Hi, Alia,

On Jul 25, 2013, at 2:24 PM, Alia Atlas 
<[email protected]<mailto:[email protected]>> wrote:

Carlos,

I'm not surprised that there is lots of vendor documentation.  It's something 
that has been around for a while.   The question is whether there is anything 
that might benefit from documentation.

Yes, I agree. If the goal of the effort is to document existing 
inconsistencies, my initial reaction is that those are well documented already 
outside the RFC series, and generally well known. On the other hand, I do not 
think that an analysis of the issue will cause harm, of course, and can result 
in a reference of use. However, the excerpt I included below states a different 
goal, which I do not think is realistic, practical, or backwards-compatible :-)

If the document would aim at cataloging different ways of configuring MTU and 
warning users about what commands mean, I would also find more value in an 
analysis of MTU configuration requirements (1812/1122/ospf/isis) in addition to 
the mostly empiric approach.


   The size of label stack, in particular,  might be interesting...


I agree that the size of the label stack is interesting -- but I sense that 
this is a bit of a rathole and orthogonal to MTU configuration for an OSPF 
adjacency to come up. It does cause fragmentation potentially but not 
interoperability problems. I wonder if the size of the label stack is/should be 
included in draft-ietf-mpls-forwarding.

Point is to have the discussion.

Exactly. This is my small contribution to it. :-)

Thanks!

-- Carlos.

Alia

On Jul 25, 2013 1:54 PM, "Carlos Pignataro (cpignata)" 
<[email protected]<mailto:[email protected]>> wrote:
Alia,

The fact that different vendors use different interpretations of "MTU" in a 
configuration interface (some L3, some L2 (with/without FCS)) is well known and 
accounted for while configuring. Many of those config interfaces show "hardware 
mtu", "protocol mtu", and other proxies for naming labels to imply which encaps 
included. There seems to be lots of publicly available documentation about this.

I was also wondering what the expected output or goal is. Would publishing an 
RFC achieve the goal stated in the I-D?

   for all vender to standardize the MTU configuration in IP
   network and MPLS network.

Or are existing RFCs + vendors' documentation + discussion lists archives 
enough?

Thanks,

-- Carlos.

On Jul 25, 2013, at 12:19 PM, Alia Atlas 
<[email protected]<mailto:[email protected]>> wrote:

Yes - my thought is perhaps a BCP??  It's clearly an interoperability problem.  
What is interesting is to discuss the issue and see what could be done.

Alia

P.S.  Thanks for the quick reading and response!


On Thu, Jul 25, 2013 at 12:04 PM, Tony Li 
<[email protected]<mailto:[email protected]>> wrote:

On Jul 25, 2013, at 8:35 AM, Alia Atlas 
<[email protected]<mailto:[email protected]>> wrote:

> We are likely to add this draft to the list to be discussed in the meeting.  
> It is discussing some general problems with MTU configuration between 
> vendors.  While it discusses this for ISIS, some aspects also apply to OSPF.
>
> Basically, there are three problems:
>
> a) Some implementations use L2 MTU and others use the L3 MTU
> b) Different implementations pad and check for different lengths.  
> Consequences are that ISIS sessions do not make it to Up.
> c) Different numbers of MPLS labels are assumed as the max, resulting in 
> fragmentation, and there are generally no controls for it.
>
> http://datatracker.ietf.org/doc/draft-liu-rtgwg-mtu-config-ps/


I'm interested in understanding what the authors would like to see from this.   
It would be well outside of the IETF's charter to say that implementations must 
configure MTU in only one manner.

At best, I see this document as saying 'be consistent', which seems like 
spitting upwind.

Tony


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

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

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

Reply via email to