Open Stack

On 2/9/2015 10:15 AM, Andreas Maier wrote:
>
> Hello,
> I would like to ask for the following feature freeze exceptions in Nova.
>
> The patch sets below are all part of this blueprint:
> https://review.openstack.org/#/q/status:open+project:openstack/nova
> +branch:master+topic:bp/libvirt-kvm-systemz,n,z
> and affect only the kvm/libvirt driver of Nova.
>
> The decision for merging these patch sets by exception can be made one by
> one; they are independent of each other.
>
> 1. https://review.openstack.org/149242 - FCP support
>
>     Title: libvirt: Adjust Nova to support FCP on System z systems
>
>     What it does: This patch set enables FCP support for KVM on System z.
>
>     Impact if we don't get this: FCP attached storage does not work for KVM
>     on System z.
>
>     Why we need it: We really depend on this particular patch set, because
>     FCP is our most important storage attachment.
>
>     Additional notes: The code in the libvirt driver that is updated by this
>     patch set is consistent with corresponding code in the Cinder driver,
>     and it has seen review by the Cinder team.
>
> 2. https://review.openstack.org/150505 - Console support
>
>     Title: libvirt: Enable serial_console feature for system z
>
>     What it does: This patch set enables the backing support in Nova for the
>     interactive console in Horizon.
>
>     Impact if we don't get this: Console in Horizon does not work. The
>     mitigation for a user would be to use the Log in Horizon (i.e. with
>     serial_console disabled), or the "virsh console" command in an ssh
>     session to the host Linux.
>
>     Why we need it: We'd like to have console support. Also, because the
>     Nova support for the Log in Horizon has been merged in an earlier patch
>     set as part of this blueprint, this remaining patch set makes the
>     console/log support consistent for KVM on System z Linux.
>
> 3. https://review.openstack.org/150497 - ISO/CDROM support
>
>     Title: libvirt: Set SCSI as the default cdrom bus on System z
>
>     What it does: This patch set enables that cdrom drives can be attached
>     to an instance on KVM on System z. This is needed for example for
>     cloud-init config files, but also for simply attaching ISO images to
>     instances. The technical reason for this change is that the "IDE"
>     attachment is not available on System z, and we need "SCSI" (just like
>     Power Linux).
>
>     Impact if we don't get this:
>        - Cloud-init config files cannot be on a cdrom drive. A mitigation
>           for a user would be to have such config files on a cloud-init
>           server.
>        - ISO images cannot be attached to instances. There is no mitigation.
>
>     Why we need it: We would like to avoid having to restrict cloud-init
>     configuration to just using cloud-init servers. We would like to be able
>     to support ISO images.
>
>     Additional notes: This patch is a one line change (it simply extends
>     what is already done in a platform specific case for the Power platform,
>     to be also used for System z).
>
> Andy
>
> Andreas Maier
> IBM Senior Technical Staff Member, Systems Management Architecture & Design
> IBM Research & Development Laboratory Boeblingen, Germany
> maiera at de.ibm.com, +49-7031-16-3654
> ________________________________________________________________________
> IBM Deutschland Research & Development GmbH
> Vorsitzende des Aufsichtsrats: Martina Koederitz
> Geschaeftsfuehrung: Dirk Wittkopp
> Sitz der Gesellschaft: Boeblingen
> Registergericht: Amtsgericht Stuttgart, HRB 243294
>
>
> __________________________________________________________________________
> OpenStack Development Mailing List (not for usage questions)
> Unsubscribe: OpenStack-dev-request at lists.openstack.org?subject:unsubscribe
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
>

FWIW, I'll sponsor these changes.  I'm already +2 on one and have 
reviewed the other two which are very close to a +2 from me, just need a 
little more work (but not drastic changes).

-- 

Thanks,

Matt Riedemann