osdir.com

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Problem with CLOUDSTACK-10240 (Cannot migrate local volume to shared storage)


It is in 4.11.2.0.
It was first added to 4.12, and then back ported. I am now waiting for
Mike's feedback to proceed with the fix for the problem he found.

On Mon, Jul 16, 2018 at 4:36 PM, Yiping Zhang <yzhang@xxxxxxxxxxx> wrote:

>
> Is this code already in ACS 4.11.1.0?
>
> CLOUDSTACK-10240 is listed as fixed in 4.11.1.0, according to release note
> here, http://docs.cloudstack.apache.org/projects/cloudstack-
> release-notes/ja/master/fixed_issues.html, but in the JIRA ticket itself,
> the "fixed version/s" field says 4.12.
>
> We are using XenServer clusters with shared NFS storages and I am about to
> migrate to ACS 4.11.1.0 from 4.9.3.0.  Since we move VM between clusters a
> lot, this is going to be a blocker for us.  Someone please confirm.
>
> Thanks
>
> Yiping
>
>
> On 7/14/18, 11:20 PM, "Tutkowski, Mike" <Mike.Tutkowski@xxxxxxxxxx>
> wrote:
>
>     Hi,
>
>     While running managed-storage regression tests tonight, I noticed a
> problem that is not related to managed storage.
>
>     CLOUDSTACK-10240 is a ticket asking that we allow the migration of a
> virtual disk that’s on local storage to shared storage. In the process of
> enabling this feature, the VirtualMachineManagerImpl.
> getPoolListForVolumesForMigration method was re-written in a way that
> completely breaks at least one use case: Migrating a VM across compute
> clusters (at least supported in XenServer). If, say, a virtual disk resides
> on shared storage in the source compute cluster, we must be able to copy
> this virtual disk to shared storage in the destination compute cluster.
>
>     As the code is currently written, this is no longer possible. It also
> seems that the managed-storage logic has been dropped for some reason in
> the new implementation.
>
>     Rafael – It seems that you worked on this feature. Would you be able
> to look into this and create a PR?
>
>     Thanks,
>     Mike
>
>
>


-- 
Rafael Weingärtner