Re: encapsulation failed? forus FALSE? sendself FALSE? Huh?

Part of the encapsulation process is constructing the layer-2 packet
header. Which depends on ARP and the mac address tables. In the original

configuration was ARP getting through? Did you have the other switch's
mac addresses in the correct VLAN? Also, is there a hub or something in
the middle of your topo? Some of your links are at half duplex. Lastly,
VLAN? 824 has to exist on all switches that are expected to carry the
traffic even if they do not have access ports or an SVI in that VLAN.
Enable vtp or manually make sure that all the VLAN? db's are constant.

> This is getting entertaining. So, I pulled VLAN 824 from all trunks
except
> the ones directly between S2 and S4. The result is that VLAN 824 is not
> being passed anywhere. However, I can now ping from S2 to S4. I Cannot
ping
> from S4 to S2. In the ICMP debug I am not even attempting the PING
packet. I
> am missing something very fundamental. This lab has been built a dozen
times
> in the IPExpert classes including by me when I attended. I think I'm
going
> to take up raising koi....
>

> >>>>
> >>>>
> >>>> On Wed, Jul 8, 2009 at 1:09 PM, Hammer <bhmccie_at_gmail.com> wrote:
> >>>>
> >>>> > Acadamy Lab 3. Good stuff.
> >>>> >
> >>>> > Give me a few minutes and I'll post relevant output.
> >>>> >
> >>>> > -Hammer
> >>>> >
> >>>> >
> >>>> >
> >>>> > On Wed, Jul 8, 2009 at 1:06 PM, Bryan Bartik
<bbartik_at_ipexpert.com>
> >>>> wrote:
> >>>> >
> >>>> >> What lab is this? Spanning-tree topology shouldn't matter since
only
> >>>> >> redundant paths are blocked. If there was a misconfiguration
with
> >>>> MST,
> >>>> >> perhaps it could be an issue. Can you post relevant parts of the
> >>>> configs
> >>>> >> (SVIs, trunk ports, MST config), "sho interface trunk".
> >>>> >>
> >>>> >>
> >>>> >> On Wed, Jul 8, 2009 at 11:42 AM, Hammer <bhmccie_at_gmail.com>
wrote:
> >>>> >>
> >>>> >>> I'm using a proctorlabs rack. S1 is a 3550. Everything is using
> >>>> SVIs.
> >>>> >>>
> >>>> >>> 3550 - 12.2.(44)
> >>>> >>> and
> >>>> >>> 3560s - 12.2.(25r)SEC
> >>>> >>>
> >>>> >>> The tasks of the lab require one instance of spanning tree (MST
0)
> >>>> so all
> >>>> >>> my
> >>>> >>> vlans are going from S2 to S1. If I tamper with the cost of the
> >>>> >>> interface, I
> >>>> >>> can bring up the S2 to S4 leg but I kill the S2 to S1 leg. So I
can
> >>>> make
> >>>> >>> it
> >>>> >>> work but only by sacrificing another path.
> >>>> >>>
> >>>> >>> -Hammer
> >>>> >>>
> >>>> >>>
> >>>> >>> On Wed, Jul 8, 2009 at 12:34 PM, Ryan West <rwest_at_zyedge.com>
> >>>> wrote:
> >>>> >>>
> >>>> >>> > Hi,
> >>>> >>> >
> >>>> >>> > Do you have a 3550 in the mix with a physical interface that
has
> >>>> an IP
> >>>> >>> > address assigned to it or is everything using SVIs? If that
is
> >>>> the
> >>>> >>> case,
> >>>> >>> > just move the address to an SVI. Verify your VLAN allow list
and
> >>>> check
> >>>> >>> your
> >>>> >>> > trunk links to make sure the VLAN you're trying to reach is
> >>>> showing up
> >>>> >>> in
> >>>> >>> > the list.
> >>>> >>> >
