DHCP Relay results in Finding Router error

[kholladay]

Here is what I have:
Avaya 4621 connected to an Extreme Networks Summit 400-48 in turn connected to a Windows 2003 DHCP/TFTP server.

VLAN-10 has the server and all ports as untagged.
VLAN-11 has the phone(s) and all ports as tagged "11".

Option 176 on VLAN10 is set as "L2Q=1,L2QVLAN=11"
Option 176 on VLAN11 is set as "MCIPADD=10.254.11.11...etc"

The Extreme Summit has "enable bootprelay" and "configure bootprelay add 10.254.10.10" configured. I have also experimented with the "...bootprelay dhcp-agent option..." settings with the same result.

"Conflict Detection Attempts" on the Windows 2003 DHCP server is set to 3.

When the phone boots it assumes VLAN0 and pulls an IP of 10.254.10.xxx but then releases it pre the option 176 to obtain an IP on VLAN11. However after the release it re-obtains the same IP on 10.254.10.xxx not 10.254.11.xxx but with the VLAN set to 11 resulting in a "Finding Router" state on the set.

If I set the port on the switch to VLAN11 and untagged the phone boots correctly, this is only an issue when the port is set to VLAN tagged.

Kyle Holladay
Certified: ACACN, ACSCI, ACSCM, TIA-CTP
"If it worked the way it should you wouldn't need me

 

[krad]

We have the same problem after swapping the W2K server with 2003 in one of our offices. (access switch = cisco 3560)

Although in our main office we swapped the dhcp server without any problem (access switch = 2950)

Who did have the same problem and found the solution?

 

[krad]

Our problem is solved!

When a W2K3 DHCP server is connected on a PC/Phone Trunk port then dhcp will not work!!!
(This is not a problem for W2K DHCP servers)

Our former config of the server port on the Cisco 3560
switch :
interface FastEthernet0/3
description *** SERVER ***
switchport access vlan 100
switchport mode access
switchport nonegotiate
switchport voice vlan 500
srr-queue bandwidth share 10 10 60 20
srr-queue bandwidth shape 10 0 0 0
priority-queue out
mls qos trust cos
no mdix auto
auto qos voip trust
no cdp enable
spanning-tree portfast trunk

Our new config of the server port on the Cisco 3560
switch :
interface FastEthernet0/3
description *** SERVER ***
switchport access vlan 100
switchport mode access
switchport nonegotiate
srr-queue bandwidth share 10 10 60 20
srr-queue bandwidth shape 10 0 0 0
priority-queue out
mls qos trust cos
no mdix auto
auto qos voip trust
no cdp enable
spanning-tree portfast


Regards

 

[kholladay]

The DHCP server is connected to port 1 on the Summit48 which is set to VLAN10 untagged. There is no phone or other switch/hub on that port. The phones are connected to ports 12-22 on the Summit48 which are set to VLAN10 untagged and VLAN11 tagged. The PCs connected to the phones successfully pull IP in the scope for VLAN10 from DHCP, as do the phones (which they shoudln't as they are tagging on VLAN11).

Kyle Holladay
Certified: ACACN, ACSCI, ACSCM, TIA-CTP
"If it worked the way it should you wouldn't need me