openstackgerrit | jiang wei proposed openstack/nova master: Add action initiator attribute to the payload https://review.openstack.org/536243 | 00:38 |
---|---|---|
openstackgerrit | Takashi NATSUME proposed openstack/nova master: Update admin/flavors document https://review.openstack.org/573063 | 01:08 |
openstackgerrit | zhangdaolong proposed openstack/nova master: Fix error log output information parameters. https://review.openstack.org/575277 | 01:11 |
openstackgerrit | Zhenyu Zheng proposed openstack/nova master: Use ThreadPoolExecutor for max_concurrent_live_migrations https://review.openstack.org/563505 | 01:16 |
openstackgerrit | Zhenyu Zheng proposed openstack/nova master: Compute: add support to abort queued live migration https://review.openstack.org/568542 | 01:42 |
openstackgerrit | Corey Bryant proposed openstack/nova master: Fix execute mock for test_convert_image_with_errors https://review.openstack.org/575305 | 02:00 |
openstackgerrit | Takashi NATSUME proposed openstack/nova master: Subject: Remove mox in unit/network/test_neutronv2.py (8) https://review.openstack.org/575311 | 02:34 |
openstackgerrit | Takashi NATSUME proposed openstack/nova master: Remove mox in unit/network/test_neutronv2.py (8) https://review.openstack.org/575311 | 02:34 |
openstackgerrit | jiang wei proposed openstack/nova master: Add action initiator attribute to the payload https://review.openstack.org/536243 | 03:03 |
openstackgerrit | Takashi NATSUME proposed openstack/nova master: Add the guideline to write API reference https://review.openstack.org/569058 | 03:07 |
*** bhagyashris has quit IRC | 03:35 | |
openstackgerrit | Zhenyu Zheng proposed openstack/nova master: Add full traceback to ExceptionPayload in versioned notifications https://review.openstack.org/564092 | 03:41 |
*** bhagyashris has joined #openstack-placement | 03:46 | |
*** takashin has left #openstack-placement | 04:31 | |
*** e0ne has joined #openstack-placement | 04:53 | |
*** e0ne has quit IRC | 05:00 | |
openstackgerrit | jiang wei proposed openstack/nova master: Add action initiator attribute to the payload https://review.openstack.org/536243 | 05:36 |
openstackgerrit | yatin proposed openstack/nova stable/queens: [Stable Only] Initialise failed_builds in IronicNodeState https://review.openstack.org/574997 | 05:41 |
openstackgerrit | Zhenyu Zheng proposed openstack/nova master: Compute: add support to abort queued live migration https://review.openstack.org/568542 | 06:14 |
openstackgerrit | Zhenyu Zheng proposed openstack/nova master: Add full traceback to ExceptionPayload in versioned notifications https://review.openstack.org/564092 | 06:17 |
openstackgerrit | Shuo Liu proposed openstack/nova-specs master: fix spelling mistake https://review.openstack.org/573479 | 06:36 |
*** tssurya has joined #openstack-placement | 06:41 | |
*** ttsiouts has joined #openstack-placement | 07:14 | |
*** efried has quit IRC | 07:26 | |
*** efried has joined #openstack-placement | 07:27 | |
*** ttsiouts has quit IRC | 07:27 | |
*** tssurya_ has joined #openstack-placement | 07:27 | |
*** ttsiouts has joined #openstack-placement | 07:31 | |
*** tssurya_ has quit IRC | 07:31 | |
*** ttsiouts has quit IRC | 07:33 | |
openstackgerrit | jiang wei proposed openstack/nova master: Add action initiator attribute to the update payload https://review.openstack.org/536243 | 07:47 |
*** tetsuro has quit IRC | 07:49 | |
*** ttsiouts has joined #openstack-placement | 07:50 | |
*** ttsiouts has quit IRC | 07:56 | |
*** ttsiouts has joined #openstack-placement | 07:57 | |
*** takashin has joined #openstack-placement | 09:22 | |
*** e0ne has joined #openstack-placement | 09:28 | |
openstackgerrit | Alexandre arents proposed openstack/nova master: Preserve images_type of instance during live migration https://review.openstack.org/570528 | 09:43 |
*** nicolasbock has joined #openstack-placement | 10:37 | |
*** ttsiouts has quit IRC | 10:47 | |
openstackgerrit | zhangdaolong proposed openstack/nova master: Fix error log output information parameters. https://review.openstack.org/575277 | 10:50 |
*** cdent has joined #openstack-placement | 11:20 | |
openstackgerrit | Balazs Gibizer proposed openstack/nova master: Transfer port.resource_request to the scheduler https://review.openstack.org/567268 | 11:25 |
