Home/Toolchest

Windows Update Repair Checks

A staged Windows Update troubleshooting path that starts read-only and escalates only when needed.

Good For

  • Windows 10/11 update failures
  • Server patching
  • 0x80070490
  • component store repair

How to Use It

  1. Record the exact update KB, error code, and install phase.
  2. Stage 1 is read-only evidence: check free disk space, pending reboot state, component store health, Windows version, and Windows Update service state.
  3. Use DISM CheckHealth and ScanHealth before any repair action so the ticket records whether corruption was detected.
  4. Stage 2 is repair work: run RestoreHealth and SFC only inside an approved maintenance window with backup, restore point, VM snapshot, or change-ticket context.
  5. Restart Windows Update services only after capturing the initial state and only when service restart is part of the approved repair plan.
  6. Record the final DISM/SFC result and the Windows Update retry result as validation evidence.

Execution Modes

  • local
  • remote-single-host

Inputs and Outputs

Inputs

  • target computer
  • KB number
  • error code
  • maintenance window

Outputs

  • verbose-console
  • operator-notes

Command Starter

Changes system state: review before running

DISM /Online /Cleanup-Image /CheckHealth
DISM /Online /Cleanup-Image /ScanHealth
Get-Service wuauserv,bits,cryptsvc
Get-ComputerInfo | Select-Object OsName, OsVersion, WindowsVersion
DISM /Online /Cleanup-Image /RestoreHealth
sfc /scannow

Validation

  • Read-only evidence clearly separates component-store health from service state and update error code.
  • DISM RestoreHealth and SFC complete without unresolved corruption when repair is approved.
  • Windows Update can rescan and attempt the target update again.
  • The original error code no longer appears for the same KB.

Reporting

  • repair transcript
  • patching evidence note
  • future CSV status export

Safety Notes

  • CheckHealth, ScanHealth, Get-Service, and Get-ComputerInfo are evidence collection.
  • RestoreHealth and sfc /scannow are repair actions and should not be presented as read-only.
  • Rollback is limited for component repair. Capture a restore point, VM snapshot, backup, or change ticket context before repair on production systems.
  • Record the starting service state and latest restore point or backup option before running repair commands so you can revert to the previous recovery path if validation fails.