DHCP Address not being given by Domain Controller
We have a multi-domain organisation that has multiple DC's. One of these DC's (in Manchester) does the DHCP for all of the connected sites except one; the Cardiff site that contains a second DC for a second Domain.
We have successfully migrated each of the satellite locations on that second Domain to pick up DHCP from the Manchester DC but pick up DNS from the Cardiff DC. This works fine and has been running without issue for months.
However, due to the ending of an ISP contract, we wanted to stop the Cardiff DC supplying DHCP addresses to machines in the Cardiff site and have them pick up those addresses from Manchester instead on a line set up with our new ISP. So, we set up a valid DHCP scope in Manchester, switched off the DHCP Service in Cardiff.... and nothing happened.
Our ISP can see the DHCP requests being made by the machines in Cardiff and can see them hit the Router in Manchester and be sent to the Manchester DC - but that's it. They never see a response from the Manchester DC to DHCP requests from the Cardiff site!
All other sites request DHCP from Manchester and get an almost instant response, but we just can't get it working from Cardiff. As a temporary measure, we are assigning DHCP from the Cardiff Router instead in order to get rid of the defunct network connection.
Does anyone know of a reason why our DC would simply refuse to assign DHCP addresses to the Cardiff site alone? The DHCP range set up is the same as was set up in Cardiff - and is the same as we are assigning from the Router. All of the settings seem to be right, but it just won't do it and we're not sure why. Clearly the requests are being sent to the Server as our ISP has followed the requests that far, so the problem must be with the DC itself somewhere - but it works everywhere else?????
Confused!
We have successfully migrated each of the satellite locations on that second Domain to pick up DHCP from the Manchester DC but pick up DNS from the Cardiff DC. This works fine and has been running without issue for months.
However, due to the ending of an ISP contract, we wanted to stop the Cardiff DC supplying DHCP addresses to machines in the Cardiff site and have them pick up those addresses from Manchester instead on a line set up with our new ISP. So, we set up a valid DHCP scope in Manchester, switched off the DHCP Service in Cardiff.... and nothing happened.
Our ISP can see the DHCP requests being made by the machines in Cardiff and can see them hit the Router in Manchester and be sent to the Manchester DC - but that's it. They never see a response from the Manchester DC to DHCP requests from the Cardiff site!
All other sites request DHCP from Manchester and get an almost instant response, but we just can't get it working from Cardiff. As a temporary measure, we are assigning DHCP from the Cardiff Router instead in order to get rid of the defunct network connection.
Does anyone know of a reason why our DC would simply refuse to assign DHCP addresses to the Cardiff site alone? The DHCP range set up is the same as was set up in Cardiff - and is the same as we are assigning from the Router. All of the settings seem to be right, but it just won't do it and we're not sure why. Clearly the requests are being sent to the Server as our ISP has followed the requests that far, so the problem must be with the DC itself somewhere - but it works everywhere else?????
Confused!
ASKER
I have checked the Logs. One of the machines that attempted to get a DHCP Address (this was last week!) has the result below in the Log:
I assume that this means that the Server attempted to assign the IP Address 192.168.1.171? If so, this address was never given to the machine and the ISP never saw the attempted assignment being sent from the Server.
30,02/18/15,11:10:27,DNS Update Request,192.168.1.171,FS-LAP-ICT-PJ2.FSL.local,,,0,6,,,
10,02/18/15,11:10:27,Assign,192.168.1.171,FS-LAP-ICT-PJ2.FSL.local,D067E53CC917,,2020114998,0,,,I assume that this means that the Server attempted to assign the IP Address 192.168.1.171? If so, this address was never given to the machine and the ISP never saw the attempted assignment being sent from the Server.
Wow that's strange, it means the request came, was SENT OUT, but THEN blocked somewhere along the way. Now the difficult part is still, who/what blocked the ACK packet from the DHCP server. The only way to do that is to follow the physical route from the server to the laptop at the other side, and for every section, check if the packet is still there (needs managed router, and a seperate PC with Wireshark). To make it simple, do the packet sniffing on the cable that plugs into the ISP/router. If the ACK packet is still there, it's means it's being filtered by the ISP, and you can do nothing at all.
I have a few questions:
Are these locations on a different subnet?
Has the ISP configured a DHCP relay agent on their router?
Are there any VLAN's in play here?
Is the scope for the Cardiff site configured correctly (has the scope option 003 for the Cardiff router)?
-saige-
Are these locations on a different subnet?
Has the ISP configured a DHCP relay agent on their router?
Are there any VLAN's in play here?
Is the scope for the Cardiff site configured correctly (has the scope option 003 for the Cardiff router)?
-saige-
ASKER
To answer IT_SAIGE:
All routes are on a 255.255.255.0 subnet (192.168.1.x and 192.168.77.x)
I can only assume that the ISP has configured a DHCP relay if it works for all of the other sites (some 30-odd sites get DHCP from Manchester)
No VLAN's anywhere.
We certainly believe that the scope is correct - it does have Option 003 for the Cardiff Router.
All routes are on a 255.255.255.0 subnet (192.168.1.x and 192.168.77.x)
I can only assume that the ISP has configured a DHCP relay if it works for all of the other sites (some 30-odd sites get DHCP from Manchester)
No VLAN's anywhere.
We certainly believe that the scope is correct - it does have Option 003 for the Cardiff Router.
Do you have the capability to setup port monitoring on the switch that your DHCP server is connected to? If you do, any chance that you can perform a packet sniff using Wireshark (or any other available packet sniffers)? You will want to monitor the traffic on both sides (one in Cardiff and the other in Winchester). I would initially setup to just monitor the ports that a test computer and the DHCP server are attached to. Just to ensure that the packets are being sent and received on both sides (could potentially be the client where the DHCP client service is acting a little flaky).
-saige-
-saige-
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
If possible, when you get this figured out could you post back with the root cause and resolution?
Thanks,
-saige-
Thanks,
-saige-
ASKER
A poor workaround - nothing more.
If you really see the requests (identifiable by mac address), followed by NACK, it means your server refused it. But I highly doubt it, and therefore, it's blocked by a switch or a router. Could be your hardware, meaning it's "your fault", could be a router from the ISP, meaning it's "their fault". A bit difficult to tell right now.
For now, until it's solved, just put a simple DHCP server on the second site (have a range that's excluded from the DC DHCP), even a simple PC will do (I usually use Dual DHCP DNS Server from SourceForge.net if I need some quick testing).