openstackgerrit | Balazs Gibizer proposed openstack/nova master: Send resource allocations in the port binding https://review.openstack.org/569459 | 11:25 |
openstackgerrit | Balazs Gibizer proposed openstack/nova master: Test boot with more ports with bandwidth request https://review.openstack.org/573317 | 11:25 |
*** nicolasbock has quit IRC | 11:42 | |
*** ttsiouts has joined #openstack-placement | 11:46 | |
openstackgerrit | Alexandre arents proposed openstack/nova master: Preserve images_type of instance during live migration https://review.openstack.org/570528 | 12:00 |
openstackgerrit | Merged openstack/nova master: Add trusted_image_certificates to REST API https://review.openstack.org/486204 | 12:01 |
openstackgerrit | Zhenyu Zheng proposed openstack/nova master: Use ThreadPoolExecutor for max_concurrent_live_migrations https://review.openstack.org/563505 | 12:03 |
openstackgerrit | Zhenyu Zheng proposed openstack/nova master: Compute: add support to abort queued live migration https://review.openstack.org/568542 | 12:07 |
*** nicolasbock has joined #openstack-placement | 12:11 | |
*** cdent has quit IRC | 12:29 | |
openstackgerrit | Corey Bryant proposed openstack/nova master: Fix execute mock for test_convert_image_with_errors https://review.openstack.org/575305 | 12:33 |
*** edmondsw has joined #openstack-placement | 12:54 | |
*** mriedem has joined #openstack-placement | 13:03 | |
openstackgerrit | Merged openstack/nova stable/queens: [Stable Only] Initialise failed_builds in IronicNodeState https://review.openstack.org/574997 | 13:11 |
efried | jaypipes: You gonna hit https://review.openstack.org/#/c/559480/ today? Should be a pretty easy re-approve (you were +A at PS21). | 13:39 |
jaypipes | efried: ack | 13:48 |
*** ttsiouts has quit IRC | 13:55 | |
*** tetsuro has joined #openstack-placement | 14:00 | |
jaypipes | efried: done. | 14:02 |
efried | cool mon | 14:02 |
*** ttsiouts has joined #openstack-placement | 14:03 | |
*** cdent has joined #openstack-placement | 14:14 | |
openstackgerrit | Balazs Gibizer proposed openstack/nova master: Add notification support for trusted_certs https://review.openstack.org/563269 | 14:17 |
openstackgerrit | Balazs Gibizer proposed openstack/nova master: Add certificate validation docs https://review.openstack.org/560158 | 14:17 |
openstackgerrit | Balazs Gibizer proposed openstack/nova master: WIP: Add trusted certs to feature support matrix docs https://review.openstack.org/574890 | 14:17 |
openstackgerrit | Balazs Gibizer proposed openstack/nova master: Remove max_size parameter from fake_libvirt_utils.fetch_*image methods https://review.openstack.org/574911 | 14:17 |
cdent | jaypipes++ on the rant | 14:19 |
* efried notes that jaypipes just got incremented in place | 14:20 | |
cdent | efried: are you familiar with irc karma trackers? I hope we never have one, but it is hard habit to break | 14:22 |
efried | cdent: No, never heard of it. | 14:22 |
cdent | basically any time you ++ a nick that person's karma is incremented | 14:24 |
cdent | and you can list it | 14:24 |
efried | whoah | 14:27 |
efried | Sounds like a popularity contest, despite anyone's best efforts to the contrary. | 14:27 |
efried | Do karma points age out, or is it also a measure of how long you've been around? | 14:27 |
cdent | it only works in small groups, which is why I think it should never happen here | 14:28 |
cdent | of course, I tend to think that chat of any form only works in small groups | 14:28 |
cdent | so... | 14:28 |
cdent | re aging, it depends on the bot | 14:28 |
cdent | it's not a universal thing | 14:28 |
*** ttsiouts has quit IRC | 14:29 | |
*** ttsiouts has joined #openstack-placement | 14:31 | |
openstackgerrit | Chris Dent proposed openstack/nova master: [placement] Add status and links fields to version document at / https://review.openstack.org/575117 | 14:32 |
cdent | efried: an opportunistic rebase on ^ to try to overcome the cellsv1 thing (has the fix merged yet?). can you put your imprimatur back on it please? | 14:33 |
efried | cdent: yes, the cellsv1 fix merged (see email). Looking... | 14:33 |
efried | cdent: You're planning to compose a placement email tomorrow? | 14:34 |
cdent | yes | 14:34 |
efried | cdent: If you remember, would you please update the link at https://wiki.openstack.org/w/index.php?title=Meetings/NovaScheduler&action=edit§ion=4 ? | 14:34 |
