[Date Prev][Date Next] [Thread Prev][Thread Next]
[Thread Index]
[Date Index]
[Author Index]
Re: Question about clumanager
- From: Chris Kloiber <ckloiber redhat com>
- To: "Discussion of Red Hat Enterprise Linux 3 (Taroon)" <taroon-list redhat com>
- Subject: Re: Question about clumanager
- Date: Sun, 05 Sep 2004 11:04:21 -0400
On Sat, 2004-09-04 at 18:35, Carlos Lopez wrote:
> Hi,
>
> I have installed a test environment with two RHAS 3.0
> hosts. I would like to test clumanager with Oracle 9i.
>
> Both hosts has two network interfaces: eth0
> (external to corporate network, and eth1 internal for
> heartbeat).
>
> node01:
>
> eth0: 192.168.1.30/24
> eth1: 10.0.1.1
>
> node02:
>
> eth0: 192.168.1.31/24
> eth1: 10.0.1.2
>
>
> When I try to config both hosts with
> redhat-config-cluster gives me an error under
> cluquorumd config. Under tiebreaker IP section, what
> IP I have to put if I would like to use heartbeart
> network (no default route exists in this network)??
>
> Until now, only works to me activating the ping
> interval option (The worse option if timeout of disc
> takes place, according to redhat documentation).
>
> Thank you very for your help.
It's my understanding that a dedicated heartbeat channel is no longer
required. Heartbeats travel over all configured interfaces. In fact, you
can channel bond both interfaces on the public side for redundancy and
the heartbeat will travel over the public interface. I have discovered
that it's a good idea if you have multiple clusters on the same network
to change the cluster name (default: "Red Hat Clumanager Cluster") and
also change the Multicast IP address under "Cluster Daemon
Properties"/"clumembd" tab of the redhat-config-cluster gui. You might
need to use the cluforce command to make the first node of a two node
cluster bring up the quorum and start the service.
--
Chris Kloiber, RHCX
Global Support Services
Red Hat, Inc.
[Date Prev][Date Next] [Thread Prev][Thread Next]
[Thread Index]
[Date Index]
[Author Index]