*** hamzy has joined #openstack-nova | 00:00 | |
*** Apoorva has quit IRC | 00:03 | |
*** zz_dimtruck is now known as dimtruck | 00:04 | |
*** annegentle has quit IRC | 00:06 | |
*** aweeks has quit IRC | 00:06 | |
*** ijw has quit IRC | 00:06 | |
*** ijw has joined #openstack-nova | 00:07 | |
*** baoli has quit IRC | 00:07 | |
*** lamt has quit IRC | 00:14 | |
*** ducttape_ has quit IRC | 00:18 | |
*** thorst_ has joined #openstack-nova | 00:19 | |
*** ducttape_ has joined #openstack-nova | 00:20 | |
*** thorst_ has quit IRC | 00:22 | |
*** thorst_ has joined #openstack-nova | 00:23 | |
*** yingjun has joined #openstack-nova | 00:23 | |
*** catintheroof has quit IRC | 00:23 | |
*** catintheroof has joined #openstack-nova | 00:24 | |
openstackgerrit | Dan Smith proposed openstack/nova: Make servers api view load instance fault from proper cell https://review.openstack.org/407238 | 00:24 |
---|---|---|
openstackgerrit | Dan Smith proposed openstack/nova: Add schedule_and_build_instances conductor method https://review.openstack.org/367557 | 00:24 |
openstackgerrit | Dan Smith proposed openstack/nova: WIP Multicell support for instance listing https://review.openstack.org/396775 | 00:24 |
openstackgerrit | Dan Smith proposed openstack/nova: Make RPCFixture support multiple connections https://review.openstack.org/396417 | 00:24 |
openstackgerrit | Dan Smith proposed openstack/nova: Setup CellsV2 environment in base test https://review.openstack.org/406380 | 00:24 |
openstackgerrit | Dan Smith proposed openstack/nova: Add SecurityGroup.identifier to prefer uuid over name https://review.openstack.org/406260 | 00:24 |
openstackgerrit | Dan Smith proposed openstack/nova: Add SingleCellSimple fixture https://review.openstack.org/405684 | 00:24 |
openstackgerrit | Dan Smith proposed openstack/nova: WIP Move instance creation to conductor https://review.openstack.org/319379 | 00:24 |
openstackgerrit | Dan Smith proposed openstack/nova: Fix up non-cells-aware context managers in test_db_api https://review.openstack.org/406379 | 00:24 |
*** yingjun_ has joined #openstack-nova | 00:25 | |
*** claudiub has joined #openstack-nova | 00:28 | |
*** catintheroof has quit IRC | 00:28 | |
*** yingjun has quit IRC | 00:29 | |
*** claudiub|2 has quit IRC | 00:31 | |
*** thorst_ has quit IRC | 00:32 | |
*** tlian has quit IRC | 00:35 | |
*** tjones has quit IRC | 00:35 | |
*** ducttape_ has quit IRC | 00:36 | |
*** mtanino has quit IRC | 00:39 | |
*** diana_clarke has quit IRC | 00:39 | |
*** toure|biab is now known as toure | 00:47 | |
*** rama_y has quit IRC | 00:47 | |
*** READ10 has joined #openstack-nova | 00:50 | |
*** tuanluong has joined #openstack-nova | 00:51 | |
*** yuntongjin has joined #openstack-nova | 00:51 | |
*** tovin07 has joined #openstack-nova | 00:52 | |
*** diana_clarke has joined #openstack-nova | 00:57 | |
*** ijw has quit IRC | 00:59 | |
*** hongbin has quit IRC | 01:00 | |
*** STEWIE925 has quit IRC | 01:01 | |
*** Apoorva has joined #openstack-nova | 01:02 | |
*** Apoorva has quit IRC | 01:02 | |
*** ducttape_ has joined #openstack-nova | 01:04 | |
*** Apoorva has joined #openstack-nova | 01:04 | |
*** Apoorva has quit IRC | 01:04 | |
*** dave-mccowan has joined #openstack-nova | 01:04 | |
*** Jeffrey4l has quit IRC | 01:06 | |
*** Jeffrey4l has joined #openstack-nova | 01:06 | |
*** jamielennox is now known as jamielennox|away | 01:07 | |
*** Jeffrey4l has quit IRC | 01:08 | |
*** Jeffrey4l has joined #openstack-nova | 01:09 | |
*** annegentle has joined #openstack-nova | 01:09 | |
*** lamt has joined #openstack-nova | 01:11 | |
*** zhangjl has joined #openstack-nova | 01:14 | |
*** annegentle has quit IRC | 01:14 | |
*** ducttape_ has quit IRC | 01:18 | |
*** Jeffrey4l has quit IRC | 01:20 | |
*** Jeffrey4l has joined #openstack-nova | 01:21 | |
*** jamielennox|away is now known as jamielennox | 01:21 | |
*** zhangjl has quit IRC | 01:24 | |
*** kevinz has joined #openstack-nova | 01:28 | |
*** david-lyle has joined #openstack-nova | 01:33 | |
*** yingjun_ has left #openstack-nova | 01:33 | |
*** mriedem has joined #openstack-nova | 01:36 | |
*** READ10 has quit IRC | 01:37 | |
*** claudiub has quit IRC | 01:37 | |
*** Jeffrey4l has quit IRC | 01:39 | |
*** Jeffrey4l has joined #openstack-nova | 01:39 | |
mriedem | dansmith: did you put up the g-r bump for https://review.openstack.org/#/c/407719/ yet? | 01:40 |
dansmith | mriedem: not the nova one | 01:40 |
dansmith | mriedem: the release created a u-c bump in requirements | 01:41 |
dansmith | that merges then we get an auto-propose right? | 01:41 |
openstackgerrit | wanghao proposed openstack/nova: Set min_disk in the image meta based on the root volume's size https://review.openstack.org/407397 | 01:42 |
mriedem | no we need the min g-r version bump | 01:42 |
mriedem | so we need both i guess | 01:42 |
mriedem | i'm on the case | 01:42 |
*** mingyu has joined #openstack-nova | 01:43 | |
*** smatzek has joined #openstack-nova | 01:43 | |
mriedem | i can't handle watching barbie's life in the dreamhouse right now | 01:43 |
melwitt | as opposed to other times ... :) | 01:43 |
mriedem | melwitt: you know it | 01:44 |
mriedem | can't wait to see what dumbass thing ken does this time | 01:44 |
melwitt | lol | 01:45 |
mriedem | dansmith: https://review.openstack.org/407794 | 01:47 |
*** nic has quit IRC | 01:47 | |
*** mingyu has quit IRC | 01:49 | |
*** yuntongjin has quit IRC | 01:49 | |
openstackgerrit | Ghanshyam Mann proposed openstack/nova: Rename few tests as per new config options https://review.openstack.org/406406 | 01:50 |
*** mingyu has joined #openstack-nova | 01:51 | |
openstackgerrit | Ghanshyam Mann proposed openstack/nova: Merge v21 view builder for limit API https://review.openstack.org/406787 | 01:51 |
*** ijw has joined #openstack-nova | 01:51 | |
*** wally__ has joined #openstack-nova | 01:52 | |
*** Guest42705 has joined #openstack-nova | 01:52 | |
*** gjayavelu has quit IRC | 01:53 | |
Guest42705 | Hi All, I want to install and configure nova-lxd on my exiting openstack(mitaka) . so that I can run the linux container using LXD, is there any user guide for this? | 01:53 |
Guest42705 | can anyone knows how to do it? | 01:54 |
*** Sukhdev has quit IRC | 01:54 | |
*** mdnadeem has joined #openstack-nova | 01:57 | |
*** ijw has quit IRC | 01:58 | |
*** yuntongjin has joined #openstack-nova | 01:58 | |
mriedem | Guest42705: see topic, this isn't a support channel. even still, lxd isn't in nova's tree, so you'd have to find support for that somewhere else | 01:59 |
mriedem | canonical maintains it i think | 01:59 |
mriedem | maybe find zigo? | 01:59 |
Guest42705 | thanks a lot\ | 02:00 |
*** Guest42705 has quit IRC | 02:03 | |
*** yuntongjin has quit IRC | 02:05 | |
*** Jeffrey4l has quit IRC | 02:06 | |
*** Jeffrey4l has joined #openstack-nova | 02:07 | |
*** smatzek has quit IRC | 02:07 | |
*** baoli has joined #openstack-nova | 02:12 | |
*** thorst_ has joined #openstack-nova | 02:14 | |
jaypipes | cmake .. && make && ctest -VV | 02:16 |
jaypipes | oopsie... wrong term | 02:16 |
*** baoli has quit IRC | 02:17 | |
*** ducnc has joined #openstack-nova | 02:18 | |
oomichi | alex_xu: hi, can you take a look at https://review.openstack.org/#/c/387727/ an https://review.openstack.org/#/c/382060/ ? | 02:18 |
*** Apoorva has joined #openstack-nova | 02:20 | |
*** zhangjl has joined #openstack-nova | 02:22 | |
*** thorst_ has quit IRC | 02:23 | |
*** jaypipes has quit IRC | 02:28 | |
gcb | gmann ping | 02:29 |
*** takashin has joined #openstack-nova | 02:30 | |
*** thorst_ has joined #openstack-nova | 02:31 | |
gcb | oomichi: can you help look at https://review.openstack.org/#/c/400697/ when you're free ? it got a +2 | 02:32 |
*** shangxdy has joined #openstack-nova | 02:32 | |
openstackgerrit | int32bit proposed openstack/python-novaclient: Fix can't process the resource with name 'help' https://review.openstack.org/403382 | 02:32 |
*** ducttape_ has joined #openstack-nova | 02:33 | |
*** browne has quit IRC | 02:33 | |
*** Apoorva has quit IRC | 02:34 | |
*** Zer0Byte__ has quit IRC | 02:36 | |
openstackgerrit | Takashi NATSUME proposed openstack/nova-specs: Amend cold-migration-with-target-ocata spec https://review.openstack.org/397193 | 02:36 |
openstackgerrit | Takashi NATSUME proposed openstack/nova: Refactor a test method including 7 test cases https://review.openstack.org/395436 | 02:37 |
openstackgerrit | Takashi NATSUME proposed openstack/nova: Refactor a test method including 3 test cases https://review.openstack.org/394738 | 02:37 |
openstackgerrit | Takashi NATSUME proposed openstack/nova: Remove mox in test_block_device_mapping_v1.py https://review.openstack.org/308746 | 02:37 |
oomichi | gcb: ok, I am checking. I did mistake a patch related to py3 before and the other project jobs were broken because we don't have py27 *tempest* jobs on nova gate | 02:37 |
openstackgerrit | Takashi NATSUME proposed openstack/nova: Stop using mox stubs in nova/tests/unit/cells https://review.openstack.org/323287 | 02:37 |
openstackgerrit | Takashi NATSUME proposed openstack/nova: api-ref: Fix 'id' (attachment_id) parameters https://review.openstack.org/349863 | 02:37 |
oomichi | gcb: I'd like to pay attention for avoiding it again | 02:37 |
*** hogepodge has quit IRC | 02:38 | |
gcb | oomichi, thanks I have been working on unit tests , Python 3 porting recently https://etherpad.openstack.org/p/nova_py35 | 02:39 |
oomichi | gcb: oh, great. That is a long list | 02:40 |
*** code-R has joined #openstack-nova | 02:41 | |
gcb | oomichi, there was work session about Python 3 in summit : https://etherpad.openstack.org/p/ocata-python-3 johnthetubaguy would like to help review. and merged some ones :-) | 02:42 |
gcb | that's the short history | 02:42 |
openstackgerrit | Tuan Luong-Anh proposed openstack/python-novaclient: Replaces uuid.uuid4 with uuidutils.generate_uuid() https://review.openstack.org/406835 | 02:43 |
gcb | oomichi , do we have any plan or discussion about port functional tests to Python 3 before ? | 02:43 |
oomichi | gcb: I don't have anything, I just reviewed these patches based on technical things by checking code, doc and blogs :) | 02:44 |
gcb | oomichi, thanks for your time :-) | 02:46 |
