* Questions about the workings of iptables
@ 2004-07-26 12:51 Small, Jim
2004-07-26 13:17 ` Antony Stone
0 siblings, 1 reply; 9+ messages in thread
From: Small, Jim @ 2004-07-26 12:51 UTC (permalink / raw)
To: netfilter
I've been using netfilter/iptables for some time and I'm planning on doing a
presentation on it to a local user group. I've used most of the major
firewalls (pf, ipfilter, iptables, PIX, and CheckPoint) and I've been
comparing features. Note--I know there are other firewalls like G2,
NetScreen, and CyberGuard and I'm not trying to knock them by leaving them
out!
Two popular firewall features I'm exploring are stateful inspection and
initial TCP sequence randomizing. I've looked in the archives and couldn't
find very much. So please allow me to ask:
1) How extensive is IPTables stateful packet filtering? Especially with
TCP and the recent reset paranoia
(http://www.uniras.gov.uk/vuls/2004/236929/index.htm), what is checked for
stateful TCP inspection? Are the sequence numbers carefully scrutinized as
part of the state check? For an excellent paper on TCP state checking, I
like the following:
http://home.iae.nl/users/guido/papers/tcp_filtering.ps.gz
Is there a listing of everything the connection tracking modules do?
If connection tracking or stateful inspection does not include TCP sequence
checking, is there a way to add it? Is it well tested/supported?
2) Can IPTables randomize initial TCP sequence numbers? If not, is there
an add-on that can? I saw IP Personality, that's not really what I want. I
would prefer something that can randomize all initial TCP sequence numbers
traversing an IPTables firewall.
As to why this is important, many clients have weak TCP ISN (Initial TCP
Sequence Number) generators. While Linux is good in this aspect, many
clients like Windows are not. And it is not always possible to replace
Windows clients or other clients with weak ISNs. Proxying might be an
option, but I would really like to know if there is a stateful firewall
option or add-on.
Thanks,
<> Jim
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: Questions about the workings of iptables
2004-07-26 12:51 Questions about the workings of iptables Small, Jim
@ 2004-07-26 13:17 ` Antony Stone
2004-07-26 14:04 ` Sven Schuster
2004-07-27 7:20 ` Ashutosh
0 siblings, 2 replies; 9+ messages in thread
From: Antony Stone @ 2004-07-26 13:17 UTC (permalink / raw)
To: netfilter
On Monday 26 July 2004 1:51 pm, Small, Jim wrote:
> 1) How extensive is IPTables stateful packet filtering? Are the sequence
> numbers carefully scrutinized as part of the state check?
No. AFAIK the connection tracking in netfilter checks only src+dst IP+port,
nothing else.
> Is there a listing of everything the connection tracking modules do?
Netfilter source code is probably your best bet here.
> If connection tracking or stateful inspection does not include TCP sequence
> checking, is there a way to add it?
I do not know of a patch to provide this. It would be in patch-o-matic if
there is one.
> 2) Can IPTables randomize initial TCP sequence numbers?
No. Netfilter doesn't change the packets on their way past, except for:
- NAT, which will change as little as necessary to achieve the required
ersult (ie source port numbers don't get changed if it can be helped)
- mangle rules which specifically change things in the headers
> If not, is there an add-on that can?
How about a Linux or BSD-based proxy server :) ?
You really want something which is going to generate its own sequence numbers
to do this properly (and a proxy server would help with your first question
above as well).
I know that's not really the answer you wanted, but I'm not aware of any way
to do these with netfilter itself.
Regards,
Antony.
--
Success is a lousy teacher. It seduces smart people into thinking they can't
lose.
- William H Gates III
Please reply to the list;
please don't CC me.
