Quoting "James B. Byrne" <byrnejb at harte-lyne.ca>: > I went to reload (iptables-restore) my iptables configuration and > obtained an error at the COMMIT statement. No further details were > provided even when I ran restore with the -v option. > > I determined that none of my backed up configuration files going > back to October will load either. This is more than passing strange > because I altered and uploaded the iptables configuration on this > host several times in December alone. These alterations certainly > applied without error at the time. > > Through painful trial and error (it is a fairly large configuration) > I discovered that I cannot add any rule using the __recent__ module. > Adding a single rule referencing that module inevitably results in > a load error reported at the following COMMIT statement. An example > of an actual rule that fails follows: > > . . . > :BRUTE_FORCE - [0:0] > . . . > -A BRUTE_FORCE -p tcp -m tcp -m state -m recent --set -i eth0 > --dport 22 --state NEW > -A BRUTE_FORCE -m comment -j RETURN --comment "Return to calling chain" > COMMIT > > Perhaps I am missing something obvious but as far as I can determine > the rule using the recent module should simply add all traffic > coming in over i/f eth0 consigned to port 22 on any ip-addr to the > DEFAULT list. I do not expect it to give an error. If I remove > this statement then the iptables file loads without error. > > An interesting thing happens if I simply add a trailing -j to the > end of recent module rule above. It fails with this specific error: > > -c packet counter not numeric > > Does anyone see what I am doing wrong? I don't think you need the -m state .. >From the iptables man page ... # iptables -A FORWARD -m recent --name badguy --rcheck --seconds 60 -j DROP # iptables -A FORWARD -p tcp -i eth0 --dport 139 -m recent --name badguy --set -j DROP Barry