*** hongbin has joined #openstack-nova | 02:46 | |
*** thorst_ has quit IRC | 02:50 | |
*** hfu has joined #openstack-nova | 02:50 | |
*** thorst_ has joined #openstack-nova | 02:50 | |
*** liuyulong_ has joined #openstack-nova | 02:52 | |
*** liuyulong has quit IRC | 02:55 | |
*** kaisers__ has joined #openstack-nova | 02:57 | |
openstackgerrit | wanghao proposed openstack/nova: Set min_disk in the image meta based on the root volume's size https://review.openstack.org/407397 | 02:59 |
*** kristian__ has joined #openstack-nova | 02:59 | |
*** thorst_ has quit IRC | 02:59 | |
oomichi | gcb: I am not sure the reason of "Replace six.iteritems(dict) with dict.items()" yet | 02:59 |
*** kaisers_ has quit IRC | 03:00 | |
*** zhangjl has quit IRC | 03:02 | |
*** kristian__ has quit IRC | 03:03 | |
*** mdnadeem has quit IRC | 03:04 | |
oomichi | gcb: according to http://python-future.org/compatible_idioms.html#iterating-through-dict-keys-values-items , direct items() without six seems inefficient on py2 | 03:04 |
*** fragatina has quit IRC | 03:06 | |
*** hfu has quit IRC | 03:07 | |
*** mingyu has quit IRC | 03:07 | |
*** hfu has joined #openstack-nova | 03:07 | |
*** sreeram_v has joined #openstack-nova | 03:08 | |
openstackgerrit | Diana Clarke proposed openstack/python-novaclient: Microversion 2.39 - Simple tenant usage pagination https://review.openstack.org/407825 | 03:11 |
openstackgerrit | liaonanhai proposed openstack/nova: Use uuidutils instead of uuid.uuid4() https://review.openstack.org/394775 | 03:13 |
openstackgerrit | OpenStack Proposal Bot proposed openstack/python-novaclient: Updated from global requirements https://review.openstack.org/406250 | 03:16 |
*** zhangjl has joined #openstack-nova | 03:17 | |
*** hfu has quit IRC | 03:19 | |
openstackgerrit | xianming.mao proposed openstack/nova: Move attach operation into the try-except https://review.openstack.org/407828 | 03:20 |
*** hfu has joined #openstack-nova | 03:20 | |
*** browne has joined #openstack-nova | 03:21 | |
*** prateek has joined #openstack-nova | 03:22 | |
*** links has joined #openstack-nova | 03:22 | |
*** Jeffrey4l has quit IRC | 03:23 | |
*** Jeffrey4l has joined #openstack-nova | 03:24 | |
*** hfu has quit IRC | 03:24 | |
*** dave-mccowan has quit IRC | 03:25 | |
*** hfu has joined #openstack-nova | 03:26 | |
*** dave-mccowan has joined #openstack-nova | 03:27 | |
*** tbachman_ has joined #openstack-nova | 03:28 | |
*** ducnc has quit IRC | 03:28 | |
*** GB21 has joined #openstack-nova | 03:29 | |
*** tbachman has quit IRC | 03:30 | |
*** tbachman_ is now known as tbachman | 03:30 | |
*** prateek has quit IRC | 03:32 | |
*** GB21 has quit IRC | 03:36 | |
*** hfu has quit IRC | 03:37 | |
*** hfu has joined #openstack-nova | 03:37 | |
*** browne has quit IRC | 03:37 | |
*** Jeffrey4l has quit IRC | 03:38 | |
*** Jeffrey4l has joined #openstack-nova | 03:38 | |
*** Amit_M has joined #openstack-nova | 03:39 | |
*** shangxdy has quit IRC | 03:40 | |
*** hfu has quit IRC | 03:43 | |
*** ducttape_ has quit IRC | 03:44 | |
*** ducttape_ has joined #openstack-nova | 03:45 | |
*** GordonTX has joined #openstack-nova | 03:46 | |
*** hfu has joined #openstack-nova | 03:47 | |
*** liuyulong__ has joined #openstack-nova | 03:47 | |
*** ducttape_ has quit IRC | 03:47 | |
*** ducttape_ has joined #openstack-nova | 03:47 | |
*** GordonTX has quit IRC | 03:48 | |
*** liuyulong_ has quit IRC | 03:49 | |
*** yufei has joined #openstack-nova | 03:50 | |
*** ducttape_ has quit IRC | 03:52 | |
*** sdatko has quit IRC | 03:54 | |
mriedem | adreznec: efried: do you guys have more coming than this anytime soon? like spawn/delete patches? https://review.openstack.org/#/c/391288/2 | 03:56 |
mriedem | that's the last progress in over a month | 03:57 |
*** thorst_ has joined #openstack-nova | 03:57 | |
*** hfu has quit IRC | 03:57 | |
*** hfu has joined #openstack-nova | 04:00 | |
*** hfu has quit IRC | 04:01 | |
*** sreeram_v has quit IRC | 04:01 | |
mriedem | jroll: any idea if we can make progress on this using a depends-on to the ironic changes? https://review.openstack.org/#/c/364413/ | 04:01 |
mriedem | hasn't been updated since september | 04:01 |
*** sreeram_v has joined #openstack-nova | 04:01 | |
mriedem | Kevin_Zheng: are you blocked on this? https://review.openstack.org/#/c/394321/ | 04:02 |
*** shangxdy has joined #openstack-nova | 04:02 | |
*** thorst_ has quit IRC | 04:04 | |
*** sdatko has joined #openstack-nova | 04:05 | |
*** dave-mccowan has quit IRC | 04:06 | |
*** sreeram_v has quit IRC | 04:06 | |
openstackgerrit | Vladik Romanovsky proposed openstack/nova: tests: avoid starting compute service twice in sriov functional test https://review.openstack.org/407572 | 04:06 |
openstackgerrit | Vladik Romanovsky proposed openstack/nova: tests: generate correct pci addresses for fake pci devices https://review.openstack.org/407730 | 04:06 |
*** mingyu has joined #openstack-nova | 04:08 | |
*** Jeffrey4l has quit IRC | 04:11 | |
*** Jeffrey4l has joined #openstack-nova | 04:11 | |
*** baoli has joined #openstack-nova | 04:12 | |
*** psachin has joined #openstack-nova | 04:12 | |
*** mingyu has quit IRC | 04:12 | |
*** tbachman has quit IRC | 04:13 | |
*** shangxdy has quit IRC | 04:14 | |
*** tbachman has joined #openstack-nova | 04:14 | |
*** pbandark has joined #openstack-nova | 04:15 | |
*** mdnadeem has joined #openstack-nova | 04:18 | |
gmann | gcb: pong, sorry i was stuck in other gate fix | 04:18 |
*** tbachman has quit IRC | 04:19 | |
*** links has quit IRC | 04:20 | |
*** armax has quit IRC | 04:21 | |
efried | mriedem (adreznec, thorst), regarding https://review.openstack.org/#/c/391288/ - we'd been waiting for core commentary and also a little held up with our CI. At this point, lots of vacation time, so we won't get spawn/delete done until January. I'm good with your strategy of ensuring that something more than the skeleton makes Ocata - we'll beef up that change set. | 04:23 |
*** ducttape_ has joined #openstack-nova | 04:27 | |
*** Jeffrey4l has quit IRC | 04:29 | |
*** Jeffrey4l has joined #openstack-nova | 04:30 | |
*** Jeffrey4l has quit IRC | 04:30 | |
*** Jeffrey4l has joined #openstack-nova | 04:31 | |
*** psachin has quit IRC | 04:32 | |
*** kaisers__ has quit IRC | 04:33 | |
*** huhaoran has joined #openstack-nova | 04:36 | |
*** code-R has quit IRC | 04:37 | |
*** ducttape_ has quit IRC | 04:39 | |
*** huhaoran has quit IRC | 04:41 | |
*** shangxdy has joined #openstack-nova | 04:46 | |
*** psachin has joined #openstack-nova | 04:49 | |
*** code-R has joined #openstack-nova | 04:53 | |
*** GB21 has joined #openstack-nova | 04:55 | |
*** links has joined #openstack-nova | 04:57 | |
*** shangxdy has quit IRC | 04:57 | |
*** hongbin has quit IRC | 04:58 | |
*** hfu has joined #openstack-nova | 04:59 | |
*** thorst_ has joined #openstack-nova | 05:01 | |
*** jamespage has quit IRC | 05:02 | |
*** baoli has quit IRC | 05:02 | |
*** baoli has joined #openstack-nova | 05:03 | |
*** jamespag` has joined #openstack-nova | 05:03 | |
*** ayogi has joined #openstack-nova | 05:03 | |
*** kaisers_ has joined #openstack-nova | 05:04 | |
*** hfu has quit IRC | 05:05 | |
*** nicolasbock has quit IRC | 05:07 | |
*** baoli has quit IRC | 05:07 | |
*** yamahata has quit IRC | 05:08 | |
*** kaisers_ has quit IRC | 05:09 | |
*** thorst_ has quit IRC | 05:09 | |
*** GB21 has quit IRC | 05:13 | |
*** leifz has quit IRC | 05:14 | |
*** robcresswell has quit IRC | 05:14 | |
*** claudiub has joined #openstack-nova | 05:14 | |
*** ameade_ has joined #openstack-nova | 05:14 | |
*** zhenguo_ has joined #openstack-nova | 05:14 | |
*** McNinja has quit IRC | 05:14 | |
*** ameade has quit IRC | 05:14 | |
*** lbragstad has quit IRC | 05:14 | |
*** fyxim has quit IRC | 05:14 | |
*** mgagne has quit IRC | 05:15 | |
*** sballe_ has quit IRC | 05:15 | |
*** dimtruck has quit IRC | 05:15 | |
*** zhenguo has quit IRC | 05:15 | |
*** dirk has quit IRC | 05:15 | |
*** macsz_ has joined #openstack-nova | 05:15 | |
*** zhenguo_ is now known as zhenguo | 05:15 | |
*** swamireddy has quit IRC | 05:15 | |
*** morgan has quit IRC | 05:15 | |
*** wasmum has quit IRC | 05:15 | |
*** ildikov has quit IRC | 05:15 | |
*** DuncanT has quit IRC | 05:15 | |
*** DuncanT_ has joined #openstack-nova | 05:15 | |
*** DuncanT_ is now known as DuncanT | 05:16 | |
*** krtaylor has quit IRC | 05:16 | |
*** gmann has quit IRC | 05:16 | |
*** betherly has quit IRC | 05:16 | |
*** jamielennox has quit IRC | 05:16 | |
*** macsz has quit IRC | 05:16 | |
*** ildikov has joined #openstack-nova | 05:16 | |
*** ameade_ is now known as ameade | 05:16 | |
*** macsz_ is now known as macsz | 05:17 | |
*** robcresswell has joined #openstack-nova | 05:17 | |
*** fragatina has joined #openstack-nova | 05:17 | |
*** fyxim has joined #openstack-nova | 05:19 | |
*** dirk has joined #openstack-nova | 05:19 | |
*** betherly has joined #openstack-nova | 05:19 | |
*** lbragstad has joined #openstack-nova | 05:20 | |
*** prateek has joined #openstack-nova | 05:20 | |
*** wasmum has joined #openstack-nova | 05:20 | |
*** sballe_ has joined #openstack-nova | 05:22 | |
gcb | gmann, got it , please help revisit python3 related patch when you're free , thanks | 05:23 |
*** swamireddy has joined #openstack-nova | 05:23 | |
*** morgan has joined #openstack-nova | 05:23 | |
*** gmann has joined #openstack-nova | 05:24 | |
gcb | oomichi: please look at section "Common patterns" in https://wiki.openstack.org/wiki/Python3 | 05:25 |
gcb | oomichi: Note: Replacing dict.iteritems()/.itervalues() with six.iteritems(dict)/six.itervalues(dict) was preferred in the past, but there was a discussion suggesting to avoid six for this. The overhead of creating a temporary list on Python 2 is negligible. | 05:26 |
*** fragatina has quit IRC | 05:26 | |
*** ratailor has joined #openstack-nova | 05:27 | |
*** fragatina has joined #openstack-nova | 05:27 | |
*** krtaylor has joined #openstack-nova | 05:28 | |
*** Sukhdev has joined #openstack-nova | 05:31 | |
*** jamielennox has joined #openstack-nova | 05:31 | |
*** ratailor has quit IRC | 05:31 | |
*** GB21 has joined #openstack-nova | 05:33 | |
gmann | alex_xu: johnthetubaguy oomichi : this seems backward incompatible right - https://review.openstack.org/#/c/402372/5 | 05:33 |