^ permalink raw reply [flat|nested] 9+ messages in thread* Re: Questions about the workings of iptables
2004-07-26 13:17 ` Antony Stone
@ 2004-07-26 14:04 ` Sven Schuster
2004-07-27 7:20 ` Ashutosh
1 sibling, 0 replies; 9+ messages in thread
From: Sven Schuster @ 2004-07-26 14:04 UTC (permalink / raw)
To: netfilter
[-- Attachment #1: Type: text/plain, Size: 1127 bytes --]
Hi Antony, hi Jim,
On Mon, Jul 26, 2004 at 02:17:33PM +0100, Antony Stone told us:
> On Monday 26 July 2004 1:51 pm, Small, Jim wrote:
>
> > 1) How extensive is IPTables stateful packet filtering? Are the sequence
> > numbers carefully scrutinized as part of the state check?
>
> No. AFAIK the connection tracking in netfilter checks only src+dst IP+port,
> nothing else.
>
> > Is there a listing of everything the connection tracking modules do?
>
> Netfilter source code is probably your best bet here.
>
> > If connection tracking or stateful inspection does not include TCP sequence
> > checking, is there a way to add it?
>
> I do not know of a patch to provide this. It would be in patch-o-matic if
> there is one.
look at the TCP window tracking patch in pom-ng. Harald has submitted it
to davem for inclusion in the mainline 2.6.9 kernel recently. Don't know
if it will be included in 2.4, too.
regards,
Sven
--
Linux zion 2.6.8-rc2 #1 Sun Jul 18 15:00:48 CEST 2004 i686 athlon i386 GNU/Linux
16:02:43 up 7 days, 17:31, 1 user, load average: 1.06, 1.08, 1.02
[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: Questions about the workings of iptables
2004-07-26 13:17 ` Antony Stone
2004-07-26 14:04 ` Sven Schuster
@ 2004-07-27 7:20 ` Ashutosh
2004-07-27 8:47 ` Antony Stone
2004-07-27 13:41 ` John A. Sullivan III
1 sibling, 2 replies; 9+ messages in thread
From: Ashutosh @ 2004-07-27 7:20 UTC (permalink / raw)
To: netfilter
> No. AFAIK the connection tracking in netfilter checks only src+dst
> IP+port,
> nothing else.
.. And the Protocol
--
Ashutosh Naik
.
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: Questions about the workings of iptables
2004-07-27 7:20 ` Ashutosh
@ 2004-07-27 8:47 ` Antony Stone
2004-07-27 13:41 ` John A. Sullivan III
1 sibling, 0 replies; 9+ messages in thread
From: Antony Stone @ 2004-07-27 8:47 UTC (permalink / raw)
To: netfilter
On Tuesday 27 July 2004 8:20 am, Ashutosh wrote:
> > No. AFAIK the connection tracking in netfilter checks only src+dst
> > IP+port, nothing else.
>
> .. And the Protocol
Er, well, yes. I kind of took this bit for granted, but you're correct; a
conntrack entry for a TCP connection between a.b.c.d:e and v.w.x.y:z will not
match UDP packets on the same IPs and ports :)
Also, conntrack entries take care of non-TCP/UDP connections (eg ICMP, ESP)
which do not have port numbers, therefore src+dst IP (and protocol) are the
only things matched here.
Regards,
Antony.
--
Normal people think "If it ain't broke, don't fix it".
Engineers think "If it ain't broke, it doesn't have enough features yet".
Please reply to the list;
please don't CC me.
^ permalink raw reply [flat|nested] 9+ messages in thread* Re: Questions about the workings of iptables
2004-07-27 7:20 ` Ashutosh
2004-07-27 8:47 ` Antony Stone
@ 2004-07-27 13:41 ` John A. Sullivan III
2004-07-27 14:22 ` Antony Stone
1 sibling, 1 reply; 9+ messages in thread
From: John A. Sullivan III @ 2004-07-27 13:41 UTC (permalink / raw)
To: Ashutosh; +Cc: netfilter
On Tue, 2004-07-27 at 03:20, Ashutosh wrote:
> > No. AFAIK the connection tracking in netfilter checks only src+dst
> > IP+port,
> > nothing else.