> >>>> >>> > The first issue that I mentioned is detailed in this post:
> >>>> >>> >
> >>>> >>> >
> >>>> >>> >
> >>>> >>>
> >>>> http://forum.internetworkexpert.com/ubbthreads.php/ubb/
> showflat/Number/14757/page/1#Post14757
> >>>> >>> >
> >>>> >>> > -ryan
> >>>> >>> >
> >>>> >>> > -----Original Message-----
> >>>> >>> > From: nobody_at_groupstudy.com [mailto:nobody_at_groupstudy.com] On
> >>>> Behalf
> >>>> >>> Of
> >>>> >>> > Hammer
> >>>> >>> > Sent: Wednesday, July 08, 2009 1:15 PM
> >>>> >>> > To: ccielab_at_groupstudy.com
> >>>> >>> > Subject: encapsulation failed? forus FALSE? sendself FALSE?
Huh?
> >>>> >>> >
> >>>> >>> > I am trying to bring up ip over a VLAN between two switches.
It's
> >>>> a
> >>>> >>> typical
> >>>> >>> > vendor lab with 4 switches and way too many connections.
> >>>> >>> >
> >>>> >>> > 1X = 1 wire trunking all vlans
> >>>> >>> > 2X = 2 wires trunking all vlans
> >>>> >>> > EC = EtherChannel trunking all vlans
> >>>> >>> >
> >>>> >>> > S1---------1x-----------S2
> >>>> >>> > | \ / |
> >>>> >>> > | \ / |
> >>>> >>> > | \ / |
> >>>> >>> > | \ / |
> >>>> >>> > 2x EC EC 2x
> >>>> >>> > | \ / |
> >>>> >>> > | \ / |
> >>>> >>> > | \ / |
> >>>> >>> > | \ / |
> >>>> >>> > | \ / |
> >>>> >>> > | X |
> >>>> >>> > | / \ |
> >>>> >>> > | / \ |
> >>>> >>> > | / \ |
> >>>> >>> > | / \ |
> >>>> >>> > | / \ |
> >>>> >>> > | / \ |
> >>>> >>> > | / \ |
> >>>> >>> > | / \ |
> >>>> >>> > | / \ |
> >>>> >>> > S3-------1X-----------S4
> >>>> >>> >
> >>>> >>> > So I need OSPF in this little party but IP first. I can ping
> >>>> east/west
> >>>> >>> > (from
> >>>> >>> > 1-2 and from 3-4) and I can ping north/south (from 1-3) but I
> >>>> cannot
> >>>> >>> ping
> >>>> >>> > from 2-4. The vlan they need is active. The vlan interface is
> >>>> up/up on
> >>>> >>> each
> >>>> >>> > of the switches. But I can't reach the other side of the
VLAN. So
> >>>> first
> >>>> >>> I
> >>>> >>> > see that spanning tree is preferring my EC links. No problem.
Jack
> >>>> up
> >>>> >>> the
> >>>> >>> > cost to oblivion and move on. Nope. Still no love. On a debug
ip
> >>>> >>> packet, I
> >>>> >>> > come up with the below:
> >>>> >>> >
> >>>> >>> > BLAH BLAH BLAH, routed via FIB
> >>>> >>> > BLAH BLAH BLAH sending
> >>>> >>> > BLAH BLAH BLAH output feature, Check hwidb(63), rtype 1,
forus
> >>>> FALSE,
> >>>> >>> > sendself FALSE, mtu 0
> >>>> >>> > BLAH BLAH BLAH encapsulation failed
> >>>> >>> >
> >>>> >>> > I only found two links on google and none for cco for the
forus
> >>>> FALSE
> >>>> >>> > sendself FALSE. Both links were questions with no answers.
Has
> >>>> anyone
> >>>> >>> seen
> >>>> >>> > this before? Is this part of my encapsulation failure? Or am
I
> >>>> barking
> >>>> >>> up
> >>>> >>> > the wrong tree? My logic was that if I was dealing with a
spanning
> >>>> tree
> >>>> >>> > issue the traffic would still exit the switch....
> >>>> >>> >
> >>>> >>> > Help?
> >>>> >>> >
> >>>> >>> >
> >>>> >>> >
> >>>> >>> > -Hammer
> >>>> >>> >
> >>>> >>> >
> >>>> >>> > Blogs and organic groups at http://www.ccie.net
> >>>> >>> >
> >>>> >>> >
> >>>>
Received on Wed Jul 08 2009 - 21:52:24 ART