[prev in list] [next in list] [prev in thread] [next in thread] 

List:       bird-users
Subject:    Re: BGP route back
From:       Roman Romanyak <roman.romanyak () gmail ! com>
Date:       2019-01-22 18:52:37
Message-ID: CAO=ctuk3-bQQ+ySKOCDfTCtRRu0QgsoftGh3gtZML4ds37ze6Q () mail ! gmail ! com
[Download RAW message or body]

Thank you for the advise, Ondrej. I've heard that some l3 switches do not
support add-path feature in VRFs, but in this case I'm using routers, so
this should work fine.


-
Roman.



On Tue, Jan 22, 2019 at 9:38 AM Ondrej Zajicek <santiago@crfreenet.org>
wrote:

> On Fri, Jan 18, 2019 at 01:14:47PM -0500, Roman Romanyak wrote:
> > Hello Bird users,
> >
> > Does anyone know if there is a way to announce a BGP route back to the
> > router it was received from.
> > I need to do this for the following scenario.
>
> Hi
>
> I recently thought about a setup where a route-reflector modifies routes
> and propagates modified routes to its clients. To avoid issue with
> announcing route back and with (like you noticed) and with withdrawing
> original route (like Alexander Zubkov noticed) there is a simple solution:
>
> Use two RIB routing tables on clients and two single-directional BGP
> sessions - first RIB table collects routes from EBGP peers and uses first
> BGP session to propagate that to route-reflector, then second BGP session
> is used to propagate all routes back from route-reflector to second table
> in clients, while second table is connected to kernel/FIB.
>
> Such setup could be done in BIRD, i am just wondering whether client part
> of such setup could be done with common L3 switches with BGP support.
>
> --
> Elen sila lumenn' omentielvo
>
> Ondrej 'Santiago' Zajicek (email: santiago@crfreenet.org)
> OpenPGP encrypted e-mails preferred (KeyID 0x11DEADC3, wwwkeys.pgp.net)
> "To err is human -- to blame it on a computer is even more so."
>

[Attachment #3 (text/html)]

<div dir="ltr"><div>Thank you for the advise, Ondrej. I&#39;ve heard that some l3 \
switches do not support add-path feature in VRFs, but in this case I&#39;m using \
routers, so this should work fine.</div><div><br></div><div><br></div><div>- \
<br></div><div>Roman.<br></div><div><br></div><div><br></div></div><br><div \
class="gmail_quote"><div dir="ltr" class="gmail_attr">On Tue, Jan 22, 2019 at 9:38 AM \
Ondrej Zajicek &lt;<a \
href="mailto:santiago@crfreenet.org">santiago@crfreenet.org</a>&gt; \
wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px \
0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">On Fri, Jan 18, 2019 \
at 01:14:47PM -0500, Roman Romanyak wrote:<br> &gt; Hello Bird users,<br>
&gt; <br>
&gt; Does anyone know if there is a way to announce a BGP route back to the<br>
&gt; router it was received from.<br>
&gt; I need to do this for the following scenario.<br>
<br>
Hi<br>
<br>
I recently thought about a setup where a route-reflector modifies routes<br>
and propagates modified routes to its clients. To avoid issue with<br>
announcing route back and with (like you noticed) and with withdrawing<br>
original route (like Alexander Zubkov noticed) there is a simple solution:<br>
<br>
Use two RIB routing tables on clients and two single-directional BGP<br>
sessions - first RIB table collects routes from EBGP peers and uses first<br>
BGP session to propagate that to route-reflector, then second BGP session<br>
is used to propagate all routes back from route-reflector to second table<br>
in clients, while second table is connected to kernel/FIB.<br>
<br>
Such setup could be done in BIRD, i am just wondering whether client part<br>
of such setup could be done with common L3 switches with BGP support.<br>
<br>
-- <br>
Elen sila lumenn&#39; omentielvo<br>
<br>
Ondrej &#39;Santiago&#39; Zajicek (email: <a href="mailto:santiago@crfreenet.org" \
target="_blank">santiago@crfreenet.org</a>)<br> OpenPGP encrypted e-mails preferred \
(KeyID 0x11DEADC3, <a href="http://wwwkeys.pgp.net" rel="noreferrer" \
target="_blank">wwwkeys.pgp.net</a>)<br> &quot;To err is human -- to blame it on a \
computer is even more so.&quot;<br> </blockquote></div>



[prev in list] [next in list] [prev in thread] [next in thread]