>
> .. And the Protocol
and I would assume there is a timer - John
--
John A. Sullivan III
Chief Technology Officer
Nexus Management
+1 207-985-7880
john.sullivan@nexusmgmt.com
---
If you are interested in helping to develop a GPL enterprise class
VPN/Firewall/Security device management console, please visit
http://iscs.sourceforge.net
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: Questions about the workings of iptables
2004-07-27 13:41 ` John A. Sullivan III
@ 2004-07-27 14:22 ` Antony Stone
2004-07-28 4:00 ` Ashutosh Naik
0 siblings, 1 reply; 9+ messages in thread
From: Antony Stone @ 2004-07-27 14:22 UTC (permalink / raw)
To: netfilter
On Tuesday 27 July 2004 2:41 pm, John A. Sullivan III wrote:
> On Tue, 2004-07-27 at 03:20, Ashutosh wrote:
> > > No. AFAIK the connection tracking in netfilter checks only src+dst
> > > IP+port,
> > > nothing else.
> >
> > .. And the Protocol
>
> and I would assume there is a timer - John
Oh yes, there are timers, but that's not part of the information from the
original packet which gets matched in future packets, which is what I was
discussing, and what I think the original question was about (!?).
Regards,
Antony.
--
"Linux is going to be part of the future. It's going to be like Unix was."
- Peter Moore, Asia-Pacific general manager, Microsoft
Please reply to the list;
please don't CC me.
^ permalink raw reply [flat|nested] 9+ messages in thread* Re: Questions about the workings of iptables
2004-07-27 14:22 ` Antony Stone
@ 2004-07-28 4:00 ` Ashutosh Naik
0 siblings, 0 replies; 9+ messages in thread
From: Ashutosh Naik @ 2004-07-28 4:00 UTC (permalink / raw)
To: netfilter
On Tue, 27 Jul 2004 15:22:40 +0100, Antony Stone
<Antony@Soft-Solutions.co.uk> wrote:
> On Tuesday 27 July 2004 2:41 pm, John A. Sullivan III wrote:
>
>> On Tue, 2004-07-27 at 03:20, Ashutosh wrote:
>> > > No. AFAIK the connection tracking in netfilter checks only src+dst
>> > > IP+port,
>> > > nothing else.
>> >
>> > .. And the Protocol
>>
>> and I would assume there is a timer - John
>
> Oh yes, there are timers, but that's not part of the information from the
> original packet which gets matched in future packets, which is what I was
> discussing, and what I think the original question was about (!?).
Yes, What actually get matched is only the 5 tuples..
srcIP+destIP+srcport+destport+Protocol
--
Ashutosh Naik
Teneoris Networks India Pvt. Ltd.
^ permalink raw reply [flat|nested] 9+ messages in thread
* RE: Questions about the workings of iptables
@ 2004-07-26 13:43 Jason Opperisano
0 siblings, 0 replies; 9+ messages in thread
From: Jason Opperisano @ 2004-07-26 13:43 UTC (permalink / raw)
To: Small, Jim, netfilter
> 1) How extensive is IPTables stateful packet filtering? Especially with
> TCP and the recent reset paranoia
> (http://www.uniras.gov.uk/vuls/2004/236929/index.htm), what is checked for
> stateful TCP inspection? Are the sequence numbers carefully scrutinized as
> part of the state check? For an excellent paper on TCP state checking, I
> like the following:
> http://home.iae.nl/users/guido/papers/tcp_filtering.ps.gz.ps.gz
the "tcp-window-tracking patch" available in patch-o-matic is based upon the paper by Guido van Rooij that you reference.
-j
^ permalink raw reply [flat|nested] 9+ messages in thread
end of thread, other threads:[~2004-07-28 4:00 UTC | newest]
Thread overview: 9+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2004-07-26 12:51 Questions about the workings of iptables Small, Jim
2004-07-26 13:17 ` Antony Stone
2004-07-26 14:04 ` Sven Schuster
2004-07-27 7:20 ` Ashutosh
2004-07-27 8:47 ` Antony Stone
2004-07-27 13:41 ` John A. Sullivan III
2004-07-27 14:22 ` Antony Stone
2004-07-28 4:00 ` Ashutosh Naik
-- strict thread matches above, loose matches on Subject: below --
2004-07-26 13:43 Jason Opperisano
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox