unable to pull ip from new dhcp

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

oddyager

Diamond Member
May 21, 2005
3,398
0
76
Originally posted by: Smilin
Ok, that network trace isn't making a lot of sense to me. Any chance you could paste the full frames in here? If you are immediately getting an error just a couple would do.

In your "trace" we're seeing a unicast discover come in meaning it was indeed sent from your relay.
Then it appears there is an ICMP destination unreachable sent back to the relay?? That makes no sense.

Right, you and me both. I don't know why its behaving like that either. I don't know how I can paste the full frames form Ethereal but basically what you are seeing is the broadcast, then followed immediately by an ICMP packet from DHCP to the relay agent going to port 67 and comes back destination unreachable. This happens 4 times and then the client stops requesting an address since it fails. This repeats of course after a period of time.

Yes, its active in AD and newserver can't service 10.x range since that is the old server's scope.
 

Smilin

Diamond Member
Mar 4, 2002
7,357
0
0
Originally posted by: oddyager
Right, you and me both. I don't know why its behaving like that either. I don't know how I can paste the full frames form Ethereal but basically what you are seeing is the broadcast, then followed immediately by an ICMP packet from DHCP to the relay agent going to port 67 and comes back destination unreachable. This happens 4 times and then the client stops requesting an address since it fails. This repeats of course after a period of time.

Yes, its active in AD and newserver can't service 10.x range since that is the old server's scope.

k, no new data means no new answers for ya, sorry. :p

To test on the local subnet do this:
1. Make and exclusion for a single IP on the old server.
2. Create a scope on the new server and exclude all but that IP.
3. Make a reservation for that IP on the new server so only the box you are testing with will pick it up.
4. Take a trace at the server as you ipconfig /release /renew at the test box.

This info will tell us if the DHCP server is hosed so we can eliminate the network config. It can't tell us if the DHCP server is good or the network is good though.

To grab those ethereal (wireshark actually I assume) frames here is one way.
1. Right click and mark all the packets you want. Grab some unicast and/or broadcast offers you see coming in plus one of those ICMPs plus anything else you think is relavent.
2. Select File | Export | File...
3. Set Packet details to All expanded. Set packet range to Marked Packets. Give it a filename.
4. Copy paste the contents of that textfile here.


If those DHCP discovers from the 20.x subnet are arriving at the new DHCP server in the 10.x subnet as broadcasts that would be bad. If they are relayed they would arrive as unicast. If you are seeing broadcast then that means broadcast info from 20.x is leaking over to 10.x and the DHCP server would try to issue a 10.x addy when it gets it. It's one of those common misconfigs I mentioned earlier.

I'm not sure what you mean by an ICMP to port 67 (icmp doesn't have ports). It will make sense when you paste those frames though.
 

dfnkt

Senior member
May 3, 2006
434
0
76
Are you trying to communicate between the VLANs? As bwatson said, routers wont forward layer3 (dhcp) broadcasts by default.....just out of curiousity, how are you doing inter-VLAN traffic, ROAS or do you have some layer 3 switches?
 

oddyager

Diamond Member
May 21, 2005
3,398
0
76
Okay, after some hair pulling, its resolved. There were two Windows issues:

1. The IP address wasn't statically assigned on the new server so won't give out dynamic addresses.

2. Conflict detection was enabled and that was why it was pinging the relay agent.

Address was set properly now and CDC was disabled and all is well.