こんにちは、大下立さん、
Thank you for posting question on Microsoft Windows Forum.
Based on the issue description, there are some reports in community about large piles of non‑present (“ghost”) devices which could accumulate on session hosts, especially in environments with frequent VM redeploys, driver updates, or ephemeral NIC churn.
Basically, it can consider to be safe for bulk-cleaning the Microsoft Hyper-V Virtual PCI Bus ghost devices, but it is crucial important to have test first (on one session-host image / snapshot), export lists and backups, and use a dry-run before doing deletes. The below basic method is for your reference.
- Get-PnpDevice **を使用して非表示/存在しない Hyper-V デバイスを列挙する **
- 削除する予定の InstanceId を確認します
- ドライランをログに記録してサポートする制御されたスクリプトで pnputil /remove-device を使用して削除します。
- 詳細については 、https://learn.microsoft.com/en-us/powershell/module/pnpdevice/get-pnpdevice?view=windowsserver2025-ps
以下の記事は、あなたの懸念に関するさらなる参考に役立つかもしれません。
- https://learn.microsoft.com/en-us/troubleshoot/windows-server/virtualization/unknown-device-vmbus-virtual-machine
- https://learn.microsoft.com/en-us/windows-hardware/drivers/devtest/pnputil-command-syntax
- https://superuser.com/questions/1776942/how-to-bulk-remove-all-hidden-devices-in-device-manager-sample-powershell-scri
上記の情報がお役に立てば幸いです。