Ciao,

sorry ... I had time only on these days to test your xml.
what I've done

- put your code in a file => stonith.xml

- cibadmin -C -o resources --xml-file stonith.xml

the above command wrote the cib.xml but the "fencedev" resource didn't  
start.
In attachment you will find the output of the hb_report command.

Thanks so much for any help

cristina



On Sep 11, 2009, at 7:07 PM, Dejan Muhamedagic wrote:

> Ciao,
>
> On Fri, Sep 11, 2009 at 12:31:30PM +0200, Cristina Bulfon wrote:
>> Ciao,
>>
>> I will open a bugzilla.. I should be expert now :-)
>
> :)
>
>> and yes the device name is the same .. do you have some xml example  
>> to
>> show me ?
>
> Sure:
>
>  <clone id="Fence-clone">
>    <instance_attributes>
>      <attributes>
>        <nvpair name="clone_max" value="2"/>
>        <nvpair name="clone_node_max" value="1"/>
>      </attributes>
>    </instance_attributes>
>    <primitive id="fencedev" class="stonith" type="external/sbd">
>      <operations>
>        <op name="monitor" interval="120s" timeout="60s"  
> prereq="nothing"/>
>        <op name="start" timeout="60s" prereq="nothing"/>
>      </operations>
>      <instance_attributes>
>        <attributes>
>          <nvpair name="sbd_device" value="YOUR disk device"/>
>        </attributes>
>      </instance_attributes>
>    </primitive>
>  </clone>
>
> Basically, all stonith configurations are the same. Just change
> the instance attributes and timeouts. This example may or may not
> work, perhaps there are some ids missing, didn't try XML for such
> a long time :)
>
> Thanks,
>
> Dejan
>
>> thanks
>>
>> cristina
>> On Sep 11, 2009, at 12:25 PM, Dejan Muhamedagic wrote:
>>
>>> Ciao,
>>>
>>> On Fri, Sep 11, 2009 at 12:04:23PM +0200, Cristina Bulfon wrote:
>>>> Ciao,
>>>>
>>>> I've configured a SBD device for checking the Shared disk trough  
>>>> the
>>>> SAN.
>>>> My 2 node cluster have installed a Dual Path Qlogic FC cards with
>>>> failover, if i try to unplug
>>>> one FC port on my active machine .. SBD does the reboot to the
>>>> machine
>>>> w/o
>>>> waiting for the failover. Reading the http://www.linux-ha.org/SBD_Fencing
>>>>
>>>> I can manage the timeout parameters but it doesn't work on my
>>>> machines, I get
>>>> invalid option
>>>>
>>>> sbd -d /dev/AFS/sdg -4 20 -1 10 create
>>>>
>>>> sbd: invalid option -- 4
>>>
>>> Please file a bugzilla for this.
>>>
>>>> I am using heartbeat 2.1.4 and I cannot move up since my python
>>>> version is 2.3.4
>>>>
>>>> For testing SBD with HA I've created a simple resource and it has
>>>> been
>>>> worked,
>>>> but I saw that most of the people create a clone. Which is the best
>>>> way to use SBD  ?
>>>
>>> I think clones should be OK, of course if you use the same device
>>> name on all nodes.
>>>
>>> Thanks,
>>>
>>> Dejan
>>>
>>>> Any hint for the xml configuration to use SBD with 2 node cluster ?
>>>>
>>>> Thanks
>>>>
>>>> PS: next week I will be read email time to time,  sorry in  
>>>> advance if
>>>> I cannot answer soon.
>>>>
>>>> cristina
>>>> _______________________________________________
>>>> Linux-HA mailing list
>>>> [email protected]
>>>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>>>> See also: http://linux-ha.org/ReportingProblems
>>> _______________________________________________
>>> Linux-HA mailing list
>>> [email protected]
>>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>>> See also: http://linux-ha.org/ReportingProblems
>>>
>>
>> _______________________________________________
>> Linux-HA mailing list
>> [email protected]
>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>> See also: http://linux-ha.org/ReportingProblems
> _______________________________________________
> Linux-HA mailing list
> [email protected]
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
>

_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

Reply via email to