Red Hat Bugzilla – Bug 691777
Bonded and vlan tagged network does not work in KVM guest [rhel-6.0.z]
Last modified: 2013-01-09 18:43:27 EST
This bug has been copied from bug #623199 and has been proposed to be backported to 6.0 z-stream (EUS).
Thanks Liang Zheng for his help. With the environment he provides, we successfully verify this bug. Test topo: dl580g7-01 dl585g7-02 (eth5.100:192.168.100.55) --eth0-|-eth5--eth5.100 (Active Slave:eth0) / | br0--bond0.100--bond0-- | | \ | (eth4.100:192.168.100.44) | --eth1-|-eth4--eth4.100 | --vnet0 (guest: 192.168.10.10) Reproduce on 2.6.32-71.el6.i686 [root@hp-dl585g7-02 ~]# arping 192.168.100.10 -I eth5.100 ARPING 192.168.100.10 from 192.168.100.55 eth5.100 Unicast reply from 192.168.100.10 [52:54:00:4D:D0:AD] 1.263ms Unicast reply from 192.168.100.10 [52:54:00:4D:D0:AD] 0.934ms guest)# ping 192.168.100.55 ( Fail ) On kernel 2.6.32-71.29.1.el6.x86_64: [root@hp-dl585g7-02 ~]# ping 192.168.100.10 -I eth5.100 PING 192.168.100.10 (192.168.100.10) from 192.168.100.55 eth5.100: 56(84) bytes of data. 64 bytes from 192.168.100.10: icmp_seq=1 ttl=64 time=1.49 ms 64 bytes from 192.168.100.10: icmp_seq=2 ttl=64 time=0.749 ms 64 bytes from 192.168.100.10: icmp_seq=3 ttl=64 time=0.792 ms on guest machine of hp-dl585g7-02.rhts.eng.nay.redhat.com: [root@dhcp-66-86-245 ~]# ping 192.168.100.55 -c4 PING 192.168.100.55 (192.168.100.55) 56(84) bytes of data. 64 bytes from 192.168.100.55: icmp_seq=1 ttl=64 time=0.871 ms 64 bytes from 192.168.100.55: icmp_seq=2 ttl=64 time=0.965 ms 64 bytes from 192.168.100.55: icmp_seq=3 ttl=64 time=0.777 ms 64 bytes from 192.168.100.55: icmp_seq=4 ttl=64 time=0.863 ms --- 192.168.100.55 ping statistics --- 4 packets transmitted, 4 received, 0% packet loss, time 3003ms rtt min/avg/max/mdev = 0.777/0.869/0.965/0.066 ms So the bug is fixed. Set Verified.
Technical note added. If any revisions are required, please edit the "Technical Notes" field accordingly. All revisions will be proofread by the Engineering Content Services team. New Contents: In certain network setups (specifically, using VLAN on certain NICs where packets are sent through the VLAN GRO rx path), sending packets from an active ethernet port to another inactive ethernet port could affect the network's bridge and cause the bridge to acquire a wrong bridge port. This resulted in all packets not being passed along in the network. With this update, the underlying source code has been modified to address this issue, and network traffic works as expected.
An advisory has been issued which should help the problem described in this bug report. This report is therefore being closed with a resolution of ERRATA. For more information on therefore solution and/or where to find the updated files, please follow the link below. You may reopen this bug report if the solution does not work for you. http://rhn.redhat.com/errata/RHSA-2011-0498.html