Thanks you Lars for your answer,
I'm reading about Pacemaker, but, for now, I have another issue:
The comand "crm_mon" show the following info:
============
Last updated: Tue Jan 26 17:25:47 2010
Current DC: nodo2.example.com
2 Nodes configured.
3 Resources configured.
============
Node: nodo2.example.com : online
Node: nodo1.example.com : online
webip (ocf::heartbeat:IPaddr): Started nodo1.example.com
Master/Slave Set: ms_drbd
drbd0:0 (ocf::heartbeat:drbd):Master nodo1.example.com FAILED
drbd0:1 (ocf::heartbeat:drbd):Slave nodo2.example.com
Resource Group: group_fs
FS_replica (ocf::heartbeat:Filesystem): Started nodo1.example.com
(unman
aged) FAILED
Failed actions:
FS_replica_stop_0 (node=nodo1.example.com, call=170, rc=-2): Timed Out
drbd0:0_monitor_59000 (node=nodo1.example.com, call=167, rc=-2): Timed
Out
I think that this issue is related with the high load average but, the
question is: How can I to fix this?
Thanks for all,
Carlos Chiriboga Calderon
On Tue, 26 Jan 2010 15:05:11 +0100, Lars Ellenberg
<[email protected]> wrote:
> On Mon, Jan 25, 2010 at 08:56:52PM -0500, Carlos Eduardo Chiriboga
Calderon
> wrote:
>>
>>
>> Hi all,
>>
>> I'm trying to configure a heartbeat cluster with 2 nodes and a drbd
>> filesystem (the cib.xml is attached to this message).
>>
>> The problem is that, when I try to move the resource from a node to
>> other, the filesystem resource try to mount before that the drbd
>> master_slave resource set the correct master node and the mount fail.
>>
>>
>> The heartbeat version that I'm using is 2.1.4 and the drbd version is
> 8.0.16
>
> in your constraints,
> you need to _promote_ drbd before starting the Filesystem.
>
> But, do yourself a BIG favour.
>
> Use Pacemaker.
>
> Using drbd 8.3.7 would be strongly recommended, too.
>
> If you are scared by OpenAIS/Corosync,
> you can still use heartbeat as the infrastructure layer.
>
> But the management layer (CRM) is called Pacemaker,
> since over two years now.
>
> So unless you want to miss out on > 2 years development...
> as additional benefit, you get to use the CRM shell.
>
> I hear you say
> Why Pacemaker? I thought I want to use Heartbeat?
> well, it is "the same thing". Only two years later.
> http://clusterlabs.org/wiki/FAQ#Organizational
>
> The Heartbeat / Linux-HA Home Page:
> http://www.linux-ha.org/wiki/Main_Page
>
> The Heartbeat User's Guide (work in progress):
> http://www.linux-ha.org/wiki/Documentation
>
> The DRBD User's Guide:
> http://www.drbd.org/docs/applications/
>
> See "Integrating DRBD with Pacemaker".
> Just realize that you get to choose the infrastructure stack, use either
> Heartbeat (3.0.2 release is "imminent") or OpenAIS/Corosync.
>
> The "Integrating DRBD with Heartbeat" chapter is
> meant for _legacy_ Heartbeat
> (i.e. 2.1.4, which you currently are trying
> to use. But, please, rather use Pacemaker!)
>
> Pacemaker User's Guide, HowTo and Examples:
> http://www.clusterlabs.org/wiki/Documentation
>
> The non-xml CRM shell for xml-o-phobic admins,
> aka Pacemaker (command line, or "curses") User Interface,
> including help, tab completion and more:
> http://www.clusterlabs.org/doc/crm_cli.html
>
> And if you are more the GUI type,
> there is also the DRBD-MC,
> which is the platform independend Pacemaker Cluster GUI
> and DRBD Management Console, integrated VNC viewer for
> Pacemaker controlled VMs.
>
> The various Wizards should get you up and running in no time.
> http://www.drbd.org/mc/management-console/
>
>
>
_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems
