Re: [PATCH v2 0/4] IPVS full NAT support + netfilter 'ipvs' match support

[Hannes Eder <heder@google.com>]

On Tue, Sep 29, 2009 at 16:51, Simon Horman <horms@verge.net.au> wrote:
> On Tue, Sep 29, 2009 at 02:35:15PM +0200, Hannes Eder wrote:
>> The following series implements full NAT support for IPVS.  The
>> approach is via a minimal change to IPVS (make friends with
>> nf_conntrack) and adding a netfilter matcher, kernel- and user-space
>> part, i.e. xt_ipvs and libxt_ipvs.
>
> Its a bit late in the day for me to review the code, but I have a few
> quick comments.
>
>>
>> Example usage:
>>
>> % ipvsadm -A -t 192.168.100.30:80 -s rr
>> % ipvsadm -a -t 192.168.100.30:80 -r 192.168.10.20:80 -m
>> # ...
>>
>> # Source NAT for VIP 192.168.100.30:80
>> % iptables -t nat -A POSTROUTING -m ipvs --vaddr 192.168.100.30/32 \
>> > --vport 80 -j SNAT --to-source 192.168.10.10
>>
>> or SNAT-ing only a specific real server:
>>
>> % iptables -t nat -A POSTROUTING --dst 192.168.11.20 \
>> > -m ipvs --vaddr 192.168.100.30/32 -j SNAT --to-source 192.168.10.10
>
> If the iptables rule is not in place does LVS just use
> its old NAT behaviour?

Yes, without iptables rules LVS NAT does DNAT.

>> First of all, thanks for all the feedback.  This is the changelog for v2:
>>
>> - Make ip_vs_ftp work again.  Setup nf_conntrack expectations for
>>   related data connections (based on Julian's patch see
>>   http://www.ssi.bg/~ja/nfct/) and let nf_conntrack/nf_nat do the
>>   packet mangling and the TCP sequence adjusting.
>>
>>   This change rises the question how to deal with ip_vs_sync?  Does it
>>   work together with conntrackd?  Wild idea: what about getting rid of
>>   ip_vs_sync and piggy packing all on nf_conntrack and use conntrackd?
>>
>>   Any comments on this?
>
>    That sounds like a reasonable suggestion.
>
>    I think that ip_vs_sync came along before conntrackd
>    and no one has given much thought to merging the functionality.

Okay, I'll dig further in this direction.

Cheers,
-Hannes