gmann | not sure cinder does not make volume in -use if no destination type | 05:34 |
*** Jeffrey4l has quit IRC | 05:35 | |
*** Jeffrey4l has joined #openstack-nova | 05:35 | |
*** kaisers_ has joined #openstack-nova | 05:39 | |
*** ducttape_ has joined #openstack-nova | 05:40 | |
*** kaisers_ has quit IRC | 05:42 | |
*** kaisers_ has joined #openstack-nova | 05:42 | |
*** ducttape_ has quit IRC | 05:45 | |
*** kaisers_ has quit IRC | 05:47 | |
*** ratailor has joined #openstack-nova | 05:47 | |
*** shangxdy has joined #openstack-nova | 05:49 | |
*** armax has joined #openstack-nova | 05:49 | |
openstackgerrit | int32bit proposed openstack/python-novaclient: [WIP]Add minDisk and minRam filters to list flavors https://review.openstack.org/407855 | 05:49 |
*** prateek has quit IRC | 05:51 | |
*** hfu has joined #openstack-nova | 05:53 | |
*** oanson has joined #openstack-nova | 05:54 | |
*** GordonTX has joined #openstack-nova | 05:56 | |
*** GordonTX has quit IRC | 05:56 | |
*** hfu has quit IRC | 05:57 | |
*** hfu has joined #openstack-nova | 05:58 | |
*** b_rake has joined #openstack-nova | 05:59 | |
*** b_rake has quit IRC | 05:59 | |
*** b_rake has joined #openstack-nova | 06:00 | |
*** kristian__ has joined #openstack-nova | 06:00 | |
*** kristian__ has quit IRC | 06:00 | |
*** hfu has quit IRC | 06:01 | |
*** kristian__ has joined #openstack-nova | 06:01 | |
*** jaosorior has joined #openstack-nova | 06:02 | |
*** b_rake has quit IRC | 06:04 | |
*** markvoelker has quit IRC | 06:05 | |
*** markvoelker has joined #openstack-nova | 06:05 | |
*** kristian__ has quit IRC | 06:05 | |
*** prateek has joined #openstack-nova | 06:07 | |
*** thorst_ has joined #openstack-nova | 06:07 | |
*** sudipto has joined #openstack-nova | 06:08 | |
*** sudipto_ has joined #openstack-nova | 06:08 | |
*** markvoelker has quit IRC | 06:10 | |
openstackgerrit | int32bit proposed openstack/python-novaclient: Add minDisk and minRam filters to list flavors https://review.openstack.org/407855 | 06:11 |
openstackgerrit | wanghao proposed openstack/nova: Set min_disk in the image meta based on the root volume's size https://review.openstack.org/407397 | 06:11 |
Kevin_Zheng | mriedem: Ah, sort of, it is mostly about should it be added to the existing extension or not | 06:12 |
*** thorst_ has quit IRC | 06:14 | |
*** wingwj has joined #openstack-nova | 06:14 | |
*** int32bit has joined #openstack-nova | 06:15 | |
*** mrsoul has quit IRC | 06:15 | |
*** mrsoul has joined #openstack-nova | 06:16 | |
*** armax has quit IRC | 06:18 | |
openstackgerrit | xianming.mao proposed openstack/nova: Wrap attach operation into the try-except https://review.openstack.org/407828 | 06:21 |
alex_xu | gmann: I thought it just make a API failed sliently. Then that sounds like a bug | 06:25 |
*** hfu has joined #openstack-nova | 06:26 | |
*** fragatina has quit IRC | 06:26 | |
*** hfu has quit IRC | 06:27 | |
openstackgerrit | xianming.mao proposed openstack/nova: Replace the assertEqual(None,A) with assertIsNone(A) https://review.openstack.org/407870 | 06:28 |
*** nandal has joined #openstack-nova | 06:28 | |
*** huhaoran has joined #openstack-nova | 06:30 | |
*** cburgess has quit IRC | 06:30 | |
*** sach-m has quit IRC | 06:31 | |
*** rmk has quit IRC | 06:31 | |
openstackgerrit | xianming.mao proposed openstack/nova: Wrap attachment operation with the try-except https://review.openstack.org/407828 | 06:31 |
*** hfu has joined #openstack-nova | 06:32 | |
*** hfu has quit IRC | 06:32 | |
*** kashyap has quit IRC | 06:32 | |
*** Cristina__ has joined #openstack-nova | 06:33 | |
*** hfu has joined #openstack-nova | 06:33 | |
*** rmk has joined #openstack-nova | 06:33 | |
*** rmk has quit IRC | 06:33 | |
*** rmk has joined #openstack-nova | 06:33 | |
*** cburgess has joined #openstack-nova | 06:33 | |
*** sach-m has joined #openstack-nova | 06:34 | |
*** Cristina_ has quit IRC | 06:36 | |
*** kashyap has joined #openstack-nova | 06:36 | |
*** sudipto has quit IRC | 06:37 | |
*** sudipto_ has quit IRC | 06:37 | |
*** diga has joined #openstack-nova | 06:38 | |
openstackgerrit | xianming.mao proposed openstack/nova: Wrap attachment operation with the try-except https://review.openstack.org/407828 | 06:38 |
*** jgriffith is now known as jgriffith_away | 06:39 | |
*** hfu has quit IRC | 06:40 | |
*** voelzmo has joined #openstack-nova | 06:43 | |
*** markus_z has joined #openstack-nova | 06:43 | |
*** kaisers_ has joined #openstack-nova | 06:43 | |
*** josecastroleon has joined #openstack-nova | 06:43 | |
*** voelzmo has quit IRC | 06:45 | |
*** voelzmo has joined #openstack-nova | 06:46 | |
*** huhaoran has quit IRC | 06:46 | |
openstackgerrit | Andia proposed openstack/nova: comment https://review.openstack.org/407877 | 06:47 |
openstackgerrit | Zhenyu Zheng proposed openstack/nova: Support tag instances when boot https://review.openstack.org/394321 | 06:47 |
*** kaisers_ has quit IRC | 06:47 | |
*** kristian__ has joined #openstack-nova | 06:48 | |
*** cfriesen has quit IRC | 06:48 | |
*** huhaoran has joined #openstack-nova | 06:49 | |
*** hfu has joined #openstack-nova | 06:49 | |
*** lpetrut has joined #openstack-nova | 06:50 | |
*** voelzmo has quit IRC | 06:50 | |
*** kristian__ has quit IRC | 06:51 | |
*** kristian__ has joined #openstack-nova | 06:51 | |
*** masber has joined #openstack-nova | 06:51 | |
*** mingyu has joined #openstack-nova | 06:54 | |
*** kristian__ has quit IRC | 06:55 | |
openstackgerrit | Andia proposed openstack/nova: quotas of instance in soft-delete vm_state https://review.openstack.org/407877 | 06:56 |
openstackgerrit | Tuan Luong-Anh proposed openstack/python-novaclient: Replaces uuid.uuid4 with uuidutils.generate_uuid() https://review.openstack.org/406835 | 06:56 |
openstackgerrit | Ghanshyam Mann proposed openstack/nova: Use bdm source type values from object fields https://review.openstack.org/407886 | 06:58 |
*** Sukhdev has quit IRC | 07:01 | |
*** kristian__ has joined #openstack-nova | 07:05 | |
*** kaisers_ has joined #openstack-nova | 07:05 | |
*** markvoelker has joined #openstack-nova | 07:06 | |
*** mjura has joined #openstack-nova | 07:09 | |
*** hfu has quit IRC | 07:09 | |
*** moshele has joined #openstack-nova | 07:09 | |
*** hfu has joined #openstack-nova | 07:09 | |
*** kaisers_ has quit IRC | 07:10 | |
*** markvoelker has quit IRC | 07:10 | |
*** andreas_s has joined #openstack-nova | 07:10 | |
*** ducttape_ has joined #openstack-nova | 07:11 | |
*** jamielennox is now known as jamielennox|away | 07:11 | |
*** sileht has quit IRC | 07:13 | |
openstackgerrit | Bhagyashri Shewale proposed openstack/nova: Return 400 when name is more than 200 characters https://review.openstack.org/398998 | 07:13 |
*** thorst_ has joined #openstack-nova | 07:13 | |
openstackgerrit | Andia proposed openstack/nova: quotas of instance in soft-delete vm_state https://review.openstack.org/407877 | 07:14 |
*** sridharg has joined #openstack-nova | 07:15 | |
*** liuyulong_ has joined #openstack-nova | 07:15 | |
*** ducttape_ has quit IRC | 07:16 | |
*** mjura has quit IRC | 07:16 | |
*** carthaca_ has joined #openstack-nova | 07:17 | |
*** mjura has joined #openstack-nova | 07:17 | |
*** liuyulong__ has quit IRC | 07:18 | |
*** fragatina has joined #openstack-nova | 07:18 | |
*** thorst_ has quit IRC | 07:19 | |
openstackgerrit | Markus Zoeller (markus_z) proposed openstack/nova: live-migration test hook for serial console https://review.openstack.org/347471 | 07:21 |
*** jaosorior has quit IRC | 07:22 | |
*** jaosorior has joined #openstack-nova | 07:23 | |
*** zhangjl1 has joined #openstack-nova | 07:27 | |
*** zhangjl has quit IRC | 07:28 | |
tdurakov | markus_z: overall looks good^, let's just wait for ci | 07:29 |
*** nkrinner_afk is now known as nkrinner | 07:30 | |
*** zhangjl1 has quit IRC | 07:31 | |
*** zhangjl has joined #openstack-nova | 07:32 | |
*** rcernin has joined #openstack-nova | 07:34 | |
*** edand has joined #openstack-nova | 07:35 | |
openstackgerrit | Zhenyu Zheng proposed openstack/python-novaclient: Check source_type and destination_type when booting with bdm provided https://review.openstack.org/402364 | 07:35 |
*** voelzmo has joined #openstack-nova | 07:39 | |
*** yamahata has joined #openstack-nova | 07:41 | |
*** pcaruana has joined #openstack-nova | 07:41 | |
*** voelzmo has quit IRC | 07:44 | |
*** voelzmo has joined #openstack-nova | 07:44 | |
*** voelzmo has quit IRC | 07:45 | |
*** voelzmo has joined #openstack-nova | 07:46 | |
*** voelzmo has quit IRC | 07:48 | |
*** ayogi has quit IRC | 07:49 | |
*** voelzmo has joined #openstack-nova | 07:49 | |
*** hfu has quit IRC | 07:50 | |
*** mvk has quit IRC | 07:51 | |
*** code-R has quit IRC | 07:51 | |
*** hfu has joined #openstack-nova | 07:51 | |
gmann | alex_xu: but, boot is all ok there it just cinder does not tag volume as in-use, may be cinder side issue ? | 07:52 |
*** gszasz has joined #openstack-nova | 07:53 | |
*** rybridges2 has quit IRC | 07:54 | |
*** mfisch has quit IRC | 07:54 | |
*** kaisers has quit IRC | 07:54 | |
*** voelzmo has quit IRC | 07:55 | |
*** voelzmo has joined #openstack-nova | 07:55 | |
*** kaisers has joined #openstack-nova | 07:55 | |
alex_xu | gmann: I feel the commit message isn't very clear. The BDM record is created in the API layer. But in the nova-compute, the BDM record with destination_type=None will be ignored totally | 07:55 |
alex_xu | gmann: so nova didn't talk with cinder at all. the logic of nova is just ignored the BDM record. So I think it is failed sliently | 07:56 |
*** kristian__ has quit IRC | 07:56 | |
*** fragatina has quit IRC | 07:58 | |
*** kristian__ has joined #openstack-nova | 07:59 | |
zigo | mriedem: FYI, lxd isn't in Debian (yet), therefore Nova in Debian can't support lxd. I don't do Nova in Ubuntu, I only maintain packages in Debian. | 08:00 |
*** moshele has quit IRC | 08:01 | |
*** mriedem has quit IRC | 08:01 | |
*** voelzmo has quit IRC | 08:01 | |
*** ratailor has quit IRC | 08:02 | |
*** ratailor has joined #openstack-nova | 08:02 | |
gmann | alex_xu: ohh | 08:03 |
