Re: BGP TTL

my point was simple ....

if a device is forwarding any packets ....like a transit router....

afer making all checkums , cef lookup;s, encriptions , adjacency table
lookps , NAT , statefull , qos decisions if the device later on drops the
packet owing to LOW ttl value. wastes the cpu cycles used in previous
operations......

you may be correct ...but only an expert can explain correct order of
operations ...and palcing the ttl value checking operation first or second
makes good use of resources.

this may also depondant on IOS code or palform or vendor ...

so let the experts guide us to correct order of operations .

On Sat, Oct 22, 2011 at 8:05 AM, CCIE KID <eliteccie_at_gmail.com> wrote:

> Hi Imran,
>
> My question is all IGP control plane packets are send with TTL of 1. Lets
> go with ur logic
>
>
> *pcket comes in with a ttl value of 1*
>
> *=> routers looks at the cef table to find an entry and determines the
> exit interface *
>
> *=> The above step may involve many cef table lookup's (recursive
> lookup)
>
> *
> *=> Now the packet is destined to itslef *( its own locaaly connected IP
> Address ), so there is nothing related to next hop or outgoing interface or
> outgoing encapsulation.
>
> *=> So there will be no decrement in the TTL vlaue. so packets with a TTL
> of 1 cannot be dropped if it sees the destined address is itself ( local
> router ) which receives the packet . ( Example is any IGP update packet ) .
> It doesnt go beyond that particular router.
>
> **=> So there will be two checksums will be created on routers. When a
> packet hits the router , it will generate a checksum to check that the
> packet has not been scrambled and does all the lookup and sends it to the
> outgoing interface and there will be a decrement in TTL and again the new
> checksum will be created ( because the TTL has been changed) and attached to
> the IP Header and send it out of the outgoing interface
>
> **=> *So thats my point. Correct me if i am wrong in the above statement
> and justify ur statement
>
> THank You
>
> *
> **
>
> *
>
>
>
> On Sat, Oct 22, 2011 at 1:15 AM, Imran Ali <immrccie_at_gmail.com> wrote:
>
>>
>>
>> On Fri, Oct 21, 2011 at 3:06 PM, Routing Freak <routingfreak_at_gmail.com>wrote:
>>
>>> Then buddy all OSPF update packets are sent with a TTL of 1 to their
>>> directly connected neighbors , so when the neighbors receive the packet ,
>>> decrement the TTL to 0 and drop the packet ahh. Naa this will not happen
>>>
>>> TTL decrement will happen only on outbound .
>>>
>>>
>>> *=> pcket comes in with a ttl value of 1*
>> **
>> *=> routers looks at the cef table to find an entry and determines the
>> exit interface *
>> **
>> *=> The above step may involve many cef table lookup's (recursive
>> lookup)*
>> **
>> *=>now looks at the adjacency table to find the layer 2 add of next hop *
>> **
>> *=> now he builds the layer 2 header *
>> **
>> *=> poor router now decriments the TTL value , and drops the packet . *
>> **
>> *it make more and more sense , first to find out if the packet have
>> enough TTL value ,THEN he can proceed with other forwarding operations ....
>> *
>>
>>
>>
>>
>>
>>> On Fri, Oct 21, 2011 at 5:20 PM, Imran Ali <immrccie_at_gmail.com> wrote:
>>>
>>>> Hi kid,
>>>>
>>>> regardsless of data plane or control pale TTL value is reduced .
>>>>
>>>> The moment any interface receives a packet the TTL will be reduced by 1
>>>> .
>>>>
>>>> On Fri, Oct 21, 2011 at 2:25 PM, CCIE KID <eliteccie_at_gmail.com> wrote:
>>>>
>>>> > Hi fellas,
>>>> >
>>>> > I am working on BGP and i find in sh ip bgp neighbor command, the
>>>> incoming
>>>> > TTL and the outgoing TTL . I understood the outgoing TTL is the one
>>>> when
>>>> > the
>>>> > control plane packets are generated , the TTL in the IP Header will be
>>>> 255.
>>>> > I find the incoming TTL to 0 . Can anyone explain me what is actually
>>>> meant
>>>> > by incoming TTL.
>>>>
>>>>
>>>>
>>>> > My understanding is the incoming TTL is the when the
>>>> > neighbor send u any BGP Control plane packet, the local router is
>>>> expecting
>>>> > this TTL value in the IP HEader. Am i right ?
>>>> > So the incoming TTL valus is 0 in my case..
>>>> >
>>>>
>>>> regardless of any operation , the TTL value is always present in ip
>>>> header
>>>> :-
>>>>
>>>> >
>>>> > Router drops a packet with a TTL value of 0. So when the control plane
>>>> > packet comes. it should be of TTL of 1.
>>>> > TTL will be only decremented on the outgoing interface and not the
>>>> incoming
>>>> > interface. THe genreal logic says that the TTL of all control plane
>>>> packets
>>>> > of all IGPS will be 1 and so when a router receives this control plane
>>>> > packet , it is destined to itself , it will update the OSPF or EIGRP
>>>> RIB
>>>> > and
>>>> > in turn do the local computation for the best path .
>>>> >
>>>> > My question is whether the where does the TTL decrement happen ? My
>>>> answer
>>>> > is outbound and for any other control plane packet it will be destined
>>>> to
>>>> > itslef , so there will be no decrement in the TTL
>>>> >
>>>> > --
>>>> > With Warmest Regards,
>>>> >
>>>> > CCIE KID
>>>> > CCIE#29992 (Security)
>>>> >
>>>> >
>>>> > Blogs and organic groups at http://www.ccie.net
>>>> >
>>>> >
>>>> _______________________________________________________________________
>>>> > Subscription information may be found at:
>>>> > http://www.groupstudy.com/list/CCIELab.html
>>>>
>>>>
>>>> Blogs and organic groups at http://www.ccie.net
>>>>
>>>> _______________________________________________________________________
>>>> Subscription information may be found at:
>>>> http://www.groupstudy.com/list/CCIELab.html
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>
>>
>
>
> --
> With Warmest Regards,
>
> CCIE KID
> CCIE#29992 (Security)

Blogs and organic groups at http://www.ccie.net
Received on Sat Oct 22 2011 - 14:55:28 ART