radium closes connection after rebooting argus sensors

Wolfgang Barth wob at swobspace.de
Mon May 21 08:04:54 EDT 2007


Hey Carter,

> I'll take a look today.  Any foolproof way of reproducing the bug?

Yes :-)

1) Stopping argus on argus sensor is uncritical:

May 21 14:00:28 elix01 radium[19858]: 2007-05-21 14:00:28 connect to
172.17.130.82:561 failed 'Connection refused'

and radium tries, and tries, and tries.

2) now simulate "host unreachable" (because I can't shutdown my productive
gateway during workhours). On argus sensor I did:

iptables -I INPUT -s <radium-host> -p TCP --dport 561 \
   -j REJECT --reject-with icmp-host-unreachable

The sensor sends an icmp-host-unreachable, if radium tries to contact
argus. In the logfile:

May 21 14:02:53 elix01 radium[19858]: 2007-05-21 14:02:53 connect to
172.17.130.82:561 failed 'No route to host'

and stops contacting the argus sensor. No more retries. I had to restart
radium.

Wolfgang

PS: don't forget deleting the iptables rule after testing, replace -I with -D.
-- 
<wob (at) swobspace de> * http://www.swobspace.de



More information about the argus mailing list