Re: [OSPF] OSPFv2 SR draft

Hi Chris,

On 24/08/16 20:31 , Chris Bowers wrote:
> Peter,
>
> The text that you propose corresponds to part of the text that I proposed, and it seems good to me.
>
> However, the last sentence of the text that I proposed in not addressed.
> ------
> If router B does not advertise the
> SR-Algorithm TLV for algorithm X, then other routers should not
> forward traffic destined for a prefix-SID for algorithm X advertised by
> some router D using a path that would require router B to forward traffic using
> algorithm X.
> ------
> Is this an oversight?

not that I disagree with the statement that you want to add.

The question is whether that belongs to the IGP SR draft, or whether 
that should be specified in a different draft.

There is already some text regarding the forwarding for a SR algorithm 
in draft-ietf-spring-segment-routing section 3.2.1., which may not be 
aligned with what you have in mind:

   "The ingress node of an SR domain validates that the path to a prefix,
    advertised with a given algorithm, includes nodes all supporting the
    advertised algorithm.  In other words, when computing paths for a
    given algorithm, the transit nodes MUST compute the algorithm X on
    the IGP topology, regardless of the support of the algorithm X by the
    nodes in that topology.  As a consequence, if a node on the path does
    not support algorithm X, the IGP-Prefix segment will be interrupted
    and will drop packet on that node.  It's the responsibility of the
    ingress node using a segment to check that all downstream nodes
    support the algorithm of the segment."

Maybe we should add/modify the text in draft-ietf-spring-segment-routing 
section 3.2.1, rather then adding anything to the OSPF/ISIS SR drafts.

thanks,
Peter

>
> Thanks,
> Chris
>
>
> -----Original Message-----
> From: Peter Psenak [mailto:ppsenak@cisco.com]
> Sent: Monday, August 22, 2016 2:32 AM
> To: Chris Bowers <cbowers@juniper.net>; OSPF List <ospf@ietf.org>
> Subject: Re: [OSPF] OSPFv2 SR draft
>
> Chris,
>
> what about this to be added in the Section 3.1:
>
>
> "A router receiving a Prefix-SID (defined in section 5) from a remote node and with an SR algorithm value that such remote node has not advertised in the SR-Algorithm sub-TLV MUST ignore the Prefix-SID sub-TLV."
>
> thanks,
> Peter
>
>
> On 19/08/16 23:33 , Chris Bowers wrote:
>> Peter,
>>
>> Please share the updated text that you plan to use with the WG, since this is a reasonably significant clarification.
>>
>> Thanks,
>> Chris
>>
>> -----Original Message-----
>> From: Peter Psenak [mailto:ppsenak@cisco.com]
>> Sent: Tuesday, August 16, 2016 10:02 AM
>> To: Chris Bowers <cbowers@juniper.net>; OSPF List <ospf@ietf.org>
>> Subject: Re: [OSPF] OSPFv2 SR draft
>>
>> Hi Chris,
>>
>> I'll update the draft along those lines.
>>
>> thanks,
>> Peter
>>
>>
>> On 16/08/16 16:02 , Chris Bowers wrote:
>>> Peter,
>>>
>>> I suggest changing the paragraph to read as below to make this clearer.
>>>
>>> =====
>>>       The SR-Algorithm Sub-TLV is optional.  It MAY only be advertised once
>>>       in the Router Information Opaque LSA.  If the SID/Label Range TLV, as
>>>       defined in Section 3.2, is advertised, then the SR-Algorithm TLV MUST
>>>       also be advertised.  If a router C advertises a Prefix-SID sub-TLV for algorithm X
>>>       but does not advertise the SR-Algorithm Sub-TLV with algorithm X, then
>>>       a router receiving that advertisement MUST ignore the Prefix-SID
>>>       advertisement from router C.  If router B does not advertise the
>>>       SR-Algorithm TLV for algorithm X, then other routers should not
>>>       forward traffic destined for a prefix-SID for algorithm X advertised by
>>>       some router D using a path that would require router B to forward traffic using
>>>       algorithm X.
>>> =====
>>>
>>> Thanks,
>>> Chris
>>>
>>>
>>>
>>> -----Original Message-----
>>> From: Peter Psenak [mailto:ppsenak@cisco.com]
>>> Sent: Monday, August 15, 2016 6:40 AM
>>> To: Chris Bowers <cbowers@juniper.net>; OSPF List <ospf@ietf.org>
>>> Subject: Re: [OSPF] OSPFv2 SR draft
>>>
>>> Hi Chris,
>>>
>>> sorry for the delay, I was on PTO during last two weeks.
>>> Please see inline:
>>>
>>> On 03/08/16 16:45 , Chris Bowers wrote:
>>>> Peter,
>>>>
>>>> Taking a looking at the whole paragraph into this sentence was
>>>> added, I am not sure how to interpret it.
>>>>
>>>>        The SR-Algorithm Sub-TLV is optional.  It MAY only be advertised once
>>>>        in the Router Information Opaque LSA.  If the SID/Label Range TLV, as
>>>>        defined in Section 3.2, is advertised, then the SR-Algorithm TLV MUST
>>>>        also be advertised.  If the SR-Algorithm TLV is not advertised by the
>>>>        node, such node is considered as not being segment routing capable.
>>>>
>>>> Is this sentence intended to imply that if a router does not
>>>> advertise the SR-Algorithm TLV including algorithm X, then any
>>>> prefix-SIDs for algorithm X advertised by that router will be ignored by other routers?
>>>
>>> in OSPF we do not have the SR capability TLV. We use SR-Algorithm TLV
>>> for that purpose. So if a router does not advertise the SR-Algorithm
>>> TLV for algorithm X, other routers should not send any SR traffic
>>> using SIDs that were advertised for algorithm X.
>>>
>>> If the router does not advertise any SR Algorithm TLV, then the node
>>> is not SR capable and no SR traffic should be forwarded to such a node.
>>>
>>> thanks,
>>> Peter
>>>
>>>
>>>>
>>>> If this is the intention, then it would be better to state is more explicitly.
>>>>
>>>> If not, then the intended meaning should be clarified.
>>>>
>>>> Thanks,
>>>> Chris
>>>>
>>>>
>>>> -----Original Message-----
>>>> From: OSPF [mailto:ospf-bounces@ietf.org] On Behalf Of Peter Psenak
>>>> Sent: Friday, July 22, 2016 3:30 AM
>>>> To: OSPF List <ospf@ietf.org>
>>>> Subject: [OSPF] OSPFv2 SR draft
>>>>
>>>> Hi All,
>>>>
>>>> following text has been added in the latest revision of the OSPFv2
>>>> SR draft, section 3.1.
>>>>
>>>> "If the SR-Algorithm TLV is not advertised by node, such node is
>>>> considered as not being segment routing capable."
>>>>
>>>> Please let us know if there are any concerns regarding this addition.
>>>>
>>>> thanks,
>>>> Peter
>>>>
>>>> _______________________________________________
>>>> OSPF mailing list
>>>> OSPF@ietf.org
>>>> https://www.ietf.org/mailman/listinfo/ospf
>>>> .
>>>>
>>>
>>> .
>>>
>>
>> .
>>
>
> .
>