cdent | sure | 14:34 |
efried | thx | 14:34 |
cdent | It will be a bit harder to write the update email, because I usually do that with the laptop, in a coffee place, but I think I can overcome that limitation | 14:35 |
efried | hehehe | 14:35 |
cdent | but it may have a different tone as I'll be standing | 14:35 |
efried | for sure | 14:35 |
openstackgerrit | Dan Smith proposed openstack/nova stable/pike: Change consecutive build failure limit to a weigher https://review.openstack.org/573248 | 14:35 |
*** tetsuro has quit IRC | 14:46 | |
cdent | efried: on placement meeting chair, I'm meh about being on the hook for that since I'm already on the hook (self imposed) for the weekly update | 14:46 |
cdent | both is ... tedious | 14:46 |
cdent | so while no one else wants it, I'd prefer to be interim | 14:46 |
*** takashin has left #openstack-placement | 15:01 | |
*** e0ne has quit IRC | 15:03 | |
openstackgerrit | Merged openstack/nova master: XenAPI: define a new image handler to use vdi streaming https://review.openstack.org/486475 | 15:08 |
openstackgerrit | Merged openstack/nova master: XenAPI: update the document related to vdi streaming https://review.openstack.org/568444 | 15:08 |
*** e0ne has joined #openstack-placement | 15:40 | |
*** ttsiouts has quit IRC | 15:45 | |
openstackgerrit | Merged openstack/nova master: Add enhanced KVM storage QoS quotas https://review.openstack.org/558530 | 15:47 |
openstackgerrit | Merged openstack/nova master: placement: Allocation.consumer field https://review.openstack.org/565405 | 16:10 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: objects: Add NUMACell.network_info https://review.openstack.org/564439 | 16:23 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: network: Unchain _get_phynet_info from _get_port_vnic_info https://review.openstack.org/564443 | 16:23 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: network: Add 'create_resource_requests' to network API https://review.openstack.org/564444 | 16:23 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: network: Retrieve tunneled status in '_get_network_info' https://review.openstack.org/564445 | 16:23 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: objects: Remove NUMATopologyLimits.obj_from_db_obj https://review.openstack.org/537412 | 16:23 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: objects: Remove legacy '_to_dict' functions https://review.openstack.org/537413 | 16:23 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: hardware: Start accounting for networks in NUMA placement https://review.openstack.org/564448 | 16:23 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: objects: Add RequestSpec.numa_networks https://review.openstack.org/564442 | 16:23 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: scheduler: Start utilizing RequestSpec.numa_networks https://review.openstack.org/564452 | 16:23 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: conf: Add '[neutron] physnets' and related options https://review.openstack.org/564440 | 16:23 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: libvirt: Start populating NUMACell.network_info field https://review.openstack.org/564441 | 16:23 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: objects: Add NUMATopologyLimits.networks https://review.openstack.org/575486 | 16:23 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: DNM: objects: Store RequestSpec.numa_networks https://review.openstack.org/575487 | 16:23 |
openstackgerrit | Stephen Finucane proposed openstack/nova master: network: Remove 'create_pci_requests_for_sriov_ports' https://review.openstack.org/575488 | 16:23 |
openstackgerrit | Dan Smith proposed openstack/nova master: Make check_can_live_migrate_destination use long_rpc_timeout https://review.openstack.org/575493 | 16:47 |
*** e0ne has quit IRC | 16:48 | |
efried | jaypipes: Well, we have support for nested providers in alloc cands in microversion 1.27. | 16:52 |
efried | jaypipes: Can't say exactly where it crept in, but it's there in current master. | 16:52 |
efried | cdent, edleafe: How do we handle this? | 16:53 |
efried | Do we need to put up code to revert the behavior in 1.27? | 16:53 |
cdent | i would hope/think not | 16:53 |
cdent | what we have, whenever it leaked in, is a microversion that has incomplete support for nested | 16:54 |
