On Tue, 6 Apr 2010, Sriram, Kotikalapudi wrote:

Responses inline.

-----Original Message-----
From: Sandra Murphy [mailto:[email protected]]
Sent: Tuesday, April 06, 2010 1:49 PM

Comments in line.

On Tue, 6 Apr 2010, Sriram, Kotikalapudi wrote:

....
<snip>

About point 2:  There is this thing called "proxy aggregation" where an
ISP discovers two external peer BGP routes that are aggregatable and
announces the aggregate.  Long long ago, IIRC, we decided that proxy
aggregation was too hard and too rare and we would declare it out of
scope.  Your point 2 would specifically disallow proxy aggregation.  Not
saying that's a bad thing, for a feature that was declared out of scope.
Just warning about the side effect.

My point 2 was not to say that proxy aggregation would be disallowed.
In that example, if the prefix owner of 10.1.0.0/16 wants the proxy aggregation 
to pass the ROA validation, then they should be aware enough to create four ROAs
for the _aggregate_ prefix (/16) as follows:
ROA-5: {10.1.0.0/16, maxlength =18, AS64496},
ROA-6: {10.1.0.0/16, maxlength =18, AS64497},
ROA-7: {10.1.0.0/16, maxlength =18, AS64498},
ROA-8: {10.1.0.0/16, maxlength =18, AS64499}.
That is, they should not merely create ROAs for the more specifics
as stated in my point 2 above, and hope that the router
algorithm will magically combine the ROAs for the more specifics
and hence figure it out that the proxy aggregation is OK.
No ... they should know that will not work.
The way to make it work is to create ROAs for the aggregate
as illustrated in ROA-5 through ROA-8 above.
With this set of ROAs for the aggregate, the four
ASs listed above can still originate the more specifics (/18s),
and when an upstream ISP's AS does proxy aggregation, then
that proxy aggregation will pass ROA validation.
That is what I think we want to happen.
It is just a matter of cautioning the user and their ISP about what specific
ROAs they should create if they wish to do proxy aggregation.

As I understand proxy aggregation, an ISP receives updates from two peers and discovers that the NLRI in the two updates coincidentally aggregate and the ISP wants to announce the aggregate. (If I'm wrong, I should be corrected - I'm thinking of description in rfc2519 and Geoff's example in http://www.potaroo.net/ispcol/2006-05/bgp.html.)

So neither origin AS has authority over the aggregate prefix. So neither origin AS would be able to issue a ROA for the proxy aggregated prefix.

--Sandy

P.S. I found the long ago discussion on the pre-wg sidr list starting with:

http://www.ietf.org/mail-archive/web/sidr/current/msg00025.html


Sriram

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

Reply via email to