Description of problem:
Mutlicast between virtual hosts gets dropped after a short period
Version-Release number of selected component (if applicable):
Instal omping on both virtual hosts. A low number of ping wil turn out OK, but giving a higher number, the multicast packet loss gets huge:
omping <remote ip> local ip> -c500 (on both hosts)
192.168.1.211 : unicast, xmt/rcv/%loss = 500/500/0%, min/avg/max/std-dev = 0.330/0.610/0.789/0.064
192.168.1.211 : multicast, xmt/rcv/%loss = 500/268/46%, min/avg/max/std-dev = 0.416/0.635/0.921/0.066
After turning off "multicast snooping", the packet loss will disappear:
echo 0 > /sys/class/net/ovirtmgmt/bridge/multicast_snooping
omping <remote ip> local ip> -c500
192.168.1.212 : unicast, xmt/rcv/%loss = 500/500/0%, min/avg/max/std-dev = 0.353/0.623/0.944/0.075
192.168.1.212 : multicast, xmt/rcv/%loss = 500/500/0%, min/avg/max/std-dev = 0.356/0.642/0.964/0.078
Steps to Reproduce:
1. install virt-node-iso-2.6.1-20120228
2. install 2 virtual (fedora) host and omping
3. start multicast test between virtual host using omping
Huge packet loss after apr. 200 ping.
0% packet loss
I've changed the component as vdsm is responsible to setting up ovirt's bridges. There's a twin bug 902922 for libvirt.
This bug is waiting for the patch from https://bugzilla.redhat.com/show_bug.cgi?id=880035#c26 to get in next release.
Closed because problem in 880035 was succesfully solved.