Mitel DHCP disaster

Apr 11, 2006 3 Replies

G'day everybody!



I'm currently facing a tough challenge. We have built the following: Mitel 3300 ICP with 5212 IP phones on an HP ProCurve layer 3 PoE switch. On the switch, we made two VLAN's (VLAN 1 for data, VLAN 2 for voice), of which the second has tagged ports for QoS. The PC's are physically linked by the phones, acting as a hub for interconnectivity.


When rebooting the phones, and thus requesting for DHCP scope, this is not functioning (reject 1 thru 10). The PC's do get a correct IP address. After untagging the ports in VLAN 2, DHCP was working properly. This situation, however, is not helping anyway, since tagged ports are necessary for QoS purposes.


Investigation shows that the phones send untagged packets when requesting DHCP, this is fixed in the firmwear of the phones.


Can anyone provide new insights on this please?


Thanks!


Johan, project leader



The phones will initially send an untagged request to the DHCP server. Your VLAN 1 DHCP server must contain certain options.(More then likey you'd want to be running Options 128-133) When the phones recieve this vlan1 address and see these options, they will then drop the address and switch to tagged vlan 2, and get a new address from your vlan 2 dhcp server. Make sense?

It's normal behavior. Check you Procurve switch to mark all untagged frames with VLAN2. Usually when you set port as a trunk, you should be able to set "native VLAN" - VLAN TAG should be used if no VLAN TAG is set.

Another thing - in most switches VLAN1 is management and/or internal VLAN, and it's not-recommneded to use it for any user traffic.

Good luck,

Mike

Cheers guys!

Problem solved! We just made the mitel dhcp server for vlan2, and the win2k server for vlan1, works like sunshine:-))

Grtz Johan

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required