Re: [spring] to drop or to forward unlabelled (Re: Question on RFC8660)

Martin Horneffer <maho@lab.dtag.de> Fri, 28 August 2020 10:13 UTC

Return-Path: <maho@lab.dtag.de>
X-Original-To: spring@ietfa.amsl.com
Delivered-To: spring@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 033693A08E5 for <spring@ietfa.amsl.com>; Fri, 28 Aug 2020 03:13:31 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.644
X-Spam-Level:
X-Spam-Status: No, score=-2.644 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_INVALID=0.1, DKIM_SIGNED=0.1, HTML_MESSAGE=0.001, NICE_REPLY_A=-0.948, SPF_HELO_NONE=0.001, SPF_NONE=0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=neutral reason="invalid (public key: not available)" header.d=lab.dtag.de
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 0JQvJExI1ROu for <spring@ietfa.amsl.com>; Fri, 28 Aug 2020 03:13:28 -0700 (PDT)
Received: from OldBailey.lab.dtag.de (OldBailey.lab.DTAG.DE [194.25.1.220]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 0DE653A0853 for <spring@ietf.org>; Fri, 28 Aug 2020 03:13:27 -0700 (PDT)
Received: from [127.0.0.1] (localhost [127.0.0.1]) by localhost (Mailerdaemon) with ESMTPSA id C7D1CC1120; Fri, 28 Aug 2020 12:13:23 +0200 (CEST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=lab.dtag.de; s=dkim; t=1598609603; h=from:subject:date:message-id:to:cc:mime-version:content-type: in-reply-to:references; bh=xczt3cE88B9dMcPR3o8A83K4xmKH0Aaumqd3FyxgogY=; b=Mie/ScMW6O+Yt6gm32IqV04QgD/1rNNtFJ0xd0ufeJUt9qvp4mO6c8I41T/JRuasNnWbpC 7H55mDQpVoTCqEgv137vKCaxz/fqYJ3xo/mXcKE4JJhzUIavBoXwJhY7QY7EmlOe75dbq8 mWbC+25eVu/cChDWFug84iZhAeO6DAPx3CWyaQneqzg64P6sbOmJAQcpHVrNgj+ymPe8xP 0Rwh7BFVQUdHfzKPnIJ1Emh912CPzjLfaOMvwbXsF26mp3dJ4jORtYxOrQUCbnY5SWoQ7M EPE90Oe7Rb4gkMwRN5VziXzFXwNAL4lfBHr19Wt8OF98+AbT7+nBjw4UgL+Pbw==
To: peng.shaofu@zte.com.cn
Cc: spring@ietf.org
References: <202008281131380547490@zte.com.cn>
From: Martin Horneffer <maho@lab.dtag.de>
Message-ID: <05d5bf15-9d7a-a093-d32f-e3af839bd41e@lab.dtag.de>
Date: Fri, 28 Aug 2020 12:13:23 +0200
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.14; rv:68.0) Gecko/20100101 Thunderbird/68.12.0
MIME-Version: 1.0
In-Reply-To: <202008281131380547490@zte.com.cn>
Content-Type: multipart/alternative; boundary="------------173249FF22F82B138A614718"
X-Last-TLS-Session-Version: TLSv1.3
Archived-At: <https://mailarchive.ietf.org/arch/msg/spring/YBja4JxSzcS0sW64P-7wx9W4j8w>
Subject: Re: [spring] to drop or to forward unlabelled (Re: Question on RFC8660)
X-BeenThere: spring@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: "Source Packet Routing in NetworkinG \(SPRING\)" <spring.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/spring>, <mailto:spring-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/spring/>
List-Post: <mailto:spring@ietf.org>
List-Help: <mailto:spring-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/spring>, <mailto:spring-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 28 Aug 2020 10:13:31 -0000

Hi PSF,

limiting the "forward unlabelled" behavior to packets with the 
bottom-of-stack bit set is an interesting idea. I haven't though about 
that yet. In my case that would definitely help and be perfectly ok.

