Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2018-01-29
16:11:45 openstackgerrit Balazs Gibizer proposed openstack/nova master: Add late server group policy check to rebuild https://review.openstack.org/525242
16:11:45 openstack bug 1735407 in OpenStack Compute (nova) "[Nova] Evacuation doesn't respect anti-affinity rules" [Medium,In progress] https://launchpad.net/bugs/1735407 - Assigned to Balazs Gibizer (balazs-gibizer)
16:11:48 mriedem artom: yeah - which was my original complaint in the earlier fix
16:12:03 dansmith sauloaugusto: the migration steps from no cells to one cell were required before pike
16:12:09 efried cdent Not sure we quite have that mechanism yet.
16:12:13 mriedem artom: we can't really roll everything back with what we have today in the conductor code
16:12:35 mriedem we'd either have to change the rpc cast to a call, or move the code that changes the instance from the api to conductor
16:12:37 cdent efried: indeed, thus why I've been modelling it the way I've been
16:12:40 efried cdent Yeah, a way to tie these brother-roots back to the compute host for purposes of allocation and deploy.
16:12:44 cdent if it is a child you know where to go
16:12:50 artom mriedem, yeah... I guess it's fine as a bandaid for now? Ideally we need to come back to this and fix it properly, maybe like what Andrey was proposing
16:12:50 mriedem hell, maybe we should have just moved the instance.save() from api to conductor...
16:12:52 cdent because
16:12:54 cdent it is a child
16:13:04 mriedem artom: yes this is a "tactical fix"
16:13:13 cdent a resource provider withing a compute-node is _obviously_ a child
16:13:22 efried cdent But if it's a child, the scheduler will happily deploy an instance with VCPU from pool1, MEMORY_MB from pool2, and DISK_GB from pool3.
16:13:24 cdent and it seems to me that the ironic model might be the thing that's weird
16:13:25 artom mriedem, heh, code for "someone else can deal with it once I'm retired" ;)
16:13:38 efried cdent Unless you *always* use a numbered granular group. Which doesn't seem like the right answer.
16:13:43 mriedem maybe - it's just something one of my old managers at ibm always said,
16:13:48 cfriesen jaypipes: for https://bugs.launchpad.net/nova/+bug/1742747 do we look at the cpu_allocation_ratio of the first host aggregate, or do we look for the first host aggregate that has cpu_allocation_ratio set?
16:13:49 openstack Launchpad bug 1742747 in OpenStack Compute (nova) "RT overrides default allocation_ratios for ram cpu and disk" [Undecided,In progress] - Assigned to Maciej Jozefczyk (maciej.jozefczyk)
16:13:54 mriedem "do we have a tactical fix while we work on the long-term strategic fix"
16:14:16 artom mriedem, not a bad way of thinking
16:14:26 efried I was doing a word puzzle the other day where the clue was "strategic" and the answer was "tactical". I thought how the IBM ppt-jockeys would flip out at that.
16:14:38 artom awaugama, you around? Since you're the cause of all this, if we build you our proposed tactical fix, you want to try and break it again?
16:14:39 sauloaugusto dansmith: Yes I did that, and I get get list o admin instances , and also create new instances at that . The problem that I can not do nothing at all other tenants .
16:14:53 artom Maybe if we let QE loose on it *before* merging it, we'll avoid the pain?
16:15:03 awaugama artom: I can do some sanity checks on it
16:15:09 mriedem artom: that never seems to happen
16:15:13 mriedem until it's in product
16:15:24 awaugama artom, and *Technically* I'm just the warning light, i'm not the cause of the issue
16:15:47 artom awaugama, hah, sorry, should have added a ;) or /s in there somewher e:)
16:15:52 awaugama ;)
16:15:58 cdent efried: these issues are why I pressed rado to start talking about and experimenting with this stuff. I think the sort of gravitational mechanics of nested providers is going to have lots of weird
16:16:23 cdent and I'd much prefer to see us working from concrete situations then not
16:16:48 efried cdent Agree. At the moment I'm actually leaning towards the thing I originally said was the worst idea: always using a single numbered request group.
16:17:59 Spazmotic Always good when the best idea is your worst idea
16:21:07 cdent efried: I guess we have time to experiement...
16:25:35 artom mriedem, https://bugs.launchpad.net/nova/+bug/1746032
16:25:36 openstack Launchpad bug 1746032 in OpenStack Compute (nova) "By rebuilding twice with the same "forbidden" image one can circumvent scheduler rebuild restrictions" [Undecided,New]
16:27:34 mriedem artom: L145 https://etherpad.openstack.org/p/nova-ptg-rocky
16:28:14 artom mriedem, oh hey and I'll actually be there this time :)
16:28:32 mriedem we can finally meet, face to face, and settle all the scores
16:28:38 openstackgerrit Balazs Gibizer proposed openstack/nova master: Reproduce bug 1724172 in the functional test env https://review.openstack.org/512553
16:28:38 openstackgerrit Balazs Gibizer proposed openstack/nova master: Enhance service restart in functional env https://review.openstack.org/512552
16:28:39 mriedem there can be only one
16:28:39 openstackgerrit Balazs Gibizer proposed openstack/nova master: cleanup evacuated instances not on hypervisor https://review.openstack.org/512623
16:28:39 openstack bug 1724172 in OpenStack Compute (nova) "Allocation of an evacuated instance is not cleaned on the source host if instance is not defined on the hypervisor" [Undecided,In progress] https://launchpad.net/bugs/1724172 - Assigned to Balazs Gibizer (balazs-gibizer)
16:28:49 artom ... one of *what*?