*** wingwj has quit IRC | 08:03 | |
*** voelzmo has joined #openstack-nova | 08:04 | |
*** irenab has quit IRC | 08:04 | |
*** irenab has joined #openstack-nova | 08:05 | |
*** ihrachys has joined #openstack-nova | 08:05 | |
*** sapcc-bot has quit IRC | 08:06 | |
*** sapcc-bot has joined #openstack-nova | 08:06 | |
*** markvoelker has joined #openstack-nova | 08:07 | |
*** kaisers_ has joined #openstack-nova | 08:07 | |
*** rybridges2 has joined #openstack-nova | 08:07 | |
gmann | alex_xu: so you mean server would not boot in that case? | 08:08 |
alex_xu | gmann: server will boot, just ignore that bdm record in the compute node, and won't create any real disk for that bdm | 08:09 |
*** wingwj has joined #openstack-nova | 08:09 | |
gmann | alex_xu: yea, that was my understanding. | 08:09 |
gmann | alex_xu: so in that case user will be 400 with that change | 08:10 |
*** gszasz has quit IRC | 08:10 | |
gmann | alex_xu: i got your point that it fail silently but server still accessible | 08:10 |
*** gszasz has joined #openstack-nova | 08:10 | |
alex_xu | gmann: yeah | 08:11 |
openstackgerrit | Takashi NATSUME proposed openstack/python-novaclient: WIP: Microversion 2.39 - Enable cold migration woth target host https://review.openstack.org/406707 | 08:11 |
gmann | alex_xu: thats why i was thinking it is backward incompatible may be.... | 08:11 |
*** markvoelker has quit IRC | 08:11 | |
gmann | alex_xu: because in api doc also we mentioned it is optional | 08:11 |
alex_xu | gmann: if remember correct, it is ignored at https://github.com/openstack/nova/blob/master/nova/virt/driver.py#L58 | 08:12 |
*** ayogi has joined #openstack-nova | 08:12 | |
*** owalsh has joined #openstack-nova | 08:13 | |
alex_xu | gmann: when it isn't specified, it is None? | 08:14 |
gmann | alex_xu: yea - it comes from here at APi layer - https://github.com/openstack/nova/blob/master/nova/block_device.py#L185 | 08:15 |
*** thorst_ has joined #openstack-nova | 08:16 | |
openstackgerrit | Zhenyu Zheng proposed openstack/nova: Don't apply multi-queue to SRIOV ports https://review.openstack.org/397545 | 08:16 |
*** xiaoyang has joined #openstack-nova | 08:16 | |
*** Jeffrey4l has quit IRC | 08:17 | |
*** lpetrut has quit IRC | 08:17 | |
*** Jeffrey4l has joined #openstack-nova | 08:17 | |
alex_xu | gmann: check the guide http://docs.openstack.org/developer/nova/api_microversion_dev.html it sounds a bug | 08:17 |
*** irenab has quit IRC | 08:18 | |
*** stvnoyes has quit IRC | 08:18 | |
*** hfu has quit IRC | 08:18 | |
alex_xu | gmann: maybe worth bring it up in the api meeting today, then double check with johnthetubaguy and sdague | 08:18 |
*** irenab has joined #openstack-nova | 08:18 | |
*** zul has quit IRC | 08:18 | |
*** stvnoyes has joined #openstack-nova | 08:19 | |
gmann | alex_xu: yea, it will be more clear from johnthetubaguy sdague feedback | 08:19 |
gmann | alex_xu: thanks for your input also | 08:20 |
alex_xu | gmann: +1 | 08:20 |
alex_xu | gmann: np | 08:20 |
*** gszasz has quit IRC | 08:20 | |
*** hfu has joined #openstack-nova | 08:20 | |
*** magic has quit IRC | 08:20 | |
gmann | alex_xu: i noted today meeting, last week i just forgot due to holiday here :) | 08:20 |
alex_xu | gmann: heh :) | 08:21 |
*** code-R has joined #openstack-nova | 08:21 | |
*** jed56 has joined #openstack-nova | 08:21 | |
*** shangxdy has quit IRC | 08:22 | |
openstackgerrit | Oleg Bondarev proposed openstack/nova: Live migration: wait for vif-plugged event on pre live migration https://review.openstack.org/246910 | 08:22 |
*** magic has joined #openstack-nova | 08:23 | |
*** sach-m has quit IRC | 08:23 | |
*** thorst_ has quit IRC | 08:23 | |
openstackgerrit | Takashi NATSUME proposed openstack/python-novaclient: WIP: Microversion 2.39 - Enable cold migration woth target host https://review.openstack.org/406707 | 08:24 |
*** pbandark has quit IRC | 08:25 | |
*** mvk has joined #openstack-nova | 08:26 | |
*** sileht has joined #openstack-nova | 08:26 | |
*** xiaoyang has quit IRC | 08:26 | |
*** slaweq_ has quit IRC | 08:27 | |
*** lpetrut has joined #openstack-nova | 08:27 | |
*** amoralej|off is now known as amoralej | 08:29 | |
openstackgerrit | Zhenyu Zheng proposed openstack/nova: Check destination_type when booting with bdm provided https://review.openstack.org/402372 | 08:30 |
*** mingyu has quit IRC | 08:31 | |
*** zul has joined #openstack-nova | 08:32 | |
*** slaweq_ has joined #openstack-nova | 08:32 | |
*** kristia__ has joined #openstack-nova | 08:34 | |
*** sach-m has joined #openstack-nova | 08:35 | |
*** kristian__ has quit IRC | 08:35 | |
*** gszasz has joined #openstack-nova | 08:39 | |
*** shaohe_feng has quit IRC | 08:41 | |
*** eliqiao_ has quit IRC | 08:41 | |
*** Yingxin has quit IRC | 08:41 | |
*** ducttape_ has joined #openstack-nova | 08:41 | |
*** roeyc has joined #openstack-nova | 08:42 | |
*** mkoderer has joined #openstack-nova | 08:42 | |
*** hfu has quit IRC | 08:45 | |
openstackgerrit | Takashi NATSUME proposed openstack/nova: Fix misuse of assertTrue https://review.openstack.org/407935 | 08:45 |
*** josecastroleon has quit IRC | 08:45 | |
*** ducttape_ has quit IRC | 08:46 | |
*** lpetrut has quit IRC | 08:46 | |
*** hfu has joined #openstack-nova | 08:46 | |
*** Yingxin has joined #openstack-nova | 08:47 | |
openstackgerrit | xianming.mao proposed openstack/nova: Wrap attachment operation with the try-except https://review.openstack.org/407828 | 08:48 |
pkholkin | alex_xu: thanks for review! | 08:50 |
alex_xu | pkholkin: np, thanks for patch also! | 08:50 |
*** sorrison_laptop has joined #openstack-nova | 08:51 | |
pkholkin | alex_xu: np! will wait Matt for review, too | 08:51 |
alex_xu | pkholkin: cool | 08:51 |
*** shaohe_feng has joined #openstack-nova | 08:52 | |
*** ralonsoh has joined #openstack-nova | 08:53 | |
*** shangxdy has joined #openstack-nova | 08:54 | |
*** josecastroleon has joined #openstack-nova | 08:57 | |
*** hfu has quit IRC | 08:57 | |
*** hfu has joined #openstack-nova | 08:58 | |
*** daemontool_ has quit IRC | 08:59 | |
*** zzzeek has quit IRC | 09:00 | |
*** sorrison_laptop is now known as sorrison_ | 09:00 | |
*** zzzeek has joined #openstack-nova | 09:01 | |
*** code-R_ has joined #openstack-nova | 09:01 | |
*** mfisch has joined #openstack-nova | 09:01 | |
*** eliqiao has joined #openstack-nova | 09:01 | |
*** mfisch has quit IRC | 09:02 | |
*** mfisch has joined #openstack-nova | 09:02 | |
*** sapcc-bot has quit IRC | 09:02 | |
*** sapcc-bot has joined #openstack-nova | 09:02 | |
*** wingwj has quit IRC | 09:04 | |
*** code-R has quit IRC | 09:04 | |
*** markvoelker has joined #openstack-nova | 09:07 | |
alex_xu | pkholkin: do you have python-novaclient side patch ready? | 09:09 |
pkholkin | alex_xu: no, I worked only only nova patch yet | 09:10 |
*** yassine has joined #openstack-nova | 09:10 | |
pkholkin | I will start | 09:10 |
*** yassine is now known as Guest78372 | 09:11 | |
*** eliqiao has quit IRC | 09:11 | |
*** lpetrut has joined #openstack-nova | 09:11 | |
alex_xu | pkholkin: thanks, just want to add it to my review list. let me know if it is ready | 09:11 |
*** markvoelker has quit IRC | 09:12 | |
pkholkin | alex_xu: ok, thanks! | 09:12 |
bauzas | mdbooth: do you think we should wait for merging https://review.openstack.org/#/c/382024/7 as per your last comment ? | 09:16 |
*** zhangjl1 has joined #openstack-nova | 09:17 | |
*** sudipto has joined #openstack-nova | 09:19 | |
*** sudipto_ has joined #openstack-nova | 09:19 | |
*** eliqiao has joined #openstack-nova | 09:20 | |
openstackgerrit | Zhenyu Zheng proposed openstack/python-novaclient: Check source_type and destination_type when booting with bdm provided https://review.openstack.org/402364 | 09:20 |
*** zhangjl has quit IRC | 09:20 | |
*** eandersson has joined #openstack-nova | 09:22 | |
openstackgerrit | Hussain Chachuliya proposed openstack/nova: Fix 500 error while passing 4-byte unicode data https://review.openstack.org/407514 | 09:24 |
*** kristia__ has quit IRC | 09:25 | |
*** paul-carlton has joined #openstack-nova | 09:26 | |
*** moshele has joined #openstack-nova | 09:26 | |
*** wingwj has joined #openstack-nova | 09:27 | |
*** wingwj has quit IRC | 09:28 | |
*** eandersson has quit IRC | 09:28 | |
*** lpetrut has quit IRC | 09:28 | |
*** lpetrut has joined #openstack-nova | 09:28 | |
openstackgerrit | Andia proposed openstack/nova: quotas of instance in soft-delete vm_state https://review.openstack.org/407877 | 09:29 |
*** efoley has joined #openstack-nova | 09:31 | |
*** sudipto has quit IRC | 09:31 | |
*** sudipto_ has quit IRC | 09:31 | |
*** sam_nowitzki has joined #openstack-nova | 09:31 | |
*** lucas-afk is now known as lucasagomes | 09:32 | |
*** matrohon has joined #openstack-nova | 09:33 | |
*** mdnadeem has quit IRC | 09:33 | |
openstackgerrit | Takashi NATSUME proposed openstack/python-novaclient: Fix doc generation errors https://review.openstack.org/407968 | 09:39 |
*** derekh has joined #openstack-nova | 09:40 | |
*** rmart04 has joined #openstack-nova | 09:42 | |
openstackgerrit | jichenjc proposed openstack/nova: Catch VolumeEncryptionNotSupported during spawn https://review.openstack.org/406109 | 09:44 |
*** eliqiao has quit IRC | 09:49 | |
*** mdnadeem has joined #openstack-nova | 09:50 | |
openstackgerrit | Hironori Shiina proposed openstack/nova: ironic: Add soft power off support to ironic driver. https://review.openstack.org/407977 | 09:50 |
mdbooth | bauzas: No | 09:51 |
mdbooth | bauzas: We can address that when it lands | 09:51 |
*** tpatzig_ has joined #openstack-nova | 09:52 | |
*** dgonzalez_ has joined #openstack-nova | 09:52 | |
*** code-R_ has quit IRC | 09:52 | |
*** david_1 has joined #openstack-nova | 09:52 | |
*** tommy_ has joined #openstack-nova | 09:52 | |
*** databus23_ has joined #openstack-nova | 09:52 | |
*** code-R has joined #openstack-nova | 09:52 | |
*** tommy_ is now known as Guest4119 | 09:52 | |
*** sapcc-bot has quit IRC | 09:52 | |
*** sapcc-bot has joined #openstack-nova | 09:53 | |
*** tpatzig_ has quit IRC | 09:54 | |
*** dgonzalez_ has quit IRC | 09:54 | |
*** david_1 has quit IRC | 09:54 | |
*** Guest4119 has quit IRC | 09:54 | |
*** databus23_ has quit IRC | 09:54 | |
*** mvk has quit IRC | 09:54 | |
openstackgerrit | jichenjc proposed openstack/nova: Catch VolumeEncryptionNotSupported during spawn https://review.openstack.org/406109 | 09:55 |