Best regards, Martin


Am 28.08.20 um 05:31 schrieb peng.shaofu@zte.com.cn:
>
>
> Hi Martin,
>
>
> For the matched LFIB entry with unlabeled outgoing forwarding 
> information, I agree your opinion that public traffic need to be 
> forwarded.
>
> I think the above LFIB entry can easiely drop VPN traffic according to 
> no bottom flag of incomming LDP/SR label.
>
>
> Regards,
>
> PSF
>
>
>
> 原始邮件
> *发件人:*MartinHorneffer <maho@lab.dtag.de>
> *收件人:*spring@ietf.org <spring@ietf.org>;
> *日 期 :*2020年08月27日 18:35
> *主 题 :**[spring] to drop or to forward unlabelled (Re: Question on 
> RFC8660)*
> Hello everyone,
>
> may I come back the the question below? Or rather let me update it a little:
>
> In case an SR-MPLS path is broken, should a node rather drop the packet,
> or forward it?
> This can happen whenever the IGP points to a certain next hop, but that
> neither supplies a valid SID, nor allows LDP-stitching for whatever
> reason. For PUSH as well as for CONTINUE.
>
> We have been using MPLS transport and a BGP free core since about two
> decades now, using LDP. In the analog case, LDP creates "unlabelled"
> entries in the LFIB, does the equivalent of a POP operation and forwards
> the packet to the next-hop as chosen by the IGP.
>
> This behavior obviously breaks any traffic that relies on a service
> label, but it can protect some traffic.
> In our case a huge percentage of all traffic still is public IPv4. This
> needs MPLS only for a transport label, be it LDP or SR-MPLS. If this
> traffic gets forwarded unlabelled, it follows an IGP default route to a
> central device, where it is 1) redirected to the correct destination and
> 2) counted in a way that operators can quickly see whether and where
> this kind of failure occurs at some point in the network.
>
> After more operational experience and several internal discussions we
> agreed that we want packets to be forwarded unlabelled rather than
> dropped. Anyone to share, or oppose this position?
>
> Best regards, Martin
>
>
> Am 31.01.20 um 16:50 schrieb Martin Horneffer:
> > Hello everyone,
> >
> > again it seems the interesting questions only show up when applying
> > something to the live network...
> >
> > We ran into something that poses a question related to RFC8660: What
> > is the exact meaning of section 2.10.1, "Forwarding for PUSH and
> > CONTINUE of Global SIDs", when the chosen neighbor doesn't provide a
> > valid MPLS path?
> >
> > The relevant sections reads:
> >
> >       -  Else, if there are other usable next hops, use them to forward
> >          the incoming packet.  The method by which the router "R0"
> >          decides on the possibility of using other next hops is beyond
> >          the scope of this document.  For example, the MCC on "R0" may
> >          chose the send an IPv4 packet without pushing any label to
> >          another next hop.
> >
> > Does the part "send an IPv4 packet without pushing any label" apply to
> > PUSH and CONTINUE, or just to PUSH?
> > Does R0 have to validate that neighbor N can correctly process to
> > packet? Or can it forward the packet regardless?
> >
> > The reason for asking is that we are now seeing issues similar to ones
> > we had when starting with LDP based MPLS about two decades ago:
> > traffic being black holed even though a path to the destination
> > exists, because the MPLS path is interrupted somewhere in the middle.
> >
> > With LDP we know the case of LFIBentries called "unlabelled". While
> > this does break connectivity for many kinds of service, e.g. those
> > relying on an additional service labels, it still works for plain
> > IP(v4) traffic. In our cases, this works perfectly fine for all
> > internal routing and control traffic. And even for IPv4 traffic that
> > gets collected by a central router that injects a default route.
> >
> > However, depending on the exact interpretation of the above paragraph,
> > an implementor might feel obliged to chose the next paragraph:
> >
> >       -  Otherwise, drop the packet.
> >
> > Which is, at least in our case, very unfortunate...
> >
> > Any advice or opinion appreciated!
> >
> >
> > Best regards, Martin
> >
> >
> >
> >
> > _______________________________________________
> > spring mailing list
> > spring@ietf.org
> > https://www.ietf.org/mailman/listinfo/spring
>
> _______________________________________________
> spring mailing list
> spring@ietf.org
> https://www.ietf.org/mailman/listinfo/spring
>
>
>
> _______________________________________________
> spring mailing list
> spring@ietf.org
> https://www.ietf.org/mailman/listinfo/spring