Typically (or at least somewhat occasionally) after a reboot of my
 shorewall[-lite] machine I find that I end up with conntrack table
entries for unNATted connections such as:

# conntrack -L -p udp --dport 5060 -d 99.232.11.14
udp      17 59 src=10.75.22.8 dst=99.232.11.14 sport=5060 dport=5060 
packets=5472 bytes=3031488 [UNREPLIED] src=99.232.11.14 dst=10.75.22.8 
sport=5060 dport=5060 packets=0 bytes=0 mark=1 use=2

These are supposed to be NATted and will be so if I flush the offending
entries from the conntrack table:

# conntrack -D -p udp --dport 5060 -d 99.232.11.14
udp      17 52 src=10.75.22.8 dst=99.232.11.14 sport=5060 dport=5060 
packets=5558 bytes=3079132 [UNREPLIED] src=99.232.11.14 dst=10.75.22.8 
sport=5060 dport=5060 packets=0 bytes=0 mark=1 use=2
conntrack v1.0.0 (conntrack-tools): 1 flow entries have been deleted.
# conntrack -L -p udp --dport 5060 -d 99.232.11.14
udp      17 3593 src=10.75.22.8 dst=99.232.11.14 sport=5060 dport=5060 
packets=1 bytes=554 src=99.232.11.14 dst=67.193.214.242 sport=5060 dport=5060 
packets=1 bytes=516 mark=257 use=2

Clearly there is some kind of a race or timing issue involved here.
Perhaps it's as simple as these conntrack entries get established
before shorewall has a chance to get loaded.  Or maybe it's a more
sinister race in shorewall's startup.  My gut instinct makes me feel
like it's the former though.

I wonder as a matter of routine, if shorewall should flush the entire
conntrack table when it loads to prevent this sort of thing.

Thoughts?

b.

Attachment: signature.asc
Description: OpenPGP digital signature

------------------------------------------------------------------------------
RSA(R) Conference 2012
Mar 27 - Feb 2
Save $400 by Jan. 27
Register now!
http://p.sf.net/sfu/rsa-sfdev2dev2
_______________________________________________
Shorewall-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/shorewall-users

Reply via email to