*** mvk has joined #openstack-nova | 09:55 | |
*** GB21 has quit IRC | 09:55 | |
bauzas | mdbooth: okay, so W'ipping it | 09:55 |
*** eliqiao has joined #openstack-nova | 10:00 | |
*** code-R_ has joined #openstack-nova | 10:03 | |
*** hfu has quit IRC | 10:05 | |
*** kaisers_ has quit IRC | 10:05 | |
*** thorst_ has joined #openstack-nova | 10:06 | |
*** code-R has quit IRC | 10:06 | |
*** hfu has joined #openstack-nova | 10:06 | |
*** huhaoran has quit IRC | 10:07 | |
*** markvoelker has joined #openstack-nova | 10:08 | |
*** irenab has quit IRC | 10:09 | |
*** irenab has joined #openstack-nova | 10:09 | |
*** sorrison_ has quit IRC | 10:09 | |
*** mpaolino has joined #openstack-nova | 10:12 | |
*** ducttape_ has joined #openstack-nova | 10:12 | |
*** mpaolino has quit IRC | 10:12 | |
*** mpaolino has joined #openstack-nova | 10:12 | |
*** markvoelker has quit IRC | 10:13 | |
*** carthaca_ has quit IRC | 10:14 | |
*** sapcc-bot has quit IRC | 10:14 | |
*** carthaca_ has joined #openstack-nova | 10:14 | |
*** david_1 has joined #openstack-nova | 10:14 | |
*** sapcc-bot has joined #openstack-nova | 10:14 | |
*** tpatzig_ has joined #openstack-nova | 10:14 | |
*** dgonzalez_ has joined #openstack-nova | 10:14 | |
*** tommy_ has joined #openstack-nova | 10:14 | |
*** databus23_ has joined #openstack-nova | 10:14 | |
*** tommy_ is now known as Guest22291 | 10:14 | |
*** WillUK2010 has joined #openstack-nova | 10:15 | |
*** thorst_ has quit IRC | 10:15 | |
*** paul-carlton2 has quit IRC | 10:15 | |
*** tpatzig_ has quit IRC | 10:16 | |
*** databus23_ has quit IRC | 10:16 | |
*** Guest22291 has quit IRC | 10:16 | |
*** dgonzalez_ has quit IRC | 10:16 | |
*** hfu has quit IRC | 10:16 | |
*** david_1 has quit IRC | 10:16 | |
*** ducttape_ has quit IRC | 10:16 | |
*** hfu has joined #openstack-nova | 10:16 | |
*** tuanluong has quit IRC | 10:17 | |
mdbooth | ftersin: Hey, I just responded to your comment on https://review.openstack.org/#/c/333263/18 | 10:18 |
mdbooth | You're right and I missed the implications of that line in cache() for Lvm backends | 10:18 |
mdbooth | However, given that it was already broken, I don't want to complicate the series by attempting to fix it early | 10:18 |
mdbooth | It gets fixed later in the series anyway | 10:18 |
mdbooth | Yeah, handling swap the same way as ephemerals breaks swap, but it was already mostly hosed anyway. | 10:19 |
*** GB21 has joined #openstack-nova | 10:19 | |
mdbooth | At least it's now consistently hosed. | 10:19 |
*** eglynn has joined #openstack-nova | 10:23 | |
*** hfu has quit IRC | 10:23 | |
eglynn | bauzas: the placement API service is not intended to be exposed via a publicURL, amiright? | 10:25 |
eglynn | (context is whether the keystone endpoint for that API should have a publicURL attribute) | 10:25 |
bauzas | eglynn: well | 10:26 |
bauzas | eglynn: we are providing an URL for the service catalog using a service type named "placement" | 10:27 |
bauzas | eglynn: but yeah, end-users shouldn't call the placement API, just the 'compute' service | 10:27 |
eglynn | yep, and this service catalog entry can have 3 URLs right ... public, internal, admin | 10:28 |
kashyap | markus_z: When you're about, can you also review this (this was discussed at Barcelona Summit) - https://review.openstack.org/#/c/313568/ -- Plugin to setup libvirt/QEMU from tar releases | 10:28 |
bauzas | eglynn: that said, we're putting publicURL endpoint https://github.com/openstack-dev/devstack/blob/master/lib/placement#L130-L139 | 10:28 |
bauzas | eglynn: yup | 10:28 |
eglynn | so just wondering if the publicURL should be left unset if it's not intended to be used | 10:28 |
bauzas | eglynn: maybe | 10:29 |
bauzas | eglynn: but we can also use the policy | 10:29 |
bauzas | good point to be discussed with the scheduler team | 10:29 |
bauzas | eglynn: could you please write an email thread about that? | 10:29 |
kashyap | markus_z: No rush, though. (Myself currently occupied with some stuff.) | 10:29 |
bauzas | eglynn: /me needs to bail out | 10:29 |
eglynn | bauzas: thanks, o/ | 10:29 |
bauzas | FWIW, folks, I need to bail out for today | 10:30 |
openstackgerrit | Pawel Koniszewski proposed openstack/nova: Change exponential function to linear https://review.openstack.org/407518 | 10:30 |
bauzas | so good bye and thanks for the fish | 10:30 |
openstackgerrit | Pawel Koniszewski proposed openstack/nova: Move migration_downtime_steps to libvirt/migration https://review.openstack.org/407517 | 10:30 |
openstackgerrit | Pawel Koniszewski proposed openstack/nova: Mark live_migration_downtime_steps as deprecated for removal https://review.openstack.org/408002 | 10:30 |
eglynn | context for the question is the puppetry for the TripleO integration ... https://review.openstack.org/#/c/406300/10/manifests/keystone/auth_placement.pp | 10:30 |
markus_z | kashyap: yep, will review that in ~2h (our CI got hit, have to fix that first) | 10:30 |
*** DuncanT has quit IRC | 10:33 | |
*** josecastroleon has quit IRC | 10:33 | |
*** DuncanT has joined #openstack-nova | 10:33 | |
*** WillUK2010 has quit IRC | 10:34 | |
* kashyap nods | 10:34 | |
*** david-lyle_ has joined #openstack-nova | 10:35 | |
*** shangxdy has quit IRC | 10:37 | |
*** david-lyle has quit IRC | 10:37 | |
*** kristian__ has joined #openstack-nova | 10:38 | |
*** kristian__ has quit IRC | 10:39 | |
*** kristia__ has joined #openstack-nova | 10:40 | |
*** kristia__ has quit IRC | 10:40 | |
*** kristian__ has joined #openstack-nova | 10:40 | |
*** yufei has quit IRC | 10:41 | |
*** kristian__ has quit IRC | 10:41 | |
*** kristian__ has joined #openstack-nova | 10:41 | |
*** kristian__ has quit IRC | 10:42 | |
tdurakov | bauzas, johnthetubaguy, hi folks, could you please take a look at https://review.openstack.org/#/c/407602/ | 10:42 |
*** kristian__ has joined #openstack-nova | 10:42 | |
*** jaosorior has quit IRC | 10:42 | |
tdurakov | ah, need to read recent logs first( | 10:44 |
*** paul-carlton has quit IRC | 10:44 | |
*** yamahata has quit IRC | 10:45 | |
*** sudipto has joined #openstack-nova | 10:46 | |
*** nandal has quit IRC | 10:46 | |
*** sudipto_ has joined #openstack-nova | 10:46 | |
*** McNinja has joined #openstack-nova | 10:47 | |
*** paul-carlton has joined #openstack-nova | 10:47 | |
*** kristian__ has quit IRC | 10:47 | |
*** mgagne has joined #openstack-nova | 10:47 | |
*** mgagne is now known as Guest2615 | 10:47 | |
*** leifz has joined #openstack-nova | 10:48 | |
*** zz_dimtruck has joined #openstack-nova | 10:50 | |
*** zz_dimtruck is now known as dimtruck | 10:50 | |
*** kristian__ has joined #openstack-nova | 10:54 | |
*** kristia__ has joined #openstack-nova | 10:55 | |
*** ftersin_ has joined #openstack-nova | 10:59 | |
*** kristian__ has quit IRC | 10:59 | |
*** erhudy has quit IRC | 11:00 | |
*** erhudy has joined #openstack-nova | 11:00 | |
*** ildikov has quit IRC | 11:00 | |
*** ildikov has joined #openstack-nova | 11:00 | |
*** claudiub has quit IRC | 11:01 | |
openstackgerrit | Matthew Booth proposed openstack/nova: libvirt: Combine injection info in InjectionInfo https://review.openstack.org/337790 | 11:01 |
*** cdent has joined #openstack-nova | 11:02 | |
*** tbachman has joined #openstack-nova | 11:02 | |
ftersin_ | mdbooth: hi. i understand that ephemeral formatting is already broken for lvm, and it is better to use the same way for ephemeral and swap formatting | 11:03 |
mdbooth | Sorry I didn't spot it earlier, btw. | 11:03 |
mdbooth | I didn't understand what you were getting at. | 11:03 |
mdbooth | sfinucan: ^^^ just lost 2 +2s due to merge conflict, btw. Would you mind fast approving? | 11:05 |
sfinucan | mdbooth: Sure | 11:05 |
mdbooth | sfinucan: Thanks | 11:05 |
*** john-davidge has joined #openstack-nova | 11:05 | |
sfinucan | mdbooth: No +W though? | 11:06 |
mdbooth | Don't know if that was because of the merge conflict already? | 11:06 |
mdbooth | sfinucan: You could review and add your own if you like :) | 11:06 |
*** rfolco has joined #openstack-nova | 11:06 | |
mdbooth | ftersin: I'm just reordering those patches, btw, and I'm going to add the missing test you spotted. | 11:07 |
*** sdague has joined #openstack-nova | 11:08 | |
sfinucan | mdbooth: iirc, I think I did. Two +2s and no serious changes anyway so I'll assume the +W was intended :) | 11:08 |
mdbooth | sfinucan: Thanks | 11:08 |
mdbooth | ftersin: Are you ok with that, btw? | 11:10 |
mdbooth | I appreciate it's not a clear cut decision. | 11:10 |
*** GB21 has quit IRC | 11:10 | |
*** tbachman has quit IRC | 11:12 | |
*** carthaca_ has quit IRC | 11:12 | |
*** sapcc-bot has quit IRC | 11:12 | |
*** david_1 has joined #openstack-nova | 11:12 | |
*** dgonzalez_ has joined #openstack-nova | 11:12 | |
*** tpatzig_ has joined #openstack-nova | 11:12 | |
*** databus23_ has joined #openstack-nova | 11:12 | |
*** carthaca_1 has joined #openstack-nova | 11:12 | |
*** sapcc-bot2 has joined #openstack-nova | 11:12 | |
*** thorst_ has joined #openstack-nova | 11:13 | |
*** tbachman has joined #openstack-nova | 11:13 | |
*** int32bit has quit IRC | 11:13 | |
mdbooth | ftersin: Hmm, on second thoughts, I don't think it will be simple to re-order those 2 commits | 11:13 |
*** tpatzig_ has quit IRC | 11:14 | |
*** databus23_ has quit IRC | 11:14 | |
*** dgonzalez_ has quit IRC | 11:14 | |
*** david_1 has quit IRC | 11:14 | |
*** zhangjl1 has quit IRC | 11:14 | |
*** tommy_ has joined #openstack-nova | 11:15 | |
cdent | "ephemeral formatting" is a funny phrase, out of context | 11:15 |
*** tommy_ is now known as Guest99726 | 11:15 | |
mdbooth | cdent: Hehe | 11:15 |
ftersin_ | mdbooth: great, thanks for the test. as for reordering: i've got it now - you have to fix lvm to analyze image_id instead of ephemeral_size to get 'generating' flag correctly | 11:15 |
gcb | sfinucan: congratulations, so can you convert +1 to +2 on https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:master+topic:bp/goal-python35 :-) | 11:16 |
*** Guest99726 has quit IRC | 11:17 | |
ftersin_ | mdbooth: i understand that you want to minimize side works for the series, but the result looks weird for me: some compatibility is broken, some code which is no longer called is being refactored | 11:18 |
mdbooth | ftersin: Which code is no longer being called? | 11:18 |
