[Sipsak] sipsak does not return after third tmeout

[John A. Sullivan III]

Hello, all, and thanks to the developers for a very handy project.  We
are using sipsak 0.9.6 with OpenNMS 1.6.7 on CentOS 5.4 running kernel
2.6.31.5 configured with --enable-timeout=200 (by the way, the web page
instructions about using --enable-retrys=xxx appear to be in error - no
such parameter).

All works splendidly well unless there is a lost packet in the network.
Upon investigating, it looks like our script for our multi-homed OpenNMS
server is hanging:

#!/bin/bash
PATH=${PATH}:/usr/local/bin:
SOURCE="$(ip route get to ${2} | grep src | sed 's/.* src \([0-9][0-9]*[\.][0-9,\.]*\).*/\1/')"
sipsak -v -H ${SOURCE} -s sip:nobody@${2}
ret=$?
if [ ${ret} -gt 1 ];then
        exit 8
fi
exit 0

We see the following of there is no network connectivity (and hence no
icmp reply to indicate an error):

[root at monitor01 download]# sipsak -v -H 192.x.x.125 -s
sip:nobody at 172.y.y.9
** timeout after 200 ms**
** timeout after 400 ms**
** timeout after 800 ms**

or, with more verbosity:

[root at monitor01 scripts]# sipsak -vv -H 192.168.124.125 -s sip:nobody at 172.30.14.9
** timeout after 200 ms**
** timeout after 400 ms**

received ICMP message: ignoring (ICMP data is not a UDP packet)
** timeout after 800 ms**

received ICMP message: ignoring (ICMP data is not a UDP packet)

received ICMP message: ignoring (ICMP data is not a UDP packet)

received ICMP message: ignoring (ICMP data is not a UDP packet)

received ICMP message: ignoring (ICMP data is not a UDP packet)

received ICMP message: ignoring (ICMP data is not a UDP packet)

The command never returns no matter how long we wait.  Did we invoke the
command incorrectly? Is this a bug (we saw a similar one for traceroute
mode)? Is there a workaround? Thanks - John
-- 
John A. Sullivan III
Open Source Development Corporation
+1 207-985-7880
jsullivan at opensourcedevel.com

http://www.spiritualoutreach.com
Making Christianity intelligible to secular society