Description of problem:

I use virt-manager on Fedora 20 to connect to the hypervisor on my RHEL 6.5 nodes. When I stop the libvirtd daemon on the nodes (this is done by the cluster when withdrawing a node from the cluster), the client throws the usual error about losing the connection.

Now, however, it keeps spawning the error, over and over again. Even if I just leave the error up, new errors spawn every few seconds. I had to kill virt-manager PID to get them to stop.


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

Client:

lemass:/home/digimer# uname -a
Linux lemass.alteeve.ca 3.13.6-200.fc20.x86_64 #1 SMP Fri Mar 7 17:02:28 UTC 2014 x86_64 x86_64 x86_64 GNU/Linux
lemass:/home/digimer# rpm -q virt-manager
virt-manager-1.0.0-5.fc20.noarch

Node:

an-c05n01:~# uname -a 
Linux an-c05n01.alteeve.ca 2.6.32-431.5.1.el6.x86_64 #1 SMP Fri Jan 10 14:46:43 EST 2014 x86_64 x86_64 x86_64 GNU/Linux
an-c05n01:~# rpm -q libvirt
libvirt-0.10.2-29.el6_5.5.x86_64


How reproducible:

100%


Steps to Reproduce:
1. Connect virt-manager from Fedora to libvirtd on rhel 6.5
2. Stop libvirtd on rhel 6.5
3.

Actual results:

Frighteningly annoying error messages that eventually consume the system.


Expected results:

One error message only.

Additional info:

Screen shot attached. Note that 'jounalctl -f' on the client and /var/log/messages on the server showed nothing.

I am marking this as "urgent" because users not familiar with pgrep/kill will have trouble stopping the errors and would likely have to reboot to regain control of their system.