<div dir="auto"><div><br><div class="gmail_extra"><br><div class="gmail_quote">On 22 May 2017 11:28, "Toke Høiland-Jørgensen" <<a href="mailto:toke@toke.dk">toke@toke.dk</a>> wrote:<br type="attribution"><blockquote class="quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="quoted-text">Dean <<a href="mailto:dluga93@gmail.com">dluga93@gmail.com</a>> writes:<br>
<br>
> Hello,<br>
><br>
> This is another try at extending the static and OSPFv3 protocols to<br>
> support Source Address Dependent Routing (SADR), also called Source<br>
> Specific Routing. This basically means that routing will take into<br>
> account not only the destination address, but the source address as<br>
> well.<br>
<br>
</div>Nice work! After a quick look at this (I'll go through it in more detail<br>
later), it looks like I can use the core parts of this for Babel as<br>
well. :)<br>
<br>
It looks like a separate channel is needed for SADR routes; (right?) but<br>
can SADR and non-SADR ipv6 routes co-exist in the same FIB?<br>
<font color="#888888"><br>
-Token<br></font></blockquote></div></div></div><div dir="auto"><br></div><div dir="auto">Yes, they can. You can use two kernel protocols, one with the SADR channel and the other with ipv6. Both sets of routes are stored in the same kernel table. </div><div dir="auto"><div class="gmail_extra"><div class="gmail_quote"><blockquote class="quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><font color="#888888">
</font></blockquote></div><br></div><div class="gmail_extra" dir="auto">But the kernel bug is still present. Using both types of routes in the same table gives undefined behavior. A hacky workaround would be to replace ::/0 sources with 2000::/3 in netlink, but it would reduce the set of accepted prefixes.</div></div></div>