Re: bridging removes port from 3550 vlan - FIXED

From: P729 (p729@cox.net)
Date: Wed Jul 02 2003 - 15:35:45 GMT-3


Sage advice.

Every trunking-capable Catalyst and ProCurve switch I've run across will, by
default, form a trunk (and EtherChannel) if asked to.

Just curious, did you also try 'debug dtp' ?

Regards,

Mas Kato
https://ecardfile.com/id/mkato

----- Original Message -----
From: "Jonathan V Hays" <jhays@jtan.com>
To: <ccielab@groupstudy.com>
Sent: Wednesday, July 02, 2003 11:01 AM
Subject: RE: bridging removes port from 3550 vlan - FIXED

> -----Original Message-----
> From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On
> Behalf Of Jonathan V Hays
> Sent: Monday, June 30, 2003 12:35 PM
> To: ccielab@groupstudy.com
> Subject: bridging removes port from 3550 vlan
>
> My lab study partner has a pair of catalyst 3550s running the
> latest EMI
> image, (C3550-I5Q3L2-M), Version 12.1(13)EA1a.
>
> Whenever we configure DLSW (or simple bridging) the fastethernet 3550
> switch ports connected to the DLSW-enabled routers are
> removed from the
> configured VLAN. You can execute "show vlan" on the switch
> and the DLSW
> ports are missing. Removing the DLSW configuration from the router
> restores the port to the VLAN. It goes without saying, but I'll say it
> anyway, we have tried various ports and rebooted switches and routers
> but that does nothing to fix this problem.
>
> The Cisco Bug Toolkit did not give me any hits on this problem.
>
> Does anyone have any ideas on this one?
>
> Thanks,
>
> Jonathan
> ______________________________________________________________

Thanks to everyone for their suggestions. The problem is FIXED.

I'm somewhat embarrassed to report the details. The fault was ours for
ignoring the details of the Catalyst 3550 Configuration Guide.

I post this in the hopes that others don't make this same mistake and to
remind people that part of being a Groupstudy member is the
responsibility to GIVE BACK and not just to beg for information and
fixes on our own problems.

My lab partner and I spent some time focusing on this problem yesterday
and fixed it, although we still do not fully understand the mechanism
involved.

We had been in the habit of taking a shortcut to assign a Catalyst 3550
port to a VLAN by simply typing "switchport access vlan vlan-id". In
most cases the switch and the attached routers in our lab functioned
just fine. However, when DLSW is configured and an ethernet port
connected to the 3550 is configured with "bridge-group 1" then the
switch will quit access mode and start TRUNKING on those ports, as
verified by debug and "show int".

The solution is to configure each port with the command "switchport mode
access" as recommended in the configuration guide. Without that command
the port is not locked into access mode.

I was wrong that about bridging. Configuring bridging alone on the
router port did not force the switch to start trunking on the connected
port. It is only when you configure DLSW that it starts doing this. Why
this is I still have no clue.

We used "sh interface" and various flavors of "debug spanning-tree" to
track down this problem.

It is interesting that these new-fangled switches <G> will trunk on a
port automatically. For example, if you just plug a crossover cable
between any two ports on two 3550s they will start trunking.

A bit of repetition should hammer this into the brains of those of us
unfamiliar with the 3550.

When you configure a 3550 port for a vlan do it like this:

switchport mode access
switchport access vlan 13

You leave off the first command at your peril!

Jonathan



This archive was generated by hypermail 2.1.4 : Wed Aug 06 2003 - 06:52:19 GMT-3