"Aijun Wang" <[email protected]> writes:
Hi, Christian:
In our concerned scenarios, it is not one or small part of the prefixes that is
important.
Instead, the loopback addresses of all PEs(for overlay BGP/BGP VPN services),
or all Ps(for SRv6 tunnel service) are all important.
Extract them out of the summary addresses and advertise them into IGP directly
is not the optimize solution-----This will certainly increase the IGP's pressure
in normal situation, and weaken the WG' efforts to introduce several solutions
for the flood reduction effects.
I don't see how this addresses my point.
Having extra prefixes advertised is not in and of itself a bad thing, even if there are a
good number of them. You gain a lot by advertising these prefixes that you've identified
as "Very Important", so unless the protocol literally can't handle it, it just
seems like Good Design (KISS) to me.
Thanks,
Chris.
[as wg member]
Best Regards
Aijun Wang
China Telecom
-----Original Message-----
From: Christian Hopps <[email protected]>
Sent: Tuesday, January 4, 2022 2:31 PM
To: Aijun Wang <[email protected]>
Cc: Christian Hopps <[email protected]>; Tony Li <[email protected]>; Robert
Raszuk <[email protected]>; Shraddha Hegde <[email protected]>; Hannes
Gredler <[email protected]>; Les Ginsberg (ginsberg)
<[email protected]>; lsr <[email protected]>; Peter Psenak
<[email protected]>
Subject: Re: [Lsr] BGP vs PUA/PULSE
On Jan 3, 2022, at 11:26 PM, Aijun Wang <[email protected]> wrote:
Hi, Christian and Tony:
Following your idea, the network will be filled with the large amounts of loop
back addresses, will be one inextensible solution which has been discussed
previously.
I stand by what I said: "... if a prefix is important enough to merit a bunch of
new protocol extensions and state, then it's important enough to simply be
left out of the summarization in the first place."
And, once such loopack address failures, the nodes in other areas will also be
notified.
No different than what is being proposed.
That is to say, such solution is not optimized both in normal situations and
failures situations.
It is optimized. It is optimized for optimal delivery to that prefix. No
unseemly protocol changes required.
So yet again, either the prefix is important (it is reasonable for it to take
up signaling/state) or it isnt (don't special case it). Nothing I've read in
the hundreds of emails has convinced me otherwise.
Thanks,
Chris.
[as wg member]
Aijun Wang
China Telecom
On Jan 4, 2022, at 05:09, Tony Li <[email protected]> wrote:
On Jan 3, 2022, at 11:23 AM, Christian Hopps <[email protected]> wrote:
And I'm saying if a prefix is important enough to merit a bunch of new
protocol extensions and state, then it's important enough to simply be left
out of the summarization in the first place.
And then people get what they want, w/o protocol changes/upgrades, and it's
using time tested and hardened IGP code and designs.
+1
T
_______________________________________________
Lsr mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/lsr
