Description of problem:
If up2date cannot reach the host, the 'up2date' command cannot be ctrl-c'd and
it takes many minutes to come back with a response:

[root@test01-ia64 tmp]# up2date
Error communicating with server. The message was:
Connection timed out


Version-Release number of selected component (if applicable):
up2date-4.4.63-3

How reproducible:
always

Steps to Reproduce:
1- ks a client to rhel 3 u6, include @ Everything in ks profile
2- ssh -Y <client>
3- on client:  install latest up2date from
/nfs/redhat/beehive/comps/dist/3.0E-qu-candidate/
4- register the client to webqa
5- on client:  run block-host.sh script to block firstboot's connections to
webqa:   ./block-host.sh xmlrpc.rhn.webqa.redhat.com  (NOTE:  block-host.sh
needs to be loaded on the client from wherever it's stored -- maybe
/usr/local/bin locally?)  This will allow pings to the host but will prevent
udp/tcp connections.
6- run 'update' on client

  
Actual results:
'up2date' sits for many minutes.  It cannot be killed by ctrl-c.  You have to
open a new session to kill it, and even then it requires kill -9.  If you don't
kill it, it will eventually return a 'Connection timed out message.'

Expected results:
Either make up2date killable via ctrl-c or reduce the time-out period.

Additional info: