|
"td" <td@xxxxxxxxxxxx> wrote:
> Hi all,
> i have installed "heartbeat" on an 2 node cluster(RedHat 7.0). The
Hi,
this is the list for LinuxVirtualServer, a loadbalancer based on Linux.
You might get more help at the heartbeat mailing list linux-ha@xxxxxxx
Send a message to linux-ha-subscribe@xxxxxx to subscribe to the list.
> heartbeate-link is a 10MBit cross-networkcable (eth1<->eth1).
> The problem is when node1(master) goes down, node2 takes over the ip. But
> when node1 is back online node2 doesn´t recognize that and i have to take
> down eth0:1 manually. Is there a way to do it automatically?
It should do it automatically. How does your haresources file look like?
And don't configure the alias address within you OS! Heartbeat cares
completely about the alias.
> Another problem is that "heartbeate" only works when the complete machine
> goes down or the heartbeat link, if only the nic(eth0 of node1) fails it
> doesn´t work.
No wonder. Heartbeat isn't designed to monitor the resources! It only
monitors the other instance of heartbeat (on the other node). For resource
monitoring you need something like mon (http://www.kernel.org/software/mon/).
> Here is my /etc/ha.d/ha.cf
> keepalive 1
> deadtime 3
You shouldn't use such a little value for the deadtime, or you might run into
a little problem in heartbeat on startup (I'm not sure if this is fixed in
the latest release).
> udpport 1001
> udp eth1
> debugfile /var/log/ha-debug
> logfile /var/log/ha-log
> logfacility local0
> node node01.xxx.com
> node node02.xxx.com
The other lines are ok.
Juri
--
Juri Haberland <juri@xxxxxxxxxxxxxx>
|
