andreas at hastexo
Jan 12, 2012, 4:59 PM
Post #4 of 11
(678 views)
Permalink
|
|
Re: Resended : Understanding how heartbeat and pacemaker work together
[In reply to]
|
|
Hello,
On 01/12/2012 11:29 PM, Niclas Müller wrote:
> That the VirtualIP isn't shown by 'ifconfig -a' is realy nice, because i
> made my failed search on this because of this howto :
>
> http://www.howtoforge.com/high_availability_loadbalanced_apache_cluster
You follow a howto from the year 2006? ... anyway, ifconfig would show
the IP because you used IPaddr and not IPaddr2 RA ....
>
>
>
>
> Im going to understand now all. I've configurated a resource for
> failover-ip from your link, but get this error page. Pacemaker cannot
> start resource failover-ip
of course you already have an interface up in this network? The resource
agent only adds secondary addresses ....
And you should find most logging information on your DC - node2 in
/var/log/{syslog,daemon.log)
Regards,
Andreas
--
Need help with Pacemaker?
http://www.hastexo.com/now
>
>
>
>
>
> root [at] node:/etc/ha.d# crm_mon -1fr
> ============
> Last updated: Thu Jan 12 17:25:43 2012
> Stack: Heartbeat
> Current DC: node2 (40dbaed1-9618-41f0-acbe-c3f0f6334cce) - partition
> with quorum
> Version: 1.0.9-74392a28b7f31d7ddc86689598bd23114f58978b
> 2 Nodes configured, unknown expected votes
> 1 Resources configured.
> ============
>
> Online: [ node1 node2 ]
>
> Full list of resources:
>
> failover-ip (ocf::heartbeat:IPaddr): Stopped
>
> Migration summary:
> * Node node1:
> failover-ip: migration-threshold=1000000 fail-count=1000000
> * Node node2:
> failover-ip: migration-threshold=1000000 fail-count=1000000
>
> Failed actions:
> failover-ip_start_0 (node=node1, call=4, rc=1, status=complete):
> unknown error
> failover-ip_start_0 (node=node2, call=4, rc=1, status=complete):
> unknown error
>
>
>
>
>
> node $id="40dbaed1-9618-41f0-acbe-c3f0f6334cce" node2
> node $id="6c62b04f-0d3a-4bc5-a084-ffba618a8e87" node1
> primitive failover-ip ocf:heartbeat:IPaddr \
> params ip="192.168.0.100" \
> op monitor interval="10s" \
> meta target-role="Started"
> property $id="cib-bootstrap-options" \
> dc-version="1.0.9-74392a28b7f31d7ddc86689598bd23114f58978b" \
> cluster-infrastructure="Heartbeat" \
> stonith-enabled="false"
>
>
>
>
>
>
> Thank'a lot, the switch in my brain make PUK and now it seems all clear
> how this work together. Syslog doesn't show any errors about the
> failover-ip resource. Ideas ?
>
>
>
>
>
> On 01/12/2012 10:55 PM, Andreas Kurz wrote:
>> On 01/12/2012 10:22 PM, Niclas Müller wrote:
>>
>>> I'm currently going to setup a Linux HA Cluster with apache and MySQL.
>>> I've created three VM with KVM Vitalization. One NetworkManager as DNS
>>> and DHCP Server, and two other as Cluster Nodes. All VMs are both Debian
>>> Squeeze minimal installations. On the Nodes i've installed the packages
>>> heartbeat and pacemaker. The configuration of heartbeat seems like
>>> correct because in the syslog there are no errors and i can read that
>>> the nodes have contact. My first impression of the software packages is,
>>> that heartbeat is only for checking the Nodes of ability. Pacemaker is
>>> for the services which are to manager. I cannot get it on that heartbeat
>>> / pacemaker (??) use the virtual ip. In both interfaces there is no
>>> eth0:0 with the configurated ip address. Is the virtual ip used by the
>>> primary node only when a service is confiugrated? Have anybody a good
>>> howto for set up a apache and mysql cluster with heartbeat and pacemaker
>>> ?
>>>
>> yes, Heartbeat is the CCM (cluster consensus and membership) layer and
>> Pacemaker relies on it to get valid information about node health and
>> uses it to transfer messages/updates to the other nodes.
>>
>> use "crm_mon -1fr" to see the current state of your cluster resources
>> you already?? configured in Pacemaker.
>>
>> if you use IPaddr2 RA you have to use "ip addr show" to see the virtual IPs.
>>
>> there are a lot of howtos available, maybe you want to start at:
>>
>> http://www.clusterlabs.org/wiki/Documentation#Examples
>>
>>
>> Regards,
>> Andreas
>>
>>
>>
>>
>> _______________________________________________
>> Linux-HA mailing list
>> Linux-HA [at] lists
>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>> See also: http://linux-ha.org/ReportingProblems
>
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
|
signature.asc
