MASQUERADE vs. SNAT

Harald Welte laforge@gnumonks.org
Wed, 30 May 2001 12:43:44 -0300


On Wed, May 30, 2001 at 01:59:28PM +0200, Alexander W. Janssen wrote:
> Hi,
> 
> the practical difference is, that MASQUERADE can only map several ports
> where SNAT can map to IP and ports together.

? There is no difference between MASQUERADE and SNAT, but ...

> Another point is that you assign a interface to the MASQUERADE-target (where
> the actual IP doesn't matter), SNAT uses a specific IP which you have to know
> beforehand (not possible with dynamic assigned IP's).

... this. This is also explained in some of the documents (NAT-HOWTO, if i
remember correctly).  In addition, all existing conntrack/nat mappings
are deleted, once a interface goes down / comes up, when you use the
MASQUERADE target.

> The more deeper difference is afaik that the MASQUERADE-ocde uses NAT-helpers
> to handle weird protocols like ICQ. I'm not quite sure, but i think that's
> it. 

???? MASQUERADE and SNAT don't doo too much at all. they just set up a 
NAT mapping. Everything else is handled by the NAT core, which is generic
and used for MASQUERADE, SNAT, DNAT, SAME, BALANCE, ...

So there is _no_ difference with regart to NAT helpers.

> Hope i'm not talking complete bullsh*t, i'm allways open for useful criticism.

ok :) see above.

> Cheers, Alex.

-- 
Live long and prosper
- Harald Welte / laforge@gnumonks.org                http://www.gnumonks.org
============================================================================
GCS/E/IT d- s-: a-- C+++ UL++++$ P+++ L++++$ E--- W- N++ o? K- w--- O- M- 
V-- PS+ PE-- Y+ PGP++ t++ 5-- !X !R tv-- b+++ DI? !D G+ e* h+ r% y+(*)