Description of problem: iptables setup with port translation from common port to high port fails to stop or restart (which is really the same thing as "stop", "start"), instead "/sbin/service iptables stop" hangs forever as does any strace attempt on the modprobe part. Modprobe utilizes 99% of CPU, driving load from 0.00 to 1.00. Services active on the high port include things such as tomcat. I'm able to reproduce this with all kernels I've tried this on including: 2.4.20-18.9 2.4.20-19.9 2.4.20-20.9 2.4.22-1.2051.nptl 2.4.22-1.2061.nptl % ps ax .... 23069 pts/3 S 0:00 /bin/sh /sbin/service iptables stop 23072 pts/3 S 0:00 /bin/sh /etc/init.d/iptables stop 23111 pts/3 R 0:38 modprobe -r ipt_REDIRECT 23204 pts/3 R 0:00 ps ax Meanwhile: % /sbin/service iptables status Firewall is stopped. Attached is a sample /etc/sysconfig/iptables which I'm able to get to hang. Version-Release number of selected component (if applicable): iptables-1.2.8-12.1 How reproducible: Always Additional info: Workaround is to reboot when modifying iptables rulesets but that really blows.
Created attachment 94772 [details] Test case /etc/sysconfig/iptables Oh yeah, haven't touched /etc/sysconfig/iptables-config, it's at its defaults.
This is a kernel issue.
Looks like a duplicate of bug #103177 which is about Valhalla, but contains a pointer to a fix.
Resolving as duplicate as earlier bug #103177 exists with the same symptoms. *** This bug has been marked as a duplicate of 103177 ***