17:02:05 <tjones> #startmeeting vmwareapi
17:02:06 <openstack> Meeting started Wed Oct 22 17:02:05 2014 UTC and is due to finish in 60 minutes.  The chair is tjones. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:02:07 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
17:02:09 <openstack> The meeting name has been set to 'vmwareapi'
17:02:14 <tjones> hi sorry im late
17:02:21 <thangp> o/
17:02:21 <tjones> who's here today?
17:02:24 <arnaud__> o/
17:03:13 <tjones> anyone else??
17:03:52 <arnaud__> short meeting apparently :)
17:03:54 <thangp> lite crew
17:04:17 <tjones> yeah
17:04:22 <tjones> ok lets be quick then
17:04:30 <thangp> are there sessions planned for paris?
17:04:32 <tjones> #link https://review.openstack.org/#/q/status:open+project:openstack/nova-specs+message:vmware,n,z
17:04:57 <tjones> i wanted to talk about the status of these bp, but the owners are not here...
17:04:59 <thangp> please add https://review.openstack.org/#/c/129389/
17:05:11 <arnaud__> tjones, the OVA bp has been approved
17:05:19 <arnaud__> I will rebase the patch today
17:05:20 <thangp> it's about supporting adapter_type as an extra-spec in cinder
17:05:23 <tjones> cool
17:05:35 <tjones> thangp: this is a cinder spec?
17:05:40 <thangp> yes
17:05:52 <tjones> does it have a nova component too??
17:05:59 <thangp> kind of related to my volume hotplug bug
17:06:07 <thangp> nova component?
17:06:19 <tjones> i mean do you need to do anything in nova to support this?
17:06:27 <tjones> or just cinder change?
17:06:36 <thangp> yes, my volume hotplug bug to make it in:
17:06:44 <thangp> https://review.openstack.org/#/c/122251/
17:07:09 <tjones> i see ok
17:07:34 <thangp> because the code will use the adapter_type of the shadow volume to determine what controller to add to the vm
17:08:15 <thangp> since the volume will always be lsilogic, it will be the only one added to the vm
17:08:15 <tjones> ok i have added it here https://etherpad.openstack.org/p/VMwareapi-kilo
17:08:29 <thangp> i like to add other adapter types
17:08:34 <thangp> thx
17:08:43 <tjones> great
17:08:57 <tjones> any other updates on specs for kilo
17:08:58 <tjones> ?
17:09:16 <tjones> rado is not here unfortunately
17:09:25 <tjones> i'll ping him
17:09:26 <thangp> not from me
17:09:34 <thangp> is gary here?
17:09:41 <tjones> no he is not
17:09:56 <thangp> i had a question why  it's necessary for the boot disk to be on the same datastore
17:10:01 <thangp> maybe you would know
17:10:19 <tjones> or arnaud__ will - what's your question?
17:10:31 <tjones> i mean - more detail about the question
17:10:40 <thangp> for attach_root_volume ... why it's necessary for the boot disk to be on the same datastore, or if it is still necessary
17:10:54 <thangp> it's related to this bug: https://review.openstack.org/#/c/128508/8/nova/virt/vmwareapi/vmops.py
17:11:00 <thangp> mbooth asked
17:11:19 <tjones> looking
17:11:48 <tjones> sorry i do not know
17:12:05 <tjones> best to ping gary
17:12:07 <thangp> np :)
17:12:12 <thangp> i'll ask gary
17:12:15 <tjones> great
17:12:36 <tjones> im happy to end early if you guys don't have anything else - all i wanted to discuss was BP for kilo today
17:12:48 <arnaud__> thangp, what do you mean by "same" datastore?
17:13:05 <arnaud__> the _relocate_vmdk_volume takes a datastore in param
17:13:55 <thangp> the code in attach_root_volume...https://github.com/openstack/nova/blob/master/nova/virt/vmwareapi/volumeops.py#L560
17:14:21 <thangp> it calls attach_volume in the end, but if its a vmdk, it does a _relocate_vmdk_volume
17:14:23 <arnaud__> yep
17:14:30 <thangp> i'm not sure why it is needed
17:15:00 <thangp> it "Pick the resource pool on which the instance resides. Move the volume to the datastore of the instance."
17:15:13 <arnaud__> yeah
17:15:27 <thangp> is it necessary?  mbooth asked, and i didnt know
17:16:21 <thangp> so can we just call attach_volume instead of attach_root_volume?  for any volumes even the boot volume
17:16:56 <arnaud__> I might missing something
17:17:18 <arnaud__> but if you have a vmdk vol, and you want to you use, you need to put it in the datastore where the vm folder has been created
17:17:20 <arnaud__> no?
17:17:36 <arnaud__> so, you have to relocate
17:18:25 <thangp> can a vmdk volume from another datastore be attached to a vm in a different datastore?
17:19:03 <arnaud__> I don't think so, if it's supposed to be the root disk
17:19:10 <thangp> arnaud__, i understand the logic, but i think mbooth is asking ^^
17:19:29 <arnaud__> let me confirm with vui
17:19:34 <arnaud__> if garyk doesn't know
17:19:36 <thangp> ok, i'll put that in the note
17:19:38 <arnaud__> k?
17:19:41 <thangp> ok, thx
17:20:18 <tjones> ok guys - anything else?  looks like no one else has joined
17:20:28 <thangp> good from here
17:20:42 <tjones> lets end early then
17:20:45 <tjones> thankd
17:20:47 <tjones> thanks
17:20:56 <arnaud__> bye
17:20:57 <tjones> #endmeeting