Possible bug in trace with conn=

[Sebastian Auriol]

Hello,

I have a host where I am using conn=worst and trace in bb-hosts. It seems that if one of the IP addresses is down, hobbit is doing a traceroute to the wrong IP address - I expect a traceroute to the IP address that is down...

In this example, IPAdress4 is unreachable but hobbit does a traceroute to IPAdress1.
Here is the conn page:


 Mon Sep 3 18:15:31 2007 conn NOT ok

Service conn on MyRouter is not OK : Host responds to ping on 4 of 5 IP's


System unreachable for 3 poll periods (567 seconds)

green IPAdress1 is alive (43 ms)
green IPAdress2 is alive (22 ms)
green IPAdress3 is alive (44 ms)
red IPAdress4 is unreachable
green IPAdress5 is alive (1 ms)
Traceroute results:
traceroute to IPAdress1 (IPAdress1), 30 hops max, 38 byte packets
 1  x.x.x.x (x.x.x.x)  0.931 ms  0.840 ms  0.820 ms
 2  y.y.y.y (y.y.y.y)  1.661 ms  2.222 ms  1.795 ms


bb-hosts:
IPAdress1	MyRouter # testip conn conn=worst,IPAdress2,IPAdress3,IPAdress4,IPAdress5 trace NET:MyBBNET

Kind regards,

Sebastian