that becomes real).

Issues — resolved

Section titled “Issues — resolved”
#IssueStatus
1C: drive space✅ Resolved. 7.9GB → 52GB free (purge + kopia relocation + Talbot’s own app cleanup). Well above the 20GB diskcheck threshold.
2diskcheck monitor reliability✅ Resolved. Root cause: bare uv path + RunLevel: Highest broke PATH resolution for the elevated Task Scheduler action. Fixed: full path (C:\Users\Admin\.local\bin\uv.exe) + RunLevel: Limited. Verified: manually triggered DiskSpaceCheck — LastTaskResult: 0, new log line 2026-07-21 17:57:19 - OK: C: 30.5GB, D: 152.7GB.
3Windows 11 upgrade✅ Decision made. PC Health Check confirms unsupported (CPU: Ryzen 7 1700, pre-dates Win11’s supported list). Recommendation: enroll in Consumer ESU (free) as the security-patch bridge; defer any hardware decision to the future AI-workstation build (which resolves this as a side effect). See Core/IT/Logs/2026-07-21_Windows11-Upgrade-Decision.md.
4C: SSD → 1TB upgradeFolded into the Windows11-Upgrade-Decision reasoning — no standalone purchase, deferred to the eventual AI-workstation build
5diskcheck util hardening🟢 Lower priority, no longer blocked — diskcheck now works, hardening is a nice-to-have, not urgent
6kopia cache relocation✅ Done via my_backup-kopia-cache-relocation — ~12GB freed
7Recurring C: maintenance (new, 2026-07-21)✅ Automated: MonthlyDriveCleanup scheduled task (weekly, RunLevel: Limited, verified via trigger, script relocated to D:\FSS\Software\Utils\Windows\ after Talbot’s correction). WSL vhdx --set-sparse fix and .wslconfig memory cap flagged, not executed — see Core/IT/Logs/2026-07-21_C-Drive-Purge-Plan.md.

Next Steps

Section titled “Next Steps”
  • Enroll in Windows 10 Consumer ESU (free tier via Microsoft account + Windows Backup sync) when convenient — no downside, extends security patches to Oct 2026
    1. done; C:\tmp\ScreenShots\explorer_wQzI3oAICW.png
  • Compact the WSL virtual disk, once (~9GB reclaimable on C:) — not automated on purpose: it requires wsl --shutdown, which kills every running WSL instance including whatever’s driving this session, and there’s no safe unattended trigger for an operation that can only run when WSL is stopped. Run this once yourself, from a plain Windows terminal (not a WSL one) — the --set-sparse true part also means it largely self-maintains afterward (space reclaims automatically as WSL deletes files), so this shouldn’t need repeating on a schedule:
    Terminal window
    wsl --shutdown
    wsl --manage Ubuntu-24.04 --set-sparse true
    1. done
  • Decide on .wslconfig memory cap — no config exists today (WSL defaults to ~50% host RAM, ~15GB observed), which already matches the “left uncapped for dev work” decision in LESSONS.md. The purge-plan’s suggested memory=8GB would be a real cut, not neutral cleanup. Confirm if you actually want that trade-off — not applied without your say-so.
    1. leave as is
  • Uninstall Azure Data Studio + DbSchema if truly unused — via Settings → Apps → Installed apps (not manual folder deletion, so registry/service entries clean up properly). Nothing found tying either to other business tooling here — looks safe to remove if you don’t use them.
    1. done
  • Whenever the AI-workstation build becomes real: new task to scope it (CPU/mobo/RAM/GPU/SSD)
    1. will do