cdent | or even wrong | 16:54 |
cdent | and what 1.28 gives us is correct | 16:54 |
cdent | you should be able to repeat your gabbi trick, downgrading the microversion until it blows up? | 16:55 |
efried | I.e. "don't worry about the fact that you get different behaviors in 1.27 depending which commit you're working against, because you shouldn't be trying to do this against nrp in the first place, you dolt" ? | 16:55 |
cdent | no, more like: CD has risks, it's okay | 16:56 |
efried | cdent: It's not even a 1.27 thing. The leak is about which code you're running. | 16:56 |
cdent | okay my statement still stands: sometimes master won't be right | 16:57 |
cdent | rarely | 16:57 |
efried | The test passes in 1.13 | 16:57 |
cdent | which test are your running? | 16:57 |
efried | here, I'll put up a DNM to show you what I've done... | 16:59 |
cdent | k, i'll be slow to respond, going on a call... | 16:59 |
openstackgerrit | Eric Fried proposed openstack/nova master: DNM: Leaked nrp-in-alloc-cands behavior https://review.openstack.org/575497 | 16:59 |
efried | cdent: ^ | 16:59 |
efried | What I think I hear you saying is that we should git r merged before we tag the next thingy, and not worry about it. | 17:00 |
*** e0ne has joined #openstack-placement | 17:00 | |
efried | off to lunch | 17:00 |
cdent | efried: pretty much | 17:01 |
cdent | efried: if you can provide a bit more detail on what's wrong and what right would look like that would be great. I'm sure I can figure it out, but with the number of calls and shit going on, don't make me think | 17:02 |
*** e0ne has quit IRC | 17:10 | |
*** e0ne has joined #openstack-placement | 17:13 | |
openstackgerrit | Elod Illes proposed openstack/nova master: Reject interface attach with QoS aware port https://review.openstack.org/570078 | 17:16 |
openstackgerrit | Elod Illes proposed openstack/nova master: Reject networks with QoS policy https://review.openstack.org/570079 | 17:16 |
*** e0ne has quit IRC | 17:19 | |
openstackgerrit | Matt Riedemann proposed openstack/nova master: Add troubleshooting item about ignored microversions https://review.openstack.org/575506 | 17:23 |
*** e0ne has joined #openstack-placement | 17:24 | |
*** e0ne has quit IRC | 17:27 | |
efried | cdent: All I know at the moment is that, at some point (maybe multiple points) in the nrp-in-alloc-candidates series, we made behavior changes without microversions. | 17:45 |
*** e0ne has joined #openstack-placement | 17:45 | |
efried | Prior to that series, if you asked for allocation candidates against a setup with trees, you would only get candidates that involved a single non-sharing provider. | 17:45 |
cdent | efried: when the series is fully merged, will we have something like the right behavior (at the proper microversions)? | 17:46 |
cdent | If so, I'd say: let's just make that happen, and try not to worry | 17:46 |
efried | cdent: Well, certainly the idea is for 1.28 (or more likely 1.29, since consumer gens will likely beat this) to behave properly for provider trees. The question is whether the behavior changes prior to that microversion are acceptable. | 17:47 |
efried | cdent: So there's some commit before which you could only get back an allocation_request with multiple providers in it if N-1 of them were sharing. But after that commit, you could get back allocation_requests with multiple non-sharing providers from the same tree. | 17:49 |
cdent | i'm asking: after extant code merges is <= 1.27 doing what we expect and is 1.28 doing what we expect. If so, let's just ship code. | 17:50 |
cdent | If something is out there there running off a commit that is doing the wrong thing, we can just tell them to use newer code | 17:50 |
efried | cdent: I guess that's what I'm getting at. If someone's using Queens today, and they have provider trees for whatever reason, and they're using GET /allocation_candidates (at an old microversion), and they're relying on the fact that, when run against those trees, they only get results that involve one non-sharing provider... | 17:52 |
efried | cdent: And then they upgrade to rocky and change nothing else... | 17:52 |
efried | cdent: Then they'll start getting different behavior (more results back). | 17:52 |