ftersin_ | mdbooth: _create_swap | 11:18 |
mdbooth | Well it is still being called, just not for Lvm | 11:19 |
ftersin_ | and you extend its interface with a parameter needed for lvm only | 11:19 |
mdbooth | Well, that's true. But when we fix it it will be needed. | 11:20 |
mdbooth | Lvm *should* be calling that, but it isn't. | 11:20 |
mdbooth | All I've done is make swap and ephemeral the same. | 11:20 |
*** thorst_ has quit IRC | 11:20 | |
ftersin_ | i know it, it will be needed indeed. but i'm not sure about time range of 'when' :) | 11:20 |
mdbooth | Yeah, that's why I haven't wanted to hold up other patches :/ | 11:21 |
mdbooth | However, all the code's there, and we are getting through it. | 11:21 |
mdbooth | ftersin: Is Lvm-backed with swap but no ephemeral currently important to you? | 11:22 |
*** kristia__ has quit IRC | 11:23 | |
*** ociuhandu has joined #openstack-nova | 11:23 | |
*** gcb has quit IRC | 11:24 | |
*** GB21 has joined #openstack-nova | 11:24 | |
ftersin_ | mdbooth: in fact i do not care about lvm, but scaleio is important for me. And without that code working properly, i cannot get scaleio driver formatting these damned ephemerals | 11:25 |
ftersin_ | see https://review.openstack.org/#/c/407440/ | 11:25 |
mdbooth | ftersin: Looking | 11:25 |
*** sam_nowitzki has quit IRC | 11:25 | |
*** geekinutah has quit IRC | 11:26 | |
*** geekinutah has joined #openstack-nova | 11:26 | |
ftersin_ | mdbooth: However it could be left as 'known issue' and fixed later | 11:26 |
mdbooth | ftersin: Actually, that fix 'if target != base' is probably sane | 11:27 |
mdbooth | We could take that as a separate patch, no? | 11:27 |
ftersin_ | mdbooth: sure | 11:27 |
mdbooth | Would that address your problems with ScaleIO? | 11:28 |
*** ociuhandu has quit IRC | 11:28 | |
*** ociuhandu has joined #openstack-nova | 11:29 | |
ftersin_ | yep. the only question is time. if a right fix depends of an enormous series... | 11:29 |
mdbooth | Well that's just a little fix we could slip in at the beginning | 11:30 |
mdbooth | In fact, I'd be prepared to put it in just before the 2 | 11:30 |
ftersin_ | as a branch of your series, right? | 11:30 |
*** openstackgerrit has quit IRC | 11:32 | |
mdbooth | ftersin: Well if I put it in my series, it would currently be number 2 | 11:33 |
mdbooth | So hopefully it would be merged within a couple of days | 11:33 |
*** cdelatte has joined #openstack-nova | 11:34 | |
*** sudipto_ has quit IRC | 11:36 | |
*** sudipto has quit IRC | 11:36 | |
*** mpaolino_ has joined #openstack-nova | 11:41 | |
sfinucan | gcb: That's one https://review.openstack.org/#/c/319532 | 11:41 |
*** openstackgerrit has joined #openstack-nova | 11:42 | |
openstackgerrit | Chris Dent proposed openstack/nova: HTTP interface for resource providers by aggregates https://review.openstack.org/407742 | 11:42 |
*** ducttape_ has joined #openstack-nova | 11:42 | |
*** mpaolino has quit IRC | 11:43 | |
*** tbachman has quit IRC | 11:44 | |
ftersin_ | mdbooth: i'd be excellent, thanks | 11:45 |
*** code-R_ has quit IRC | 11:46 | |
mdbooth | ftersin: I'm going to try to incorporate your connect() calls in my ensure interface too, btw | 11:46 |
*** ducttape_ has quit IRC | 11:47 | |
*** smatzek has joined #openstack-nova | 11:54 | |
mdbooth | ftersin: Do you know if there's a bug for Lvm ephemerals already, btw? | 11:54 |
*** catintheroof has joined #openstack-nova | 11:55 | |
openstackgerrit | Chris Dent proposed openstack/nova-specs: Update the generic resource pools to reflect reality https://review.openstack.org/407562 | 11:55 |
*** sahid has joined #openstack-nova | 11:55 | |
openstackgerrit | Takashi NATSUME proposed openstack/python-novaclient: Microversion 2.39 - Enable cold migration woth target host https://review.openstack.org/406707 | 11:57 |
ftersin_ | mdbooth: i don't | 11:57 |
mdbooth | ftersin: Ok. I'll have a search and create one if I can't find it. | 11:57 |
openstackgerrit | int32bit proposed openstack/nova: Choose VMware datastore by the least provisioned space https://review.openstack.org/402632 | 11:57 |
openstackgerrit | Stephen Finucane proposed openstack/nova: conf: Improve help text for network options https://review.openstack.org/341203 | 11:57 |
openstackgerrit | Stephen Finucane proposed openstack/nova: conf: Deprecate all nova-net related opts https://review.openstack.org/395683 | 11:57 |
*** ociuhandu has quit IRC | 11:58 | |
openstackgerrit | Stephen Finucane proposed openstack/nova: conf: Deprecate all nova-net related opts https://review.openstack.org/395683 | 11:59 |
openstackgerrit | Stephen Finucane proposed openstack/nova: conf: Improve help text for network options https://review.openstack.org/341203 | 11:59 |
*** code-R has joined #openstack-nova | 12:00 | |
*** paul-carlton1 has joined #openstack-nova | 12:01 | |
*** kaisers_ has joined #openstack-nova | 12:01 | |
*** nicolasbock has joined #openstack-nova | 12:01 | |
*** zenoway has joined #openstack-nova | 12:03 | |
*** paul-carlton has quit IRC | 12:03 | |
*** zenoway has quit IRC | 12:03 | |
cdent | edleafe, bauzas: should we consider moving the scheduler meeting if it is going to be right in the middle of bauzas needing to do the school run? | 12:03 |
*** zenoway has joined #openstack-nova | 12:03 | |
mdbooth | ftersin: Hey, there might be a problem with that patch. | 12:03 |
mdbooth | Won't it cause ephemerals and swap to be overwritten in some cases? | 12:04 |
mdbooth | Maybe hard reboot? | 12:04 |
mdbooth | For ScaleIO, maybe cold migration? | 12:04 |
mdbooth | Wait, maybe not | 12:06 |
ftersin_ | mdbooth: scaleio is a shared block storage | 12:06 |
ftersin_ | its disks are accessible globally, from every host | 12:07 |
mdbooth | ftersin: Yeah, ignore me. That call is already protected by the other self.exists() check | 12:08 |
openstackgerrit | Sergey Nikitin proposed openstack/nova: Added constraints for 'hacking' lib for pep8 https://review.openstack.org/408055 | 12:08 |
*** paul-carlton1 has quit IRC | 12:09 | |
*** ratailor has quit IRC | 12:09 | |
*** dave-mccowan has joined #openstack-nova | 12:09 | |
*** markvoelker has joined #openstack-nova | 12:09 | |
*** gibi has quit IRC | 12:14 | |
*** markvoelker has quit IRC | 12:15 | |
*** irenab has quit IRC | 12:18 | |
*** irenab has joined #openstack-nova | 12:18 | |
*** nkrinner has quit IRC | 12:18 | |
*** thorst_ has joined #openstack-nova | 12:18 | |
openstackgerrit | Sergey Nikitin proposed openstack/nova: Added constraints for 'hacking' lib for pep8 https://review.openstack.org/408055 | 12:19 |
*** prateek has quit IRC | 12:19 | |
*** rtheis has joined #openstack-nova | 12:19 | |
*** nkrinner has joined #openstack-nova | 12:21 | |
openstackgerrit | Li Yingjun proposed openstack/nova: Adds view builders for hypervisors controller https://review.openstack.org/335282 | 12:22 |
*** thorst_ has quit IRC | 12:25 | |
*** baoli has joined #openstack-nova | 12:26 | |
*** baoli has quit IRC | 12:26 | |
*** edmondsw_ has joined #openstack-nova | 12:26 | |
*** baoli has joined #openstack-nova | 12:26 | |
*** nkrinner has quit IRC | 12:28 | |
*** oanson has quit IRC | 12:28 | |
*** GB21 has quit IRC | 12:28 | |
*** xiaoyang has joined #openstack-nova | 12:31 | |
*** catintheroof has quit IRC | 12:32 | |
*** jaosorior has joined #openstack-nova | 12:32 | |
*** catintheroof has joined #openstack-nova | 12:32 | |
*** bvanhav__ has joined #openstack-nova | 12:32 | |
*** magic has quit IRC | 12:33 | |
*** ociuhandu has joined #openstack-nova | 12:34 | |
*** magic has joined #openstack-nova | 12:34 | |
markus_z | wznoinsk: got a second? I have a CI related question | 12:36 |
*** xiaoyang has quit IRC | 12:36 | |
*** catintheroof has quit IRC | 12:37 | |
markus_z | Our s390x CI needs a custom flavor, none of the pre-configured flavors in Devstack are suitable for the tempest tests. Where should I trigger the creation of such a custom flavor in our CI? | 12:37 |
sdague | markus_z: local.sh ? | 12:38 |
sdague | or maybe with a devstack plugin | 12:38 |
sdague | you could run that at post-extra phase | 12:39 |
BobBall | I've used pre_test_hook for that sort of thing in the past | 12:39 |
markus_z | It's already possible in pre_test_hook? | 12:39 |
sdague | markus_z: if you are running devstack-gate, yeh, you can run arbitrary code there | 12:39 |
sdague | but, you might want to just bundle it into a devstack plugin so that people could more easily do it manually | 12:40 |
markus_z | sdague: yep, we're using that | 12:40 |
*** cdent has quit IRC | 12:40 | |
*** xiaoyang has joined #openstack-nova | 12:40 | |
*** takashin has left #openstack-nova | 12:41 | |
markus_z | sdague: OK, cool, then I use a short-term solution in pre_test_hook and a mid/long-term solution as plugin. | 12:41 |
*** nkrinner has joined #openstack-nova | 12:42 | |
*** portdirect has joined #openstack-nova | 12:42 | |
*** magic has quit IRC | 12:42 | |
markus_z | thanks | 12:43 |
*** magic has joined #openstack-nova | 12:43 | |
alex_xu | nova api subteam meeting is in 15 mins at #openstack-meeting-4 | 12:45 |
*** xiaoyang has quit IRC | 12:45 | |
*** thorst_ has joined #openstack-nova | 12:47 | |
*** lucasagomes is now known as lucas-hungry | 12:49 | |
openstackgerrit | Sergey Nikitin proposed openstack/nova: Revert "reduce pep8 requirements to just hacking" https://review.openstack.org/408069 | 12:49 |
*** gcb has joined #openstack-nova | 12:50 | |
*** edmondsw_ has quit IRC | 12:50 | |
*** baoli has quit IRC | 12:51 | |
*** catintheroof has joined #openstack-nova | 12:54 | |
*** kylek3h__ has quit IRC | 12:56 | |
sfinucan | markus_z: Bit late, but we do it in the tests themselves | 12:58 |
sfinucan | for intel-nfv-ci at least | 12:58 |
*** kevinz has quit IRC | 12:59 | |
markus_z | sfinucan: Do you have an example? The reason for the custom flavor is the image we built. It's bigger than the default cirros which doesn't run on s390x. | 12:59 |
*** jaosorior has quit IRC | 12:59 | |
*** gmann_ has joined #openstack-nova | 12:59 | |
sfinucan | markus_z: https://github.com/openstack/intel-nfv-ci-tests/blob/master/intel_nfv_ci_tests/tests/scenario/test_cpu_pinning.py#L118-L137 | 13:00 |
alex_xu | johnthetubaguy: sdague are you around for api meeting? | 13:00 |
