Re: Ping across auto IPV6 Tunnel - 50%

Ok, problem found, I only noticed it when I was going to copy my configs
into this mail and send them to you all. Here are my configs, I will not
say yet what the problem is, but it is found in my configs (usually the case
...).

Andrew

PS - any other comments are most welcome!!!

~~~~~~~~

R1:

ipv6 unicast-routing
!
interface Tunnel0
 no ip address
 no ip redirects
 ipv6 address 2002:101:101::1/64
 tunnel source Loopback0
 tunnel mode ipv6ip 6to4
!
interface Loopback0
 ip address 1.1.1.1 255.255.255.255
 ipv6 address 2001:1:1:1::1/64
 ipv6 rip 1 enable
!
ipv6 route 2002::/16 Tunnel0
ipv6 route 2001:3:3:3::3/128 2002:303:303::3
!

R3 is across the frame cloud. Here are the configs from R3:

ipv6 unicast-routing
!
interface Loopback0
 ip address 3.3.3.3 255.255.255.0
 ipv6 address 2001:3:3:3::3/64
 ipv6 rip 1 enable
!
interface Tunnel0
 no ip address
 no ip redirects
 ipv6 address 2002:303:303::3/64
 tunnel source Loopback0
 tunnel mode ipv6ip 6to4
!
ipv6 route 2001:1:1:1::/64 2002:101:101::1
ipv6 route 2001:1:1:1::/64 Tunnel0
ipv6 route 2002::/16 Tunnel0

On Tue, Sep 29, 2009 at 12:27 AM, ALL From_NJ <all.from.nj_at_gmail.com> wrote:

> For sure ... that was one of my first thought. Since the ipv6 addresses
> are really v4 addresses, I consulted the ipv4 routing table in both
> directions. There is only a single path to each loopback and physical
> address
>
> Roy, I appreciate your response. Any thoughts on additional debugs or show
> commands?
>
>
>
> On Tue, Sep 29, 2009 at 12:18 AM, Roy Waterman <roy.waterman_at_gmail.com>wrote:
>
>> Hi Andrew
>>
>> So you've verified that there is no equal cost load balancing going on?
>>
>> Regards
>> Roy
>>
>> 2009/9/29 ALL From_NJ <all.from.nj_at_gmail.com>
>>
>>> Hey team,
>>>
>>> Got this working tonight, auto ipv6 tunnels.
>>>
>>> Divin provided me this link:
>>>
>>> http://ardenpackeer.com/routing-protocols/tutorial-ipv6-tunnels-part-2-automatic-6to4-tunnels/
>>> <-- this helped me much.
>>>
>>> Conceptually it helped me to consider this a 'hack' or 'cheat'. Kind of
>>> neat in some ways too I guess ... by defining the IPv4 address as part of
>>> the IPv6 address, you are really telling the router to find the remote
>>> peer
>>> by it's IPv4 address. Of course ... the IPv4 routing table needs to be
>>> right before this will work ...
>>>
>>> Any who ... my question is related to ping. Has anyone else seen 50%
>>> ping
>>> timeout when using auto tunnels? I get all responses to other frame
>>> relay
>>> ipv6 peers or Ethernet peers etc ... . Here is the output, kind of looks
>>> like a pacing issue, but the fact that other pings work 100% of the time
>>> does not make sense.
>>>
>>> R1(config)#do ping 2001:3:3:3::3 sou 2001:1:1:1::1 re 10
>>>
>>> Sending 10, 100-byte ICMP Echos to 2001:3:3:3::3, timeout is 2 seconds:
>>> Packet sent with a source address of 2001:1:1:1::1
>>> !.!.!.!.!.
>>> Success rate is 50 percent (5/10), round-trip min/avg/max = 32/32/36 ms
>>>
>>> I have static routes going to the remote tunnel peer.
>>>
>>> Trace works just fine, no timeouts, single hop. I tried it with and
>>> without
>>> ipv6 cef. I also did a debug ipv6 pack det and debug ipv6 icmp ...
>>> nothing
>>> out of the ordinary ...
>>>
>>> Any thoughts? Seen this?
>>>
>>> --
>>> Andrew Lee Lissitz
>>> all.from.nj_at_gmail.com
>>>
>>>
>>> Blogs and organic groups at http://www.ccie.net
>>>
>>> _______________________________________________________________________
>>> Subscription information may be found at:
>>> http://www.groupstudy.com/list/CCIELab.html
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>
>>
>> --
>> Regards
>> Roy
>>
>
>
>
> --
> Andrew Lee Lissitz
> all.from.nj_at_gmail.com
>

-- 
Andrew Lee Lissitz
all.from.nj_at_gmail.com
Blogs and organic groups at http://www.ccie.net