I labbed a network (rented virtual via proctor labs) consisting of 4 Multi layer switches (3550), 4 routers (IP routing disabled and acting like hosts) and a DHCP server (R6)
I configured SVIS on all the MLS and setup OSPF advertising the 172.16.x.y/24 networks and the 46.46.46.0/24 between SW4 and R6. Please note I neglected to show the 46.46.46.0/24 network between CAT4 and R6 in the picture. CAT4 has the ip address 46.46.46.4/24 on its fast ethernet interface connected to R6. R6 is...you guessed it 46.46.46.6
see pic here
https://www.mediafire.com/?yy12k9c0ai74u5w
There was full reachability between all switches and DHCP server, sourcing from the various SVIs.
I then configured a DHCP pool on R6 so that it read
ip dhcp pool VLAN10
default-router 172.16.10.1
network 172.16.10.0 /24
ip dhcp pool VLAN20
default-router 172.16.20.3
network 172.16.20.0 /24
ip dhcp pool VLAN30
default-router 172.16.30.2
network 172.16.30.0 /24
ip dhcp pool VLAN40
default-router 172.16.40.2
network 172.16.40.0 /24
ip excluded-address 172.16.10.1 172.16.10.5
ip excluded-address 172.16.20.1 172.16.20.5
ip excluded-address 172.16.30.1 172.16.30.5
ip excluded-address 172.16.40.1 172.16.40.5
I applied the ip address dhcp command on the hosts and unshut the interface. However the hosts did not pick up a IP address from dhcp server so i then typed ip helper-address 46.46.46.6 on all the SVIS of the switches.
I ran debug ip dhcp server packet on R6 but although it appeared to be receiving dhcp packets I could not determine why it was not leasing any out
The funny thing is, is that I have done this same lab many many times with success but today (after a switching break of a few months) i could not get it working. My own conclusion is that there is something amiss with my configuration of ip helper-address?
any ideas?