efried | So is that acceptable on the reasoning that "Provider trees are unsupported/undefined/unpredictable prior to microversion 1.2{8|9}, so anything you do against them, including GET /allocation_candidates, is at your own risk" ? | 17:53 |
cdent | I'm not sure we are communicating each other's comments/questions to each other well. /me thinks | 17:53 |
efried | cdent: I think you're asking whether, after this series merges, 1.27 will go back to behaving properly | 17:54 |
*** tssurya has quit IRC | 17:54 | |
efried | cdent: If it does, then we're okay, because we were only "broken" for a few commits in the middle of an active development cycle. | 17:56 |
cdent | i had the impression from what you were saying that during rocky we've merged code that changed how things behave (at a variety of microversions) | 17:56 |
cdent | but sometimes it seems like you are saying: queens is also broken | 17:56 |
efried | cdent: I'm saying that first thing - we've merged code during Rocky that changes behavior. | 17:57 |
efried | I don't think q is broken. | 17:57 |
cdent | my statement is: whatever the changes are between the beginning of rocky and the end of rocky, as long as at the end of rocky the diff between rocky and queens is as expected, we are okay | 17:57 |
cdent | because running any unreleased version of rocky is an acceptance of risk | 17:57 |
edleafe | that sounds like an argument for one microversion per release :) | 17:58 |
openstackgerrit | Chris Dent proposed openstack/nova master: Optional separate database for placement API https://review.openstack.org/362766 | 17:58 |
openstackgerrit | Chris Dent proposed openstack/nova master: Isolate placement database config https://review.openstack.org/541435 | 17:58 |
openstackgerrit | Chris Dent proposed openstack/nova master: Ensure that os-traits sync is attempted only at start of process https://review.openstack.org/553857 | 17:58 |
cdent | hush your mouth edleafe :) | 17:58 |
cdent | striving to be CD is good. failing sometimes is okay too. | 17:59 |
efried | cdent: Gotcha. | 17:59 |
openstackgerrit | Chris Dent proposed openstack/nova master: Add PLACEMENT_DB_ENABLED=True to the nova-next job https://review.openstack.org/564067 | 17:59 |
cdent | efried: if you disagree, that's cool. I'm trying to see a path to keep us moving well, accepting a bit of slack. | 18:00 |
cdent | it sounds like in this case we probably could have had some better tests to insure that behaviors were observable rather than you discovering them by chance | 18:01 |
cdent | such as: when nested exist, acs shoud looks like this at microversion x, y, z | 18:01 |
cdent | anyway, I'm being called away to dinner | 18:01 |
efried | right, I think that's what we don't have, agree we should add it. | 18:01 |
efried | Thanks for the talk. | 18:01 |
*** e0ne has quit IRC | 18:27 | |
openstackgerrit | Matt Riedemann proposed openstack/nova master: Fix nits from trusted certs notification change https://review.openstack.org/575521 | 18:30 |
*** tssurya has joined #openstack-placement | 18:33 | |
*** tssurya has quit IRC | 18:35 | |
*** tssurya has joined #openstack-placement | 18:35 | |
*** cdent has quit IRC | 18:55 | |
*** e0ne has joined #openstack-placement | 19:15 | |
jaypipes | efried: sorry, been out all day. what's happened? | 19:33 |
jaypipes | efried: something about a microversion problem? | 19:33 |
efried | jaypipes: Yeah, the nrp-in-alloc-cands series has already enabled nrp-in-alloc-cands in current master. | 19:34 |
jaypipes | efried: nothing *uses* the returned values from n-r-p in alloc cands, though. | 19:35 |
jaypipes | efried: or am I missing something? | 19:35 |
jaypipes | efried: as in, is there anything that is yet asking for granular request groups? | 19:36 |
efried | Nothing to do with granular. | 19:36 |
jaypipes | efried: because granular request groups are required for n-r-p to even be useful to the scheduler. | 19:36 |
efried | no | 19:36 |
efried | they're not | 19:36 |
jaypipes | efried: and there isn't anything yet actually creating nested providers (on the compute side) so... | 19:37 |
efried | jaypipes: A potentially busted scenario. In Queens, we support creating nrps, but if you GET /allocation_candidates against a nrp, the only candidates you get back are ones that are satisfied by a single RP (anywhere in the tree). In Nova, that'll always be the root RP, because that's always where the VCPU and MEMORY_MB live. But even in Nova, there wouldn't be anything stopping you from putting other resources in n | 19:37 |
