Hi Jeff,

Inline comments

From: Jeffrey Haas [mailto:[email protected]]
Sent: Monday, July 20, 2015 13:47
To: LITKOWSKI Stephane SCE/IBNF
Cc: [email protected]
Subject: Re: Comments on draft-shaikh-rtgwg-policy-model


On Jul 20, 2015, at 1:35 PM, 
<[email protected]<mailto:[email protected]>> 
<[email protected]<mailto:[email protected]>> wrote:

-          Regarding tags, as pointed today, I would like to have “tag” to be a 
generic local identifier rather than pointing only to IS-IS and OSPF. Any route 
within a RIB may have a local tag (this local tag can be learned from the 
routing protocol or set by configuration or policy). So setting a tag does not 
refer to any igp-action.

I believe we do wish to keep this field used for IGP route tagging since that's 
a protocol component.
[SLI] Do we really need to differentiate from a policy point of view ? from an 
import policy perspective, matching a tag, means learning the tag value 
available in the protocol (if available) and when the route ins inserted into 
RIB the tag value is copied from the protocol value if not overrided by import 
policy action; from an export policy perspective (talking about export from rib 
to protocol), matching a tag means matching the tag value in the RIB (which may 
come from protocol or not),  setting a tag means fill the protocol field if 
available. From a RIB point of view, the tag associated with the route is 
protocol agnostic, even if the protocol does not support tags in encoding you 
may associate a local tag for policy processing.

Having two types of tags is also possible : protocol-tag and local-tag but I 
see more complexity and do not see more flexibility : but maybe there is some 
use case that I do not see.


(Fascinating issue, while researching ISIS implementations I found that the 
protocol permits 64bit tagging, but didn't find anyone that implemented them.)
[SLI] ☺ we have it in IS-IS yang model , tag64 if I remember correctly.

I do realize that there is also a need to do general route mark-up so that 
policy engines can operate off of that.  I know that for certain route types, 
the "tag" field might be able to be used for this purpose; e.g it has no 
semantics in a given BGP implementation unless the route is redistributed in 
the IGP.  (And I'm not even sure that field survives export, I'd have to check 
the code.)

Perhaps more appropriate would be to have a new mark-up field, name TBD, to 
cover this purpose.  It could then be implemented as appropriate for a given 
route type and vendor implementation?

-- Jeff


_________________________________________________________________________________________________________________________

Ce message et ses pieces jointes peuvent contenir des informations 
confidentielles ou privilegiees et ne doivent donc
pas etre diffuses, exploites ou copies sans autorisation. Si vous avez recu ce 
message par erreur, veuillez le signaler
a l'expediteur et le detruire ainsi que les pieces jointes. Les messages 
electroniques etant susceptibles d'alteration,
Orange decline toute responsabilite si ce message a ete altere, deforme ou 
falsifie. Merci.

This message and its attachments may contain confidential or privileged 
information that may be protected by law;
they should not be distributed, used or copied without authorisation.
If you have received this email in error, please notify the sender and delete 
this message and its attachments.
As emails may be altered, Orange is not liable for messages that have been 
modified, changed or falsified.
Thank you.

_______________________________________________
rtgwg mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/rtgwg

Reply via email to