From: Blanco Lam (b@gclamb.com)
Date: Wed Feb 05 2003 - 18:25:48 GMT-3
Hi Matt,
I think you can try a couple of things:
1) What happens when you move the PC 2/IP phone to the same switch as the
Server 2? Does it get an IP address then?
2) What happens when you move PC 1 into VLAN 2 and PC 2 into VLAN 1?
3) Try spanning the port of S1 and S2 and compare the difference. You
can use something like Ethereal to capture the packets for analysis.
Do the same for PC1 and PC2
If you say that the IP phone does a continous DHCP request, then I don't think
portfast is an issue. Yes, initially it may fail whilst it goes through
the different states until it gets into Forwarding, but thereafter, you should
be able to get an IP address.
Portfast on the trunk shouldn't matter either as long as no loops exist.
However, according the docs, even if a loop does exist, STP should move a link
back into blocking state to eliminate a loop. Yet it also says it may not
always work! ok.... :o)
Hope this helps.
Regards,
Blanco
---- Original message ----
>Date: Wed, 5 Feb 2003 19:09:39 -0000
>From: "Matthew Poole" <matthew.poole@blueyonder.co.uk>
>Subject: Re: Catalyst and DHCP forwarding.
>To: "Roberto Giana" <Roberto.Giana@econis.com>, <ccielab@groupstudy.com>
>Cc: <stephen.pennington@baileytelecom.co.uk>
>
>Hi,
>Thanks for replying.
>
>We discussed this earlier after somebody else suggested it, and he has tried
>it with portfast enabled - though he has tried that many things he's losing
>track!
>
>He has a non Cisco IP phone (there is such a thing ;-)) which repeatedly
>does DHCP requests indefinitely and doesn't time out like a PC which I feel
>discounts this theory. Also the config is exactly the same as with 12.0(5)
>and this issue didn't occur with that.
>
>Were you aware that there is a portfast command for trunks now, we spotted
>it in the latest switching release?
>
>http://makeashorterlink.com/?D26F15A53
>
>I believe he's on site tomorrow and is going to try your suggestion and
>others.
>
>If anybody else has any suggestions they would be appreciated.
>
>Thanks,
>Mat.
>----- Original Message -----
>From: "Roberto Giana" <Roberto.Giana@econis.com>
>To: "'Matthew Poole'" <matthew.poole@blueyonder.co.uk>;
><ccielab@groupstudy.com>
>Sent: Wednesday, February 05, 2003 7:00 PM
>Subject: RE: Catalyst and DHCP forwarding.
>
>
>> Hi Matthew
>>
>> Hmm.... I would recommend to turn off "spanning-tree portfast" on the
>trunks
>> and enable "spanning-tree portfast" on the access ports. The portfast
>> feature was intended for access ports, not for uplink ports.
>> It's a known issue, that clients can have problems with getting an IP when
>> booting up without the portfast feature activated on their access ports.
>> In your situation I would do further investigations after that.
>>
>>
>> Best regards
>> Roberto
>>
>>
>> -----Original Message-----
>> From: Matthew Poole [mailto:matthew.poole@blueyonder.co.uk]
>> Sent: Mittwoch, 5. Februar 2003 11:36
>> To: ccielab@groupstudy.com
>> Subject: OT:Catalyst and DHCP forwarding.
>>
>>
>> I realise it's a little off topic but thought I'd give you all a crack at
>> this
>> one.
>>
>> I have a colleague who has been configuring 2 2950's. They are connected
>> together via a dot1q trunk and they both have just 2 VLAN's. Two PC's on
>> separate VLAN's on S1 can get DHCP addresses from servers in S2 which are
>in
>> the same VLAN's - i.e. PC 1 and Server 1 in VLAN 1, PC 2 and server 2 in
>> VLAN
>> 2 - no intervlan routing.
>>
>> This all works fine on 12.05, however when he upgrades to 12.1 (he tried 4
>> versions) the PC in VLAN 2 can no longer get an IP address from the DHCP
>> server, VLAN 1 works fine. Interestingly, pings within VLAN 2 work fine if
>> he
>> gives the PC a static address.
>>
>> He has had this same issue with 3550's and 4500's as well on 2 other
>sites -
>> if anybody has done this uograde and had this same issue we'd love to here
>> from you and how you got round it.
>>
>> The configs are below, the only thing I can see are portfast on the trunk
>> and
>> no switchport mode access on the access ports. As I said previously it's
>> only
>> when he upgrades that he has these problems - they work fine on 12.05.
>>
>> SWITCH 1
>> no service pad
>> service timestamps debug uptime
>> service timestamps log uptime
>> no service password-encryption
>> !
>> hostname switch1
>> !
>> vtp mode transparent
>> !
>> vlan 2
>> vlan 3
>> !
>> interface FastEthernet0/1
>> description *** 802.1Q Port ***
>> switchport trunk encapsulation dot1q
>> switchport mode trunk
>> spanning-tree portfast
>> no ip address
>> !
>> interface FastEthernet0/2
>> description *** Access Port VLAN 1 ***
>> switchport access vlan 1
>> !
>> interface FastEthernet0/3
>> description *** Access Port VLAN 2 ***
>> switchport access vlan 2
>> !
>> Other ports as above
>> !
>> interface Vlan1
>> ip address 192.168.1.1 255.255.255.0
>> no ip route-cache
>> !
>> ------------------------------------------------------------
>>
>> SWITCH 2
>> no service pad
>> service timestamps debug uptime
>> service timestamps log uptime
>> no service password-encryption
>> !
>> hostname switch2
>> !
>> vtp mode transparent
>> vlan 2
>> vlan 3
>> !
>> interface FastEthernet0/1
>> description *** 802.1Q Port ***
>> switchport trunk encapsulation dot1q
>> switchport mode trunk
>> spanning-tree portfast
>> no ip address
>> !
>> interface FastEthernet0/2
>> description *** Access Port VLAN 1 ***
>> switchport access vlan 1
>> !
>> interface FastEthernet0/3
>> description *** Access Port VLAN 2 ***
>> switchport access vlan 2
>> !
>> Other ports as above
>> !
>> interface Vlan1
>> ip address 192.168.1.2 255.255.255.0
>> no ip route-cache
>> .
>.
.
This archive was generated by hypermail 2.1.4 : Sat Mar 01 2003 - 11:06:11 GMT-3