1

Hi Everyone,

Hoping I might be able to get some help with a problem we’re currently facing. Very recently, we had our virtual machines on all three of our Hyper-V servers reboot, at the same time, for a reason we’ve yet to figure out. This event did not affect physical servers.

Since this has happened, clients on the network are not receiving an IP address from our DHCP server (everything works well with a static IP). We’ve also seen a lot of BAD ADDRESS entries pop up. I’m not quite sure where to go from here, having tried for a while now and was hoping somebody would be able to point me in the right direction?

Kind regards,

KL

3 Spice ups
2

A DHCP “bad address” typically means the IP that was offered to the client is already in use by another device. Depending on the status of your DHCP server it’s possible it has lost knowledge of all prior leases so it’s starting from scratch, however, the clients that were not rebooted are hanging onto previously issued IP addresses that the DHCP server isn’t aware of.

If you keep having bad address issues, reboot all devices that are receiving DHCP leases. Instead of reboot you can also do ipconfig /release followed by ipconfig /renew on Windows clients.

As to why all VMs were restarted at once - could it have been due to Windows updates? Either on the VMs themselves or maybe the Hyper-V had updates applied that required a restart of Hyper-V services?

3

Thank you for this information. I have released the IP and rebooted a good percentage of the devices that were causing problems but it doesn’t seem to have improved the overall situation. The biggest head scratch is that on our scope, we have over 200 IP addresses unallocated but not a single device is grabbing them. If I do an IPCONFIG /RENEW it will sit there until timeout. I’ve used Wireshark to best monitor what is happening, but I’m not seeing any offers. It’s certainly an interesting problem.

Those higher than me handle Hyper-V and the VMs, but I will pass on the suggestion about Windows Update.

4

Is your DHCP server a Windows server, possibly a Domain Controller? Investigate its logs through Event Viewer. Possibly the DHCP service isn’t running or is in an error state for whatever reason.

1 Spice up
5

Is it possible there is another DHCP server on the network all of the sudden? Check the DHCP clients for which server actually issued the lease via ipconfig /all and verify it is in fact your DHCP server. That could be another possible explanation for the BAD addresses.

6

The role is installed on Windows Server 2016, separate from AD. I’ve checked the DHCP audit logs and I can see clients renewing etc and to my surprise at this moment no IDs pertaining to anything bad. I did find in the event viewer this error:

PTR record registration for IPv4 address [[x.x.x.x]] and FQDN [NAME].[DOMAIN] failed with error 9005 (DNS operation refused.)

I’m not quite sure what to do about this.

7

Check your AD/DNS server for reasons why this might be happening.

8

it is usually okay to remove all BAD ADDRESS entries. they are probably causing your issue.

9

Hi Everyone,

Thank you for all the help so far, we’re slowly getting there. Does anyone perhaps know why DHCP would issue an IP address, which I can see in the leases and its record in DNS but the client still remains with an APIPA address? It’s almost like the DHCP server is offering an address to the client and making a record but the client’s IP, DNS and Gateway aren’t updating.

Kind regards,

KL

10

Thanks, everyone! All sorted!

11

What was the final fix?

1 Spice up
12

In DHCP scope options, 66 was configured and pointing to a server that didn’t exist.