[one-users] suspend vs stop

[Shi Jin]

Interesting.
I didn't do anything else but simply stop and resume and it went to pending.
However, after removing a few other instances, the pending VM did go running.

Shi
On Wed, Sep 2, 2009 at 3:49 AM, Javier Fontan<jfontan at fdi.ucm.es> wrote:
> Hello,
>
> Resume after stop lets the scheduler select a physical host with
> enough resources to resume the VM. It should have worked, if it
> continued in pending state it is possibly because there was no free
> physical host to run it.
>
> Does it make sense to you?
>
> Bye
>
> On Wed, Sep 2, 2009 at 5:41 AM, Shi Jin<jinzishuai at yahoo.com> wrote:
>> Sorry, I didn't read the documentation very carefully.
>> Now I found this about the status:
>> stop (virtual machine state is transferred back to OpenNebula for a possible reschedule)
>> suspend (virtual machine state is left in the cluster node for resuming)
>>
>> So I understand the difference. For the stopped one, I do find the image file in the var directory (I am using nfs so it simply didn't delete the images directory after onevm delete"). But I am still confused on how to use to "onevm stop".
>> Is it only useful in the following sequence?
>> * use the VM and change its contents.
>> * shutdown the VM in the OS
>> * onevm stop
>> * onevm delete
>> * The reuse the image file
>> This way I get a updated image file for the VM.
>>
>> Other than the above scenario, I cannot think of any use of it. If that is the case, why wouldn't "stop" combine "delete" into it?
>> I am sure I am missing something important.
>
> --
> Javier Fontan, Grid & Virtualization Technology Engineer/Researcher
> DSA Research Group: http://dsa-research.org
> Globus GridWay Metascheduler: http://www.GridWay.org
> OpenNebula Virtual Infrastructure Engine: http://www.OpenNebula.org
>



-- 
Shi Jin, Ph.D.