16:29:00 mriedem idk, i'm just thinking about the quickening
16:29:04 mriedem even though that was scotland
16:29:29 artom Man, your references are out of control!
16:31:40 edleafe cdent: I've been following along, but don't have a clear picture of how things are arranged. Is there a diagram or something that would illustrate this?
16:32:22 cdent edleafe: a) which "this"?, b) I wish I never showed up today because I simply don't have the capacity to actual engage well with this topic
16:32:31 openstackgerrit Matt Riedemann proposed openstack/nova master: Rollback instance.image_ref on failed rebuild https://review.openstack.org/538961
16:32:32 mriedem artom: ok here we go, single patch ^
16:32:55 edleafe cdent: the arrangement of entities in vmware
16:33:09 edleafe and which is the "root" that placement would be selecting
16:34:20 mriedem stable/newton is still open....but not sure we want to hold up EOL for this fix
16:35:03 Spazmotic Korea in.. 6 days
16:35:05 Spazmotic Whew
16:36:06 dansmith man, newton is like the undead release
16:36:15 efried edleafe I was considering drawing that up, along with an executive summary of this conversation, and putting it all in a response to the ML thread.
16:36:32 artom Can we call U 'undead'? As an hommage?
16:36:53 cdent edleafe: not that I'm aware of at this time, but I'll check with rado. if efried is able to do something (even if it has errors), that provides at least a starting point
16:37:10 cdent the "root" I was thinking of was "compute-node"
16:37:11 mriedem i'll say it again, the path to LTS upstream is just never EOL'ing anything
16:37:21 edleafe cdent: ok, thanks. I assumed that there must be some internal documentation that would be helpful
16:38:12 cdent edleafe: could very well be, but if there are pretty pictures I haven't seen them. That doesn't mean it doesn't exist. I interact _very_ little on that side of things.
16:38:55 Spazmotic Anyone know any of the cores familiar with XenAPI drivers? Besides Garbutt that is, I got his attention on but was curious if any others were familiar that I could poke :D
16:39:14 openstackgerrit Stephen Finucane proposed openstack/nova master: mypy: Add 'types' module https://review.openstack.org/538225
16:39:14 openstackgerrit Stephen Finucane proposed openstack/nova master: mypy: Add type annotations to 'nova.pci' https://review.openstack.org/538224
16:39:17 edleafe cdent: ok, cool. If you come across something, post it so we can all better understand what is needed
16:41:07 efried Spazmotic Maybe bauzas ?
16:41:36 artom Wasn't Bob Ball xenapi? Is he still around?
16:41:37 Spazmotic I know the driver isn't looked at too frequently by folks, so just didn't want to bark up the wrong trees wwhen I had questions or was looking for some eyeballs.
16:41:42 Spazmotic Yeah bob's still around but he's no longer core
16:42:09 Spazmotic He always gives me good ideas though so.. love me some Bob
16:47:27 openstackgerrit Matt Riedemann proposed openstack/nova master: Rollback instance.image_ref on failed rebuild https://review.openstack.org/538961
16:47:28 mriedem artom: jaypipes: added a variable for the original image ref and a comment about CastAsCall
16:47:51 mriedem Spazmotic: johnthetubaguy knows the xenapi stuff
16:48:01 mriedem and is still core, but not working as much upstream these days
16:48:09 artom mriedem, and defiantly ignored my grammar nit ;)
16:48:19 Spazmotic Yeah he's been busy with family stuff.. I got in touch with him the other day though, but figured i'd ask since we need 2 cores
16:48:35 johnthetubaguy yeah, I am around for some questions, much of the time
16:48:47 openstackgerrit Matt Riedemann proposed openstack/nova master: Rollback instance.image_ref on failed rebuild https://review.openstack.org/538961
16:48:48 mriedem artom: there you go
16:48:49 Spazmotic Now look what you did
16:48:52 johnthetubaguy (still trying to juggle back more reviewing time)
16:49:04 mriedem it's is always something i'll think of since school as a possessive
16:49:23 artom mriedem, cheers :)
16:49:40 Spazmotic hehe spend time with your family john, more important than reviews :D
16:49:43 jaypipes mriedem: danke
16:51:35 mriedem johnthetubaguy: if you are around, i'll badger you about https://review.openstack.org/#/c/527541/
16:51:59 johnthetubaguy mriedem: sure thing
16:53:59 johnthetubaguy mriedem: is the TODO about jay's join magic important?
16:54:08 mriedem no
16:54:29 jaypipes quoi?

Earlier   Later