Hey gang.

I finally finished switching all of my static IPs to a /22 network. I went into the DHCP mmc and did the following:

  1. Created the new reservations in the new scope (/22)
  2. Turned off the /24 scope
  3. Turned up the /22 scope

I went to workstations and did an ipconfig /renew… and things when splat. THe machines released their IPs fine, but then would time out and inevitably give me the time out error and auto assign the 169.254 address.

I went back to the server and turned off the nic that was on the /24 network in order to force all traffic onto the /22 network. Restarted the DHCP Server service. And still problems.

Turned on the nic on the /24 network again. Turned on the /24 scope and created an exception for that whole range figuring if a client couldn’t get an IP on the /24 network, it’d be forced to get a new IP on the new scope.

Still nothing. The odd thing is, some of the machines that had reservations on the /22 network, picked up the IPs fine…

Any thoughts? Or am I going to be working the weekend to resolve this before Monday?

Thanks!

3 Spice ups

If you don’t mind posting it, what is your Network ID for the /22? If you mind, then what class is it?

Do you have your switches and/or firewalls between the server and PC’s in the /22 subnet? What about the adapter on the DHCP server itself? Is it on the /22 subnet?

Old network … 10.211.10.x/24

New Network … 10.211.0.x/22

All switches have been moved to the 10.211.0.x/22 network

DHCP / DNS server had 2 adapters… one sitting on /24 and the other on /22.

Firewall served as the gateway for both networks and handled the routing between both networks.

I had a similar issue last summer, ended up being VLAN/IP configuration on my switches. They way the switches got setup over time VLAN server and passwords started to mismatch.

If it comes down to it, run wireshark on a client and a server to see if the DHCP requests are going through.

Ian… I thought about that … since all the switches had moved over to the 10.211.0.x/22 network, and the workstations were still on the 10.211.10.x/24 network, I wasn’t sure if the DHCP requests were passed… I enabled 3com’s UDP Helper just to make sure the switches passed the request to the DHCP server.

I came in today and the majority of the machines switched over. I’m still working on a few stragglers though.

Update… looks like I was a little impatient regarding this. All workstations (or at least 95% of them) have moved over. I think it has to do with the DHCP lease times, and some of them were still within the first 50% of the DHCP lease… even forcing the renew didn’t work.

Thanks for the place to look IanFagonOKS.

I have also found that sometimes you need to do ipconfig/release before you do ipconfig/renew.