Martin Pieuchot([email protected]) on 2016.06.08 20:50:29 +0200:
> On 08/06/16(Wed) 19:51, Sebastian Benoit wrote:
> > [...] 
> > i dont see why this would be a problem
> > 
> > however:
> > 
> > +           ... if we were going to use
> > +                * the last available route, but it got removed, we'll hit
> > +                * the end of the list and then pick the first route.
> > 
> > would this make the first route get more traffic than the others,
> > statistically?
> 
> What do you mean by "statistically"?  The comment apply to route lookups
> done by a CPU while another CPU is modifying the multipath list.  This
> should be a completely negligible amount of lookups.

thanks for explaining that. i thougt this might happen more often.

> Now assuming that you are voluntarily generating a lot of multipath route
> changes to trigger this case, what you call "the first" route will most
> likely be a different route after every change.  So it should not matter ;)

Reply via email to