Nat Code Confusion
Hi,
The confused one is back. As usual, a couple of grey areas:
1. Under the data_fixup section for nat, e.g. amanda, there's this line:
newip = ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.ip;
What's the rationale for selecting the direction to use?
2. Under the nat_expected similar for amanda, the following lines:
newdstip = ct->tuplehash[IP_CT_DIR_REPLY].tuple.src.ip;
newsrcip = master->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.ip;
Again, what's the rationale for selecting the direction and in each case, ct or master to use?
3. Just a wild observation. Is the Harald the official spokesman for netfilter as it seems he's the only guy answering all the questions.
Ok, that's it.
Ricky
_________________________________________________________________
Keep track of Singapore stock prices. http://www.msn.com.sg/money/
Quoted 8 lines · Ricky Foo
The confused one is back. As usual, a couple of grey areas:
1. Under the
data_fixupsection for nat, e.g. amanda, there's this line:newip = ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.ip;
What's the rationale for selecting the direction to use?
Quoted 7 lines
2. Under the
nat_expectedsimilar for amanda, the following lines:newdstip = ct->tuplehash[IP_CT_DIR_REPLY].tuple.src.ip; newsrcip = master->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.ip; Again, what's the rationale for selecting the direction and in eachcase, ct or master to use?
A connection tracking entry is directed, the same way the TCP/UDP session was established (DIR_ORIGINAL is the direction of the first packte in the connection).
In the conntrack helper you need to carefully chose the right direction, since the REPLY tuples get modified by the nat code.
Quoted 2 lines
3. Just a wild observation. Is the Harald the official spokesman for netfilter as it seems he's the only guy answering all the questions.
Well, it's not really true that I answer all questions. I jus read through this list every couple of days and catch up with all unreplied stuff.
Answering to lots of questons on the mailinglists was one of the reasons why I was invited to the coreteam some three years ago, so it seems to be some kind of 'tradition'.
Apart from that, I'm the only core team member who gets (partially) sponsored by a company to work on netfilter/iptables, so I'm likely to have the most time of all of us to spend on the project.
Quoted 2 lines
Ok, that's it.
Ricky
Signature · 5 lines
- Harald Welte <[email protected]> http://www.netfilter.org/ ============================================================================ "Fragmentation is like classful addressing -- an interesting early architectural error that shows how much experimentation was going on while IP was being designed." -- Paul Vixie