*** lamt has quit IRC | 13:01 | |
markus_z | sfinucan: Ah, those are tests you run additionally to the existing ones in Tempest, right? | 13:01 |
*** voelzmo has quit IRC | 13:02 | |
sfinucan | markus_z: Ah, yes. Sorry - you mean all tests with a custom flavor? | 13:02 |
markus_z | yes, exactly | 13:02 |
markus_z | Something for `flavor_ref` in the tempest.conf | 13:02 |
markus_z | BobBall: Can you point me to your pre_test_hook solution please? | 13:03 |
sdague | alex_xu: I am, johnthetubaguy is on honeymoon iirc | 13:05 |
alex_xu | sdague: cool, we are at #openstack-meeting-4 | 13:06 |
sdague | sorry, apparently my hexxchat config didn't add -4 to my channel lists when I readded | 13:06 |
alex_xu | johnthetubaguy: enjoy your honeymoon :) | 13:06 |
*** tbachman has joined #openstack-nova | 13:06 | |
*** ducttape_ has joined #openstack-nova | 13:06 | |
tdurakov | sdague: hi, could you please approve https://review.openstack.org/#/c/407602/ | 13:06 |
sdague | tdurakov: +A | 13:07 |
*** efoley_ has joined #openstack-nova | 13:07 | |
tdurakov | thanks! | 13:07 |
tdurakov | markus_z: hope it will be last recheck for serial console hook) | 13:08 |
*** gibi has joined #openstack-nova | 13:08 | |
*** amoralej is now known as amoralej|lunch | 13:08 | |
*** efoley has quit IRC | 13:08 | |
*** mriedem has joined #openstack-nova | 13:09 | |
*** tblakes has joined #openstack-nova | 13:10 | |
*** jaosorior has joined #openstack-nova | 13:10 | |
markus_z | tdurakov: :) yeah, it was odd that grenade didn't even reach step 2 of the hooks | 13:11 |
*** markvoelker has joined #openstack-nova | 13:12 | |
*** code-R has quit IRC | 13:12 | |
openstackgerrit | Francesco Santoro proposed openstack/os-vif: os-vif: add vif_name to VIFVHostUser class https://review.openstack.org/390225 | 13:12 |
openstackgerrit | Francesco Santoro proposed openstack/os-vif: os-vif: add new port profiles to enable fast path vhostuser https://review.openstack.org/375533 | 13:12 |
*** kristian__ has joined #openstack-nova | 13:13 | |
*** catinthe_ has joined #openstack-nova | 13:14 | |
mriedem | jroll: https://review.openstack.org/#/c/407977/ | 13:16 |
mriedem | just fyi | 13:16 |
*** catintheroof has quit IRC | 13:16 | |
*** mdnadeem has quit IRC | 13:16 | |
*** markvoelker has quit IRC | 13:16 | |
*** heroux has joined #openstack-nova | 13:17 | |
*** voelzmo has joined #openstack-nova | 13:17 | |
jroll | mriedem: O_o | 13:17 |
*** kristian__ has quit IRC | 13:17 | |
mriedem | gonna have the same issue here https://review.openstack.org/#/c/403745/ | 13:17 |
*** jed56 has quit IRC | 13:17 | |
*** markvoelker has joined #openstack-nova | 13:19 | |
*** code-R has joined #openstack-nova | 13:19 | |
*** claudiub has joined #openstack-nova | 13:20 | |
*** matrohon has quit IRC | 13:21 | |
*** code-R has quit IRC | 13:23 | |
*** dinobot_ has joined #openstack-nova | 13:23 | |
*** ducttape_ has quit IRC | 13:23 | |
*** ducttape_ has joined #openstack-nova | 13:26 | |
*** cdent has joined #openstack-nova | 13:26 | |
*** kylek3h_away has joined #openstack-nova | 13:26 | |
*** hfu has joined #openstack-nova | 13:27 | |
*** code-R has joined #openstack-nova | 13:28 | |
*** kevinz has joined #openstack-nova | 13:28 | |
*** ayogi has quit IRC | 13:31 | |
*** code-R has quit IRC | 13:32 | |
BobBall | markus_z: Sorry - correction - you can't use pre-test-hook as it's before devstack is actually set up; I was mis-remembering. | 13:33 |
*** jaypipes has joined #openstack-nova | 13:33 | |
markus_z | BobBall: Yeah, the logs look like this is the behavior | 13:34 |
markus_z | But I can use the hook to write a proper local.sh, I guess | 13:35 |
BobBall | markus_z: potentially you could define your own gate_hook which sets DEVSTACK_GATE_TEMPEST=0 and runs devstack-vm-gate.sh then run tempest with a static command, since you know how you want to run it, you can build your tempest command from https://github.com/openstack-infra/devstack-gate/blob/master/devstack-vm-gate.sh#L760 | 13:35 |
mriedem | pkholkin: some small things inline here https://review.openstack.org/#/c/405461/ | 13:35 |
mriedem | pkholkin: but it's very close | 13:35 |
BobBall | Yeah - a local.sh should work I guess | 13:35 |
*** code-R has joined #openstack-nova | 13:35 | |
mriedem | BobBall: while you're here, has this started? https://blueprints.launchpad.net/nova/+spec/add-os-xenapi-library | 13:35 |
BobBall | mriedem: Absolutely | 13:36 |
mriedem | on the nova side? | 13:36 |
BobBall | mriedem: https://review.openstack.org/#/c/406059/ | 13:36 |
mriedem | nothing is linked to the bp | 13:36 |
pkholkin | mriedem: ops, just update api-ref) | 13:36 |
pkholkin | thanks! will look | 13:36 |
mriedem | BobBall: the bp link in that nova change is wrong | 13:36 |
mriedem | so it's not linked to the bp in LP | 13:36 |
BobBall | mriedem: Indeed. the 'add' is missing. Will comment on the change to that effect too. | 13:37 |
markus_z | BobBall: Thanks, I'm fine with a short-term solution and then I'll take a deeper dive into the CI world. | 13:37 |
mriedem | BobBall: i got it | 13:37 |
mriedem | pkholkin: it's a couple other small things too | 13:37 |
pkholkin | mriedem: yes, looking | 13:38 |
jaypipes | morning supernovas. | 13:38 |
*** sudipto has joined #openstack-nova | 13:38 | |
*** code-R has quit IRC | 13:38 | |
*** sudipto_ has joined #openstack-nova | 13:38 | |
BobBall | mriedem: We're planning to have a series of 3 changes in Nova; one to make use of os-xenapi (so it shows what the refactor is), one to remove the now-unused files and then one to change to use the new API. We figured reviews would be easier this way. Thoughts? | 13:39 |
mriedem | jaypipes: o/ | 13:39 |
BobBall | What about us normal non-super-but-still-novas? hi jaypipes :) | 13:40 |
jaypipes | BobBall: you're still a supernova, Bob :) | 13:40 |
* BobBall likes jaypipes | 13:40 | |
jaypipes | heh | 13:41 |
*** lamt has joined #openstack-nova | 13:41 | |
tbachman | hey folks — I’m doing some debugging on some sqlalchemy (unrelated project), and came across this patch: https://review.openstack.org/#/c/395253/ | 13:42 |
*** ducttape_ has quit IRC | 13:42 | |
tbachman | The commit message for the revert says “Currently causing nova periodic jobs to fail." | 13:42 |
tbachman | Is anyone familiar with this change, and what was failing in nova? | 13:42 |
*** ljxiash has joined #openstack-nova | 13:42 | |
* tbachman is also going to ask on openstack-osla | 13:43 | |
mriedem | tbachman: harlowja is your guy for that | 13:44 |
tbachman | mriedem: thx! | 13:44 |
mriedem | it was blocking a new release of oslo.context i think | 13:44 |
jaypipes | mdbooth: I presume you're working through rebasing your imagebackend series? :) | 13:44 |
mdbooth | jaypipes: Yep | 13:44 |
jaypipes | coolio :) | 13:44 |
mriedem | BobBall: seems ok, i don't really have a comment for how that's going to look | 13:45 |
*** kfarr has joined #openstack-nova | 13:45 | |
openstackgerrit | Zhenyu Zheng proposed openstack/nova: Don't apply multi-queue to SRIOV ports https://review.openstack.org/397545 | 13:46 |
jaypipes | edleafe, cdent: how about "member_of" as the name of the "resource providers in these aggregates" querystring param? | 13:46 |
cdent | jaypipes: will that need an 'aggregate' qualifier? | 13:47 |
cdent | naming is hard | 13:47 |
* cdent goes shopping | 13:47 | |
jaypipes | cdent: I don't think so... aggregates is just our group/membership bucket. | 13:47 |
cdent | true | 13:47 |
jaypipes | cdent: I don't imagine having another grouping mechanism. | 13:47 |
cdent | also true | 13:48 |
cdent | I'm happy with "member_of" | 13:48 |
* cdent waits for edleafe to wake up before making the change | 13:49 | |
*** rcernin has quit IRC | 13:49 | |
*** tbachman has quit IRC | 13:50 | |
*** tlian has joined #openstack-nova | 13:51 | |
*** rcernin has joined #openstack-nova | 13:51 | |
openstackgerrit | Ludovic Beliveau proposed openstack/nova: Fix nova-serialproxy when registering cli options https://review.openstack.org/407602 | 13:51 |
lbeliveau | sfinucan: sdague: ^ patch was in merge conflict | 13:52 |
sfinucan | lbeliveau: done | 13:53 |
*** tyrola has quit IRC | 13:53 | |
lbeliveau | sfinican: nevermind ... just a rebase, no need to do anything, my bad :) | 13:53 |
*** Guest2615 is now known as mgagne | 13:54 | |
*** mgagne has quit IRC | 13:54 | |
*** mgagne has joined #openstack-nova | 13:54 | |
*** ljxiash_ has joined #openstack-nova | 13:55 | |
*** shangxdy has joined #openstack-nova | 13:55 | |
openstackgerrit | Matthew Booth proposed openstack/nova: libvirt: Mock imagebackend template funcs in ImageBackendFixture https://review.openstack.org/400871 | 13:56 |
openstackgerrit | Diana Clarke proposed openstack/nova: Simple tenant usage pagination https://review.openstack.org/386093 | 13:57 |
*** ljxiash has quit IRC | 13:58 | |
*** lucas-hungry is now known as lucasagomes | 13:58 | |
*** ducttape_ has joined #openstack-nova | 13:59 | |
*** kfarr has quit IRC | 13:59 | |
*** bkopilov has quit IRC | 13:59 | |
*** links has quit IRC | 13:59 | |
*** bkopilov has joined #openstack-nova | 13:59 | |
*** bvanhav_ has joined #openstack-nova | 14:01 | |
*** bvanhav__ has quit IRC | 14:02 | |
*** catintheroof has joined #openstack-nova | 14:02 | |
*** prateek has joined #openstack-nova | 14:03 | |
*** lamt has quit IRC | 14:03 | |
*** catinthe_ has quit IRC | 14:04 | |
*** esberglu has joined #openstack-nova | 14:05 | |
edleafe | cdent: jaypipes: been awake; just getting the family out the door :) | 14:05 |
*** lamt has joined #openstack-nova | 14:06 | |
jaypipes | edleafe: no prolemo. | 14:06 |
edleafe | cdent: 'member_of' is fine. Big improvement | 14:06 |
jaypipes | or problemo.. | 14:06 |
jaypipes | cdent: cool. ship it. | 14:06 |
cdent | edleafe: cool, can do america | 14:06 |
pkholkin | mriedem: quick comment here, everywhere in the code I found self.assertFalse(mocked.called) pattern | 14:06 |
*** prateek_ has joined #openstack-nova | 14:06 | |
*** shangxdy has quit IRC | 14:06 | |
*** mdrabe has joined #openstack-nova | 14:07 | |
*** amoralej|lunch is now known as amoralej | 14:07 | |
*** mdnadeem has joined #openstack-nova | 14:07 | |
*** andymaier has joined #openstack-nova | 14:07 | |
*** andymaier_ has joined #openstack-nova | 14:07 | |
edleafe | cdent: while you're on that - this looks like something that really should be in an API utility: https://review.openstack.org/#/c/407742/2/nova/api/openstack/placement/handlers/resource_provider.py@186 | 14:08 |
