[one-users] Problem with shared_lvm driver

[Alan Jur?i?]

Citiram Carlos Martín Sánchez <cmartin at opennebula.org>:
> Hi,
>
> When the VM is running, onevm delete and onevm cancel perform the same
> driver operations, but with cancel OpenNebula waits for the VMM driver
> success before deleting the disks.
>
> onevm delete is more like a last resort to remove the VM from OpenNebula in
> case of error. The VM is removed immediately and OpenNebula tries to clean
> up, but the right way to destroy a VM is onevm cancel.
>
> In 4.0 we have synchronized the onevm delete driver operations, but the VM
> is still removed immediately and if the cleanup fails it is not checked.
>

Thanks, that pretty much solves my problem.

I'm having one other issue with shared_lvm driver though. When trying  
to add new image to shared_lvm datastore (source is raw image) it  
seems that datastore/shared_lvm/stat script is not working properly. I  
had to call fs_du function in cp scirpt to get size calculated for  
lvcreate. The problem is that 0M size is still written to database and  
I have to fix it manually to get things working (saveas for example).

Not sure if this is OpenNebula 3.4 issue or driver problem. Any help  
regarding this would be much appreciated.

Regards,
Alan