efried | Now you upgrade to Rocky, but change nothing else, including the microversion you were using to GET /allocation_candidates. But now instead of getting back only candidates satisfied by a single non-sharing provider, you start getting back candidates satisfied by multiple non-sharing providers. | 19:38 |
efried | Bam, behavior change, potential for bustage. | 19:38 |
efried | It won't be a problem in Nova. | 19:38 |
efried | But in theoretical other-consumer-of-placement-land, it would be. | 19:38 |
jaypipes | efried: I'm failing to see how it would be a problem... | 19:39 |
efried | What Chris asserted was that, as long as by the end of this series, we're back to pre-1.28 behaving like it used to with respect to nrps, we're fine. | 19:39 |
openstackgerrit | Matt Riedemann proposed openstack/nova master: Add certificate validation docs https://review.openstack.org/560158 | 20:13 |
openstackgerrit | Matt Riedemann proposed openstack/nova master: WIP: Add trusted certs to feature support matrix docs https://review.openstack.org/574890 | 20:13 |
openstackgerrit | Matt Riedemann proposed openstack/nova master: Remove max_size parameter from fake_libvirt_utils.fetch_*image methods https://review.openstack.org/574911 | 20:13 |
openstackgerrit | Matt Riedemann proposed openstack/nova master: Fix nits from trusted certs notification change https://review.openstack.org/575521 | 20:13 |
efried | jaypipes: Sorry, multiple context switches | 20:14 |
efried | jaypipes: But what we need is some test cases that show what the behavior of pre-1.28 GET /a-c looked like against nrps so we can be relatively sure we've matched it once this series is fully landed. | 20:15 |
efried | jaypipes: It's a problem if a consumer has come to rely on GET /a-c at e.g. microversion 1.13 behaving a certain way despite the presence of child providers, which have been available to create since Queens. | 20:16 |
*** ttsiouts has joined #openstack-placement | 20:19 | |
*** e0ne_ has joined #openstack-placement | 20:26 | |
*** ttsiouts has quit IRC | 20:28 | |
*** ttsiouts has joined #openstack-placement | 20:28 | |
*** e0ne has quit IRC | 20:29 | |
jaypipes | efried: I guess I'm just thick. I don't see the specific problem that is being worried about. perhaps you could elaborate? if a client calls GET /a-c with 1.13 and gets a certain result and then calls GET /a-c with 1.27 and gets a different result, what's wrong with that? I thought 1.27 was all about changing the result of GET /a-c to include all resource class inventories in provider_summaries instead of only the inventories for resources | 20:34 |
jaypipes | that were requested, right? so the client is already expecting that behaviour change, no? or are you saying that before 1.27 we were not returning multiple providers in *a single provider tree* a single allocation request, but after 1.27 we *are* returning multiple providers in a single provider tree? because we have been returning multiple providers (sharing and non-sharing) in a single allocation request for a long time now. | 20:34 |
jaypipes | efried: I'm not trying to be obtuse, honest. I'm just struggling to understand what the core of the issue is. | 20:35 |
efried | jaypipes: It's that second thing I'm saying - except the distinction is that we've only been returning an alloc_req that spans N providers if N-1 of them are sharing. | 20:36 |
efried | jaypipes: Whereas now, in this pre-1.28 but post-beginning-of-this-series master version of 1.13, we'll return an alloc_req that spans N providers even if >1 of them are non-sharing. | 20:37 |
jaypipes | efried: ok, understood. | 20:37 |
efried | i.e. a superset of the allocation_request results we used to return | 20:37 |
jaypipes | efried: agree with cdent then... not sure it's a huge issue if we "close the loop" by the end of this cycle. | 20:37 |
efried | jaypipes: right, I agree with that, so the action to take is to make tests that prove that. | 20:38 |
jaypipes | efried: ack | 20:38 |
efried | Which is going to be a little tricky to do because we're broken now. | 20:38 |
jaypipes | efried: you keep using this word "broken".. :) | 20:38 |
efried | Sorry, "altered" | 20:38 |
