[Pacemaker] Only Xen primitives not in groups migrate on failover

[Andrew Beekhof]

On Tue, Apr 27, 2010 at 9:24 PM, Roberto Giordani <r.giordani at libero.it> wrote:
> Hi Andrew,
> probably ther is my mistake.
> I'd like to add a second file system mounted on the guest xen, this file
> system is not the main files sytem where run the O.S., but is a file system
> that use the xen machine for file sharing for example.

Not relevant.

Think about it, the guest is accessing a file on the second FS and all
of a sudden the block device disappears.
Thats what would happen during a migration.

> Is it not possible?

No groups. End of story.

>
> Regards,
> Roberto
> On 04/27/2010 09:09 PM, Andrew Beekhof wrote:
>>
>> On Tue, Apr 27, 2010 at 8:54 PM, Roberto Giordani<r.giordani at libero.it>
>>  wrote:
>>
>>>
>>> Hello,
>>> this mean that if i create a group with a file system + xen the live
>>> migration doen't work?
>>>
>>
>> Correct.
>> You'd need to use a cluster filesystem (or some other arrangement that
>> could be cloned sanely) and colocate the xen instance with that.
>>
>>
>>>
>>> Regards,
>>> Roberto.
>>> On 04/27/2010 08:41 PM, Andrew Beekhof wrote:
>>>
>>>>
>>>> On Tue, Apr 27, 2010 at 6:39 PM, Michael Brown<michael at netdirect.ca>
>>>>  wrote:
>>>>
>>>>
>>>>>
>>>>> Greetings,
>>>>>
>>>>> I've discovered that if I have ocf::hb:Xen resources as primitives,
>>>>> they
>>>>> migrate to other nodes on failover.
>>>>>
>>>>> But if they are members of a group, the domUs stop/start. Am I doing it
>>>>> wrong?
>>>>>
>>>>>
>>>>
>>>> No.  The semantics of groups prohibit live migration.
>>>>
>>>>
>>>>
>>>>>
>>>>> I'd like to use the groups to minimize the amount of config I
>>>>> need for each VM.
>>>>>
>>>>> Here's my current config: http://pastebin.com/LSqWvtEx
>>>>>
>>>>> guest1 stops/starts when triggered
>>>>> guest3 migrates when triggered
>>>>>
>>>>> Thanks,
>>>>>
>>>>> Michael
>>>>>
>>>>> --
>>>>> Michael Brown               | `One of the main causes of the fall of
>>>>> Systems Consultant          | the Roman Empire was that, lacking zero,
>>>>> Net Direct Inc.             | they had no way to indicate successful
>>>>> ☎: +1 519 883 1172 x5106    | termination of their C programs.' - Firth
>>>>>
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> Pacemaker mailing list: Pacemaker at oss.clusterlabs.org
>>>>> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
>>>>>
>>>>> Project Home: http://www.clusterlabs.org
>>>>> Getting started:
>>>>> http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
>>>>>
>>>>>
>>>>
>>>> _______________________________________________
>>>> Pacemaker mailing list: Pacemaker at oss.clusterlabs.org
>>>> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
>>>>
>>>> Project Home: http://www.clusterlabs.org
>>>> Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
>>>>
>>>
>>>
>>> _______________________________________________
>>> Pacemaker mailing list: Pacemaker at oss.clusterlabs.org
>>> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
>>>
>>> Project Home: http://www.clusterlabs.org
>>> Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
>>>
>>
>> _______________________________________________
>> Pacemaker mailing list: Pacemaker at oss.clusterlabs.org
>> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
>>
>> Project Home: http://www.clusterlabs.org
>> Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
>
>
>
> _______________________________________________
> Pacemaker mailing list: Pacemaker at oss.clusterlabs.org
> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
>
> Project Home: http://www.clusterlabs.org
> Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf