Yes, it's a regression. The impact is that upgraders to 4.4 may have existing VMs that don't boot when rebooted, or templates that won't work, because the virtual hardware is changed and not what was expected.
On Thu, Jul 17, 2014 at 1:56 AM, Daan Hoogland <daan.hoogl...@gmail.com> wrote: > Marcus, Amogh, WHat is the work to do and the ETA? As I read your > mail, it seems we are not going to revert but I have no idea on what a > forward fix would be. Just adding mappings to KVMGuestOsMapper? > > @Marcus: is this regression? As in do I agree with you making this a > blocker. You seem to imply so and I want to read you write it;) > > On Thu, Jul 17, 2014 at 9:14 AM, Erik Weber <terbol...@gmail.com> wrote: > > Considering adding centos 7 and rhel 7 while at it. > > > > Erik > > 17. juli 2014 00:32 skrev "Marcus" <shadow...@gmail.com> følgende: > > > >> Yes, thats it. I have no idea how these mappings are created, but we > need > >> to ensure that both upgrades and new installs work > >> > >> > >> On Wed, Jul 16, 2014 at 3:25 PM, Amogh Vasekar < > amogh.vase...@citrix.com> > >> wrote: > >> > >> > Hi, > >> > > >> > It is probably due to missing mapping in guest_os_hypervisor table. I > had > >> > created default entries for KVM using : > >> > http://bit.ly/1rrML8n > >> > > >> > I can add the missing mapping to it. > >> > > >> > Thanks, > >> > Amogh > >> > > >> > On 7/16/14 2:08 PM, "Marcus" <shadow...@gmail.com> wrote: > >> > > >> > >I have another big issue to report with 4.4 in testing these > artifacts. > >> I > >> > >register a CentOS 6.5 template, and launch VMs for it on KVM, and the > >> > >disks/nics are not virtio. > >> > > > >> > >It looks like it is due to the following commit, which changes how > the > >> KVM > >> > >agent determines if a template is PV enabled. The > >> 'getPlatformEmulator()' > >> > >is empty. I have reopened the associated bug and commented on it. > >> > > > >> > > - DiskDef.diskBus diskBusType = > >> > >getGuestDiskModel(vmSpec.getOs()); > >> > >+ DiskDef.diskBus diskBusType = > >> > >getGuestDiskModel(vmSpec.getPlatformEmulator()); > >> > > > >> > >commit 02bd3d0671b0cde46f8aa7892f20aa0bb0d48d1c > >> > >Author: Amogh Vasekar <amogh.vase...@citrix.com> > >> > >Date: Wed May 7 15:16:55 2014 -0700 > >> > > > >> > > CLOUDSTACK-6358: As a part of supporting dynamic guest OS > defined by > >> > >user, removing the hard-coded dependencies. > >> > > This patch is for KVM > >> > > > >> > > 1. Local testing on KVM > >> > > 2. Successfully got up system VMs > >> > > 3. Successfully created a CentOS VM > >> > > 4. Snapshots are not supported for KVM > >> > > > >> > > Signed off by :- Nitin Mehta<nitin.me...@citrix.com> > >> > > > >> > > > >> > > > >> > >On Wed, Jul 16, 2014 at 2:57 PM, Alena Prokharchyk < > >> > >alena.prokharc...@citrix.com> wrote: > >> > > > >> > >> Daan, during 4.4 testing, I found the following regression bug: > >> > >> > >> > >> https://issues.apache.org/jira/browse/CLOUDSTACK-7118 > >> > >> > >> > >> > >> > >> "Unable to expunge vms in error state" > >> > >> > >> > >> There is a workaround for it - I¹ve mentioned it in the bug - yet > >> > >>ideally > >> > >> I would prefer the fix to be cherry-picked from 4.4-forward: > >> > >> > >> > >> commit 822b38761f1655ef3d5d3648985a197f5c2c1262 > >> > >> > >> > >> > >> > >> > >> > >> Thank you, > >> > >> Alena. > >> > >> > >> > >> On 7/16/14, 4:47 AM, "Daan Hoogland" <daan.hoogl...@gmail.com> > wrote: > >> > >> > >> > >> >On Thu, Jul 3, 2014 at 1:41 AM, Marcus <shadow...@gmail.com> > wrote: > >> > >> >> I tested this release with > >> > >> >> cherry-picking 2ec7359b4eb501b0d9e80ed87af7a54938e9d505 from > >> > >> >>4.4-forward. > >> > >> > > >> > >> > > >> > >> >I have cherry-picked and pushed. I will address the last remaining > >> > >> >issue by editting the pom.xml from build_asf.sh and roll out a new > >> RC. > >> > >> > > >> > >> >Hope all of you have continued testing and have now more issues to > >> > >> >block this release!!!! > >> > >> > > >> > >> >-- > >> > >> >Daan > >> > >> > >> > >> > >> > > >> > > >> > > > > -- > Daan >
