Ok, here is a diff I did on the two different ipfs.  The ones with <
at the beginning are the rules without keep state after rl0 and the
ones with > are what happened after I added keep state to rl0.  I
don't really see much difference, except that the number of blocked
packets is greater for very last rule, though I expect that that is
mostly a function of time, as I had that set of rule running for over
an hour.  Do you see anything here that might explain my difficulties?
I would prefer to know exactly what was going on so that I might be
able to troubleshoot further problems on my own with greater success.

Thanks again for your help.

< 0 @1 pass out quick on rl0 all
< 0 @2 pass out quick on lo0 all
---
0 @1 pass out quick on rl0 all keep state
8 @2 pass out quick on lo0 all
4c4
< 0 @4 pass out quick on vr0 proto udp from any to 64.59.184.13/32
port = domain keep state
---
54 @4 pass out quick on vr0 proto udp from any to 64.59.184.13/32 port = domain 
keep state
6c6
< 0 @6 pass out quick on vr0 proto udp from any to 64.59.184.15/32
port = domain keep state
---
14 @6 pass out quick on vr0 proto udp from any to 64.59.184.15/32 port = domain 
keep state
13c13
< 0 @13 pass out quick on vr0 proto tcp from any to any port = ftp
flags S/FSRPAU keep state
---
4 @13 pass out quick on vr0 proto tcp from any to any port = ftp flags S/FSRPAU 
keep state
29,31c29,31
< 0 @29 block out log first quick on vr0 all
< 0 @1 pass in quick on rl0 all
< 0 @2 pass in quick on lo0 all
---
120 @29 block out log first quick on vr0 all
4922 @1 pass in quick on rl0 all keep state
8 @2 pass in quick on lo0 all
35,36c35,36
< 4 @6 pass in quick on vr0 proto tcp from any to 192.168.0.254/32
port = 50505 flags S/FSRPAU keep state
< 3 @7 pass in quick on vr0 proto udp from any to 192.168.0.254/32
port = 50505 keep state
---
2328 @6 pass in quick on vr0 proto tcp from any to 192.168.0.254/32 port = 
50505 flags S/FSRPAU keep state
5605 @7 pass in quick on vr0 proto udp from any to 192.168.0.254/32 port = 
50505 keep state
39c39
< 0 @10 block in quick on vr0 from 192.168.0.0/16 to any
---
1 @10 block in quick on vr0 from 192.168.0.0/16 to any
53c53
< 0 @24 block in quick on vr0 proto icmp from any to any icmp-type echo
---
8 @24 block in quick on vr0 proto icmp from any to any icmp-type echo
55c55
< 0 @26 block in log first quick on vr0 proto tcp/udp from any to any
port = netbios-ns
---
4 @26 block in log first quick on vr0 proto tcp/udp from any to any port = 
netbios-ns
57c57
< 0 @28 block in log first quick on vr0 proto tcp/udp from any to any
port = netbios-ssn
---
45 @28 block in log first quick on vr0 proto tcp/udp from any to any port = 
netbios-ssn
64c64
< 0 @35 block in log first quick on vr0 all
---
1236 @35 block in log first quick on vr0 all


On 6/9/06, Larry Moore <[EMAIL PROTECTED]> wrote:

----- Original Message -----
From: "Trevor Osatchuk" <[EMAIL PROTECTED]>
To: "Larry Moore" <[EMAIL PROTECTED]>
Cc: <[email protected]>
Sent: Saturday, June 10, 2006 12:02 AM
Subject: Re: Problems with Azureus


> Larry, I tired all of your suggestions one at a time and all the
> iterations to see what difference they made.  the only one that made
> the difference was adding keep state on the pass in/out on rl0, my
> internal interface.  I thought that keep state was the way ipf kept
> track of 'established' connections.  If a session was started on a
> particular port, or in this case an interface, once it was allowed
> through keep state would then let the session continue with no further
> checking.  Why would it matter if I had keep state on rl0 since I am
> passing everything be default?  It certainly had made a difference,
> but why?
>

I suspect it's to do with the compiled in default operation of "block all".

Not seeing the results of ipfstat -iohn would lead to guessing though
perhaps you used this command before and after making keep-state rules on
rl0 and observed the changes.

Cheers,

Larry.


Reply via email to