*** andymaier_ has quit IRC | 14:08 | |
jaypipes | pkholkin: yes, because older versions of unittest.mock don't have an assert_not_called() method. | 14:08 |
jaypipes | bauzas: lemme know when you have some time to chat about your patch that adds the POST /requests thing | 14:08 |
pkholkin | jaypipes: so, this new one is better? | 14:08 |
edleafe | cdent: not a showstopper, of course, but a pattern to keep in mind | 14:08 |
cdent | edleafe: I was going to wait on that. YAGNI and all that. | 14:08 |
*** prateek has quit IRC | 14:08 | |
jaypipes | pkholkin: you mean the old one? | 14:08 |
openstackgerrit | Diana Clarke proposed openstack/python-novaclient: Microversion 2.39 - Simple tenant usage pagination https://review.openstack.org/407825 | 14:09 |
jaypipes | pkholkin: I always use self.assertFalse(mock_obj.called) | 14:09 |
edleafe | cdent: sure, but YAGNI could also mean "you ARE gonna need it" :) | 14:09 |
pkholkin | jaypipes: what do you prefer? | 14:09 |
pkholkin | hm | 14:09 |
pkholkin | Matt asked me to use assert_not_called() instead of that | 14:09 |
cdent | edleafe: I want to make sure that we don't enable complex query processing until we need complex (generic) query processing. | 14:09 |
cdent | Because we don't want complex queries except where explicitly necessary | 14:09 |
* cdent send edleafe to ingy | 14:10 | |
cdent | when we get there, we'll deal with it | 14:10 |
snikitin | mriedem: Hi! Now results of CI pep8 job are false positive for some cases because of bug https://bugs.launchpad.net/nova/+bug/1648054 | 14:11 |
openstack | Launchpad bug 1648054 in OpenStack Compute (nova) "Upper constraints for 'hacking' lib doesn't work for pep8" [Medium,In progress] - Assigned to Sergey Nikitin (snikitin) | 14:11 |
snikitin | I created a fix for it https://review.openstack.org/#/c/408069 but I'm not sure that this is a best solution. I wanted to consult with you. could you look, please? | 14:11 |
*** ducttape_ has quit IRC | 14:12 | |
jaypipes | pkholkin: go with whatever mriedem wants I guess :) | 14:12 |
gcb | mriedem: about https://blueprints.launchpad.net/nova/+spec/goal-python35 , does this BP include functional tests Python 3 porting ? | 14:12 |
pkholkin | jaypipes: agree ) | 14:14 |
mriedem | pkholkin: the error message from assert_not_called is probably more useful than 'False != True' | 14:14 |
mriedem | gcb: i think it's any and all things to make nova py3 compliant | 14:14 |
mriedem | except the mox > mock conversion which is a separate blueprint | 14:15 |
*** baoli has joined #openstack-nova | 14:15 | |
openstackgerrit | Stephen Finucane proposed openstack/nova: conf: move console opt to console group https://review.openstack.org/389647 | 14:16 |
openstackgerrit | Stephen Finucane proposed openstack/nova: console: Provide a ConsoleDriver interface https://review.openstack.org/392174 | 14:16 |
openstackgerrit | Stephen Finucane proposed openstack/nova: Use stevedore for console driver https://review.openstack.org/392173 | 14:16 |
gcb | mriedem, got it. I have been working on porting unittests to Python 3 - https://etherpad.openstack.org/p/nova_py35. so we can start functional tests porting at the same time ? | 14:17 |
*** parora has joined #openstack-nova | 14:17 | |
cdent | edleafe: did you see my comments earlier about perhaps moving the scheduler meeting? | 14:17 |
edleafe | cdent: yes | 14:17 |
mriedem | snikitin: you might want to ask sdague. i don't like the alternative of duplicating the version constraints in both test-requirements.txt and tox.ini, although i don't think the version range on hacking changes much | 14:17 |
edleafe | Should we discuss that at the next scheduler meeting? :) | 14:17 |
mriedem | and with wheels, installing from test-requirements isn't as painful as it used to be | 14:17 |
*** annegentle has joined #openstack-nova | 14:19 | |
*** prateek_ has quit IRC | 14:19 | |
*** prateek has joined #openstack-nova | 14:19 | |
*** diga has quit IRC | 14:20 | |
snikitin | mriedem ty! I will ping sdague | 14:20 |
cdent | edleafe: I was going to put it on the agenda, and had that paradox | 14:20 |
cdent | so resigned myself to here | 14:20 |
*** jbernard has quit IRC | 14:21 | |
edleafe | cdent: well, since bauzas is the impetus for the move, and he's gone for the day, maybe discuss it tomorrow? | 14:21 |
*** mtanino has joined #openstack-nova | 14:22 | |
*** parora has quit IRC | 14:22 | |
cdent | ah, I did not know he was gone for the day | 14:22 |
openstackgerrit | int32bit proposed openstack/nova: Choose VMware datastore by the least provisioned space https://review.openstack.org/402632 | 14:22 |
cdent | bouncers-- | 14:22 |
*** READ10 has joined #openstack-nova | 14:23 | |
*** gcb has quit IRC | 14:24 | |
sdague | snikitin: where is the patch in question? | 14:25 |
*** jbernard has joined #openstack-nova | 14:26 | |
openstackgerrit | Pavel Kholkin proposed openstack/nova: [proxy-api] microversion 2.39 deprecates image-metadata proxy API https://review.openstack.org/405461 | 14:26 |
pkholkin | mriedem: ok, pushed new patch set with fixes ^ | 14:26 |
*** mriedem has quit IRC | 14:28 | |
*** krtaylor has quit IRC | 14:28 | |
*** prateek_ has joined #openstack-nova | 14:29 | |
*** krtaylor has joined #openstack-nova | 14:29 | |
*** ljxiash_ has quit IRC | 14:29 | |
*** ducttape_ has joined #openstack-nova | 14:30 | |
*** Jeffrey4l has quit IRC | 14:30 | |
snikitin | sdague https://review.openstack.org/#/c/408069/ here | 14:30 |
*** prateek has quit IRC | 14:30 | |
*** Jeffrey4l has joined #openstack-nova | 14:31 | |
*** ljxiash has joined #openstack-nova | 14:31 | |
*** matrohon has joined #openstack-nova | 14:31 | |
openstackgerrit | Chris Dent proposed openstack/nova: rt: pass the nodename to public methods https://review.openstack.org/398470 | 14:31 |
openstackgerrit | Chris Dent proposed openstack/nova: HTTP interface for resource providers by aggregates https://review.openstack.org/407742 | 14:31 |
openstackgerrit | Chris Dent proposed openstack/nova: rt: remove fluff from test_resource_tracker https://review.openstack.org/398471 | 14:31 |
openstackgerrit | Chris Dent proposed openstack/nova: rt: pass nodename to internal methods https://review.openstack.org/398469 | 14:31 |
openstackgerrit | Chris Dent proposed openstack/nova: placement: create aggregate map in report client https://review.openstack.org/407309 | 14:31 |
openstackgerrit | Chris Dent proposed openstack/nova: Support filtering resource providers by aggregate membership https://review.openstack.org/407629 | 14:31 |
openstackgerrit | Chris Dent proposed openstack/nova: placement: correct improper test case inheritance https://review.openstack.org/406363 | 14:31 |
openstackgerrit | Chris Dent proposed openstack/nova: rt: explicitly pass compute node to _update() https://review.openstack.org/398472 | 14:31 |
openstackgerrit | Chris Dent proposed openstack/nova: rt: use a single ResourceTracker object instance https://review.openstack.org/398473 | 14:31 |
cdent | sigh, caught by the automatic rebase again | 14:32 |
*** ducttape_ has quit IRC | 14:32 | |
*** ljxiash has quit IRC | 14:33 | |
*** ljxiash has joined #openstack-nova | 14:33 | |
*** eharney has joined #openstack-nova | 14:34 | |
openstackgerrit | Francesco Santoro proposed openstack/nova: [WIP] convert libvirt driver to use os-vif for fast path vhostuser https://review.openstack.org/385061 | 14:36 |
*** smatzek has quit IRC | 14:36 | |
openstackgerrit | Feodor Tersin proposed openstack/nova: Implement ScaleIO image backend https://review.openstack.org/407440 | 14:36 |
*** gouthamr has joined #openstack-nova | 14:37 | |
*** ljxiash has quit IRC | 14:38 | |
sdague | snikitin: ah | 14:39 |
*** shangxdy has joined #openstack-nova | 14:39 | |
sdague | it would kind of suck to have to install everything in there to get contraints only | 14:39 |
*** parora has joined #openstack-nova | 14:39 | |
*** hfu has quit IRC | 14:40 | |
snikitin | sdague, agree. That is why I need an advice | 14:41 |
sdague | so, this doesn't change that often. I'd personally be fine with keeping the version pin in tox.ini | 14:41 |
*** matrohon has quit IRC | 14:41 | |
sdague | hacking<0.11,>=0.10.0 | 14:42 |
sdague | we're really changing that at most once a cycle | 14:42 |
*** prateek_ has quit IRC | 14:42 | |
*** hfu has joined #openstack-nova | 14:43 | |
*** jaosorior has quit IRC | 14:43 | |
*** prateek_ has joined #openstack-nova | 14:43 | |
*** sridharg has quit IRC | 14:44 | |
tdurakov | sdague: so in case of hacking version updated by the bot, we need a manual follow up patch to update it in tox.ini too, right? | 14:44 |
sdague | tdurakov: sure | 14:44 |
*** kaisers has quit IRC | 14:45 | |
snikitin | sdague, mriedem is not happy with this solution :( He said "i don't like the alternative of duplicating the version constraints in both test-requirements.txt and tox.ini" | 14:45 |
openstackgerrit | Dan Smith proposed openstack/nova: Make servers api view load instance fault from proper cell https://review.openstack.org/407238 | 14:45 |
openstackgerrit | Dan Smith proposed openstack/nova: Add schedule_and_build_instances conductor method https://review.openstack.org/367557 | 14:45 |
openstackgerrit | Dan Smith proposed openstack/nova: WIP Multicell support for instance listing https://review.openstack.org/396775 | 14:45 |
openstackgerrit | Dan Smith proposed openstack/nova: Make RPCFixture support multiple connections https://review.openstack.org/396417 | 14:45 |
openstackgerrit | Dan Smith proposed openstack/nova: Setup CellsV2 environment in base test https://review.openstack.org/406380 | 14:45 |
openstackgerrit | Dan Smith proposed openstack/nova: Add SecurityGroup.identifier to prefer uuid over name https://review.openstack.org/406260 | 14:45 |
openstackgerrit | Dan Smith proposed openstack/nova: Add SingleCellSimple fixture https://review.openstack.org/405684 | 14:45 |
openstackgerrit | Dan Smith proposed openstack/nova: WIP Move instance creation to conductor https://review.openstack.org/319379 | 14:45 |
openstackgerrit | Dan Smith