[Babel-users] the routing atomic update wet paint - because *I* care

[Dave Taht]

On Tue, Apr 7, 2015 at 6:19 AM, Matthieu Boutier
<boutier at pps.univ-paris-diderot.fr> wrote:
>> If you work with atomic route replacement even putting ALL of them
>> into a netlink message (or as many as you can fit in) works.
>
> What I understand is that we can't (in general) work with atomic
> *next-hop* replacement (interface index and metric may change).
>
> I proposed a workaround where instead of using two distinct messages
> for "del(r)" and "add(r)" we use one message with "del(r); add(r)".
> Even if it's not necessarily atomic (is it?), it should be faster
> (only one system call, since it was what frightened Dave).

No "fright" here.

Doing an atomic_replace(old_route,new_route) seems *desirable* at high
forwarding rates, and measuring the impact on tcp flows when a route
switch is staged (*with* reordering, *without* loss) is interesting in
the context of evaluating aqm and fq algorithms. Route flaps do nasty
things to networks in general.

It can wait til after babel-1.6 in particular. I'll have a testbed
setup this summer that will pair up an 8 port edgerouter pro (weak
cpu) with an 8 port rangeley (much less weak cpu) to fiddle with as a
fault tolerant configuration, that I will regularly be driving to
saturating loads.

ship 1.6 soon please so it lands in edgerouter os 1.7. :) Sorry for
the distractions.

> Matthieu
>



-- 
Dave Täht
We CAN make better hardware, ourselves, beat bufferbloat, and take
back control of the edge of the internet! If we work together, on
making it:

https://www.kickstarter.com/projects/onetswitch/onetswitch-open-source-hardware-for-networking