Tony,
On 14/10/2021 07:50, Tony Li wrote:
Les,
If you’re advertising loopbacks that are outside of the summarized
space, then you end up with reachability and liveness. Yes, there’s a
cost in scalability… it ain’t free.
the whole point is to summarize loopbacks of PE devices, so above would
not help.
thanks,
Peter
Tony
On Oct 13, 2021, at 10:36 PM, Les Ginsberg (ginsberg)
<[email protected] <mailto:[email protected]>> wrote:
Tony –
Given that IGPs provide the ability to advertise summaries (in the
interest of scalability), I think it is reasonable to say that being
able to advertise reachability changes of endpoints covered by the
summary also can be considered as a legitimate use of the IGP.
IGPs certainly advertise reachability and do update those
advertisements in response to liveness changes.
So, I don’t agree w your conclusion that this isn’t logically a
routing protocol function.
It is fair for you to think the solution is good or bad – happy to
hear more thoughts from you on that. But I don’t think arguing that
the routing protocol has no business being involved in this is on the
mark.
Les
*From:*Lsr <[email protected] <mailto:[email protected]>>*On
Behalf Of*Tony Li
*Sent:*Wednesday, October 13, 2021 8:53 PM
*To:*Gyan Mishra <[email protected] <mailto:[email protected]>>
*Cc:*lsr <[email protected] <mailto:[email protected]>>
*Subject:*Re: [Lsr] "Prefix Unreachable Announcement" and "IS-IS and
OSPF Extension for Event Notification"
Hi Gyan,
I fully understand the request.
I still question whether it should be solved by routing. This is not
a path computation problem. It’s reachability and more significantly,
liveness.
That really seems like it’s looking for a slightly different
architectural tool.
Yours in IS-IS,
Tony
On Oct 13, 2021, at 6:04 PM, Gyan Mishra <[email protected]
<mailto:[email protected]>> wrote:
Hi Tony
This is a real world problem for large scale networks where an
service providers core network is broken up into OSPF areas or
ISIS levels and each PEs loopback BGP next hop attribute for
1000s of PEs within an area are summarized at the P router ASBR
ISIS L1-2 router or OSPF ABR and the next hop attribute loopback0
component prefix of the summary goes down.
So we need a mechanism to signal via PUA or event notification
mechanism that the component went down that is part of the summary
to force immediately control plane convergence to avoid black hole
of traffic during the failure.
Kind Regards
Gyan
On Wed, Oct 13, 2021 at 8:48 PM Tony Li <[email protected]
<mailto:[email protected]>> wrote:
Hi,
I’ve said it many times before, in many different venues: “BGP
is not a dump truck!”.
Is the fact that I didn’t mention the IGPs taken as some
indication that they are fair game?
No, the IGPs aren’t a dump truck either.
We have a clear, unambiguous way of signaling individual
system outages already. You advertise a loopback address as a
/32 or /128.
If your host goes down, the address is no longer routable. Done.
If this technique is not sufficient, then perhaps it doesn’t
belong in routing.
Tony
_______________________________________________
Lsr mailing list
[email protected] <mailto:[email protected]>
https://www.ietf.org/mailman/listinfo/lsr
--
<~WRD0000.jpg> <http://www.verizon.com/>
*Gyan Mishra*
/Network Solutions Architect /
/[email protected] <mailto:[email protected]>/
/M 301 502-1347/
_______________________________________________
Lsr mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/lsr
