VR HealthCheck fails when there are no VMs deployed

### problem

A VR HA pair or single VR report the health check as failed when there are no VMs deployed to that VR. This causes alerts in the logs, triggering investigations.

Under the VR health check:
"dhcp_check.py" reports "No VMs running data available, skipping" (which is good)
"webserver.service" reports "warning" "service down at last check xxxxxxx" (which is bad)

This issue goes away if a VM is deployed and returns if it is deleted. If there is an VR HA pair, the failure report bounces back and forth between the 2 VR in the HA pair.

The attached screenshot has r-517/r518 in an HA pair as well as r-519/r-520.

### versions

This is on CloudStack 4.22.

### The steps to reproduce the bug

1. Create a VR Router (HA or not) and an isolated network
2. Deploy a VM to that isolated network
3. Delete the VM
4. The next health check will then report as failed

### What to do about it?

The VR healthcheck should not fail if there are no VMs deployed.

dhcp_check.py と webserver.service の health-check パスから始め、次に、その VM を削除した後、ネットワークを分離した VR で問題を再現します。VM がデプロイされていない VR が正常な状態を維持し、報告された failure alert を出力しないこと（HA ペアの場合も含む）が完了の条件です。

索引モデルが issue の本文から書いたものです。