[lvs-users] Multicast message not send out from LVS so connection is not sync

Yau (Kevin) Leung kvleung at gmail.com
Wed Jul 8 02:36:13 BST 2009

Dear all,

I'm running LVS-DR mode on 2 X LVS.  The OS is Centos 5.3, 32bit
with piranha-0.8.4-11.el5.  In lvs.cf, the syncdaemon = 1.  But connections
are not sync and I found that the multicast messages are not sent out from
In /var/log/messages, I clearly

MASTER's log:
Jul  7 22:22:53 lvs1 kernel: IPVS: sync thread started: state = MASTER,
mcast_ifn = eth0, syncid = 0

BACKUP's log:
Jul  7 22:23:22 lvs2 kernel: IPVS: sync thread started: state = BACKUP,
mcast_ifn = eth0, syncid = 0

in eth0, eth0:1, eth0:2 and eth1 of both LVS,  all are showing " UP

In the MASTER and BACKUP, I can't see any multicast message with the address, but I can see other multicast messages.
[root at lvs1 ~]# tcpdump multicast -nn
tcpdump: verbose output suppressed, use -v or -vv for full protocol decode
listening on eth0, link-type EN10MB (Ethernet), capture size 96 bytes
09:32:38.977443 IP x.x.x.x.1985 > HSRPv0-hello 20:
state=active group=121 addr=x.x.x.x
09:32:39.865486 802.1d unknown version

I have stopped iptables during testing so it shouldn't not be the firewall
blocking the messages.  Is there any clue why the multicast message is not
sent out?  Thanks a million.


More information about the lvs-users mailing list