Re: [PATCH] NETFILTER module xt_hmark new target for HASH MARK

[Hans Schillstrom <hans@schillstrom.com>]

On Thursday, February 03, 2011 17:01:27 Pablo Neira Ayuso wrote:
> On 03/02/11 16:42, Pablo Neira Ayuso wrote:
> > On 03/02/11 15:23, Hans Schillstrom wrote:
> >> On Thu, 2011-02-03 at 14:51 +0100, Pablo Neira Ayuso wrote:
> >>> On 03/02/11 14:34, Hans Schillstrom wrote:
> >>> this assumption is not valid in NAT handlings.
> >>
> >> That's true, because I want to avoid conntrack
> >>
> >>> If you want consistent hashing with NAT handlings you'll have to make
> >>> this stateful and use the conntrack source and reply directions of the
> >>> original tuples (thus making it stateful). That may be a problem because
> >>> some people may want to use this without enabling connection tracking.
> >>
> >> What about a compilation switch or a sysctl ?
> > 
> > or better some option for iptables.
> 
> Hm, this is actually not straight forward to implement, you'll have to
> use hook functions to avoid the module dependencies with conntrack and
> that's pretty annoying.
> 
> I don't come up with a good solution for this.

A configuration switch might be OK.

> 
> >>> Are you using this for (uplink) load balancing?
> >>
> >> Actually in both ways 
> >>  - in front of a bunch of ipvs
> 
> to make some preliminary load-sharing between the load balancers?

Yes that's right
and in the payloads send the return traffic in the same path.

> 
> >>  - and in the payloads for outgoing traffic.
> 
> and then to select the uplink, right?
> 

Yes.
It also has the same role for cluster originated traffic to spread the load over multiple interfaces,
and catch the return traffic.