jaypipes | efried: are you saying there is no way to tell which behaviour we're currently displaying? | 20:38 |
efried | which is the word you use for a neutered pet. | 20:38 |
jaypipes | lol | 20:38 |
efried | jaypipes: Well, I can show you the altered behavior in the current master... | 20:39 |
efried | jaypipes: https://review.openstack.org/575497 | 20:39 |
efried | this copies the fixture update (to make the fixture nested) and one of the gabbi tests from the top of the series | 20:39 |
efried | jaypipes: but sets the microversion to 1.13 and proves we still get the nested results. | 20:40 |
jaypipes | efried: k, good example, thanks. | 20:40 |
jaypipes | that gabbit clears it up for me, thanks | 20:40 |
jaypipes | (that's why I like gabbits, btw...) | 20:40 |
jaypipes | declarative HTTP diffs FTW. | 20:41 |
efried | jaypipes: So the thing to do will probably be some extension of this, but composed against queens, and then carried along until it is no longer broken in rocky. | 20:41 |
efried | sorry, altered | 20:41 |
efried | except in that case we'll actually write the case so it's working in queens, but broken in rocky. | 20:41 |
openstackgerrit | Matt Riedemann proposed openstack/nova master: Add trusted certs to feature support matrix docs https://review.openstack.org/574890 | 20:43 |
openstackgerrit | Matt Riedemann proposed openstack/nova master: Remove max_size parameter from fake_libvirt_utils.fetch_*image methods https://review.openstack.org/574911 | 20:43 |
openstackgerrit | Matt Riedemann proposed openstack/nova master: Fix nits from trusted certs notification change https://review.openstack.org/575521 | 20:43 |
openstackgerrit | Matt Riedemann proposed openstack/nova master: Time how long pre_live_migration() takes https://review.openstack.org/574481 | 20:47 |
openstackgerrit | Eric Fried proposed openstack/nova-specs master: WIP: Nova-Cyborg workflow https://review.openstack.org/575545 | 20:53 |
mriedem | efried: likely better to put that in here https://docs.openstack.org/nova/latest/reference/index.html#forward-looking-plans | 21:22 |
mriedem | the specs repo will complain about following the template | 21:22 |
efried | mriedem: I took the whole template out. I'm just using this as a dumping ground for this thing as a strawman, not really expecting to build a real spec around it. | 21:22 |
mriedem | efried: right, but it could live in the forward looking plans section as a strawman | 21:23 |
efried | mriedem: Do things in there not have to follow the spec template? | 21:23 |
mriedem | there have also been a couple of nova/cyborg workflow specs in cyborg's specs repo | 21:23 |
mriedem | efried: correct | 21:23 |
efried | mriedem: Not in their specs repo, which is confusing. But yeah, I've seen those, and they're in parts wrong and mostly too high-level. | 21:23 |
mriedem | yeah... https://specs.openstack.org/openstack/cyborg-specs/ | 21:24 |
mriedem | admin guide for cyborg specs... | 21:24 |
mriedem | sudo cat cyborg-specs | 21:24 |
efried | mriedem: https://review.openstack.org/#/q/project:openstack/cyborg-specs | 21:24 |
efried | I did start there, but there was a bunch of stuff I would have had to bootstrap about the specs repo, and I wanted to get this out quick without a lot of paperwork :) | 21:25 |
mriedem | was thinking of https://review.openstack.org/#/c/448228/ | 21:26 |
mriedem | which was then https://review.openstack.org/#/c/508572/ | 21:26 |
mriedem | yeah really high level | 21:27 |
*** tssurya has quit IRC | 21:30 | |
openstackgerrit | Merged openstack/nova master: [placement] Add status and links fields to version document at / https://review.openstack.org/575117 | 21:53 |
*** edmondsw has quit IRC | 21:54 | |
openstackgerrit | Merged openstack/nova master: Use oslo.messaging per-call monitoring https://review.openstack.org/566696 | 21:55 |
openstackgerrit | Matt Riedemann proposed openstack/nova master: Handle broken instance mapping in InstanceLister https://review.openstack.org/575556 | 22:01 |
openstackgerrit | Merged openstack/nova master: Add notification support for trusted_certs https://review.openstack.org/563269 | 22:03 |
*** e0ne_ has quit IRC | 22:24 | |
*** nicolasbock has quit IRC | 22:34 | |
*** ttsiouts has quit IRC | 23:17 | |
*** takashin has joined #openstack-placement | 23:58 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!