Hi,
after upgrading from SP1 to SP2 the ucast-heartbeat connection of a
two-node-cluster does not work any longer.
The cluster is a combination of xen and heartbeat. XEN and Xen-Network
are being startet before heartbeat starts up.
Componentes before:
2 x SLES 10 SP1:
Heartbeat 2.0.8, XEN 3.0.4 (from SLES-Distribution)
Components after upgrade:
2 x SLES 10 SP2:
Heartbeat 2.1.3, XEN 3.2.0 (from SLES-Distribution)
Due to the changes in xen-network-bridging the subsequent
heartbeat-configuration had to be changed.
With sp1 relevant configuration of heartbeat (ha.cf):
ucast eth0 IP1
ucast eth0 IP2
ucast eth1 IP3
ucast eth1 IP4
ucast eth5 IP5
ucast eth5 IP6
With sp2 relevant configuration of heartbeat (ha.cf):
ucast BRIDGE1 IP1
ucast BRIDGE1 IP2
ucast BRIDGE2 IP3
ucast BRIDGE2 IP4
ucast BRIDGE3 IP5
ucast BRIDGE3 IP6
Connection test:
LOCALHOST=`uname -n`
REMOTEHOST=`cl_status listnodes -n|grep -v $LOCALHOST`
for i in `cl_status listhblinks $REMOTEHOST 2>/dev/null |sort -u|awk
'{print $1}'`
do
echo -n Checking heartbeat-line to $REMOTEHOST $i...
cl_status hblinkstatus $REMOTEHOST $i -m
done
Now heartbeat says that only BRIDGE2 is up - whereas before eth0, eth1
and eth5 were up.
I believe this to be a heartbeat-bug. Is this known out there?
What is the workaround?
Kind regards, Nils Hildebrand
_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems
