From: Erick B. (erickbe@xxxxxxxxx)
Date: Thu Nov 08 2001 - 00:49:19 GMT-3
Guys,
I'm aware that dot1Q VLAN1 is untagged, but in my lab
and question I was pinging between VLAN2 and VLAN3
which are tagged VLANs. It didn't work until I added
another sub-interface configured for VLAN1 on the
router. After it worked, I removed this sub-interface
and it continued to work. I then rebooted router and
switch and it still worked. I have seen similar/same
odd behavior in some customers networks I've worked on
from time to time. I'm not sure how easy this problem
would be to produce on the fly as I haven't been able
to get it to act that way today.
On the switch side, it was to set to 802.1q and none
of the auto-stuff was on. Spanning tree was off to. I
just made changes on router side and it started
working even though router config was good already.
As for code, router IOS I was using was 12.0(7)T but I
also saw similar behavior with a recent 12.1(x) image.
I'm picky when I see strange behavior and try to get
to the bottom of it.
Thanks, Erick
--- Bryan Osoro <bosoro@hotmail.com> wrote:
> Reason for this happening is normal
>
> Dot1q in Cisco's implementation does NOT tag VLAN
> traffic on the "Native
> VLAN" Native VLAN on ports is 1, by default.
> Therefore VLAN 2 and 3
> traffic is getting to the router with Dot1q
> information in every frame,
> and VLAN 1 is NOT getting tagged. The router
> considers VLAN 1 to not be
> "encapsulated," and I bet if you had a debug ip
> packet detail on you'd
> see encapsulation failed when the PC sends the ping
> to the router. Up
> until 12.1 IOS could not do anything to fix this,
> now there is a native
> vlan command that makes the router aware of the
> switches behavior.
> Other fix is to change the native vlan on the switch
> to be some number
> that is not in use (i.e. 888) or some random number.
>
> -Bryan
>
> -----Original Message-----
>
> Guys
>
> This should do the trick, note the table regarding
> dot1q support and IOS
> versions.
>
> http://www.cisco.com/warp/customer/473/50.shtml
>
> Thanks
>
> Steve
>
> -----Original Message-----
>
> Erick,
>
> Do a 'sh trunk':
>
> WP-4006-1> sh trunk
> * - indicates vtp domain mismatch
> Port Mode Encapsulation Status
> Native vlan
> -------- ----------- ------------- ------------
> -----------
> 1/1 on dot1q trunking
> 1
> 1/2 auto dot1q trunking
> 1
> 2/1 auto dot1q trunking
> 1
> 2/2 auto dot1q trunking
> 1
>
> If your trunk port doesn't show 'trunking' as a
> status, it won't work.
> Try
> setting the mode on both ends to on, rather than
> auto or desirable. Are
> you
> running fairly recent GD code on the router and
> switch?
>
> Chuck
>
> -----Original Message-----
> From: nobody@groupstudy.com
>
> Hello there,
>
> I've been playing around with some VLAN stuff lately
> and have seen some odd behavior with trunking and I
> can't find a answer so thought I'd see if anyone
> here
> knew, because it is bugging me.
>
> Real basic setup.
>
> Cisco router with FE connecting to a switch. VLANs
> 1,
> 2, and 3. Workstation on switch in VLAN 2 and VLAN
> 3.
>
> Dot1Q trunk with all VLANs defined. Sub-interfaces
> on
> router for dot1q. Ping from PC in VLAN2 to VLAN3 and
> don't get further then router interface for that
> VLAN.
> Change to ISL encaps and everything works fine.
> Change
> back to Dot1Q and things continue to work. Save
> configs and reload switch and routers and things
> continue to work.
>
> I also had a similar thing when I didn't configure a
> VLAN1 sub-interface on the router (just 2 and 3).
> Once
> I added VLAN1 thing started working. Took VLAN1 away
> for kicks and things still worked. Weird.
>
> Any one have any thoughts???
>
> Thanks, Erick
This archive was generated by hypermail 2.1.4 : Fri Jun 21 2002 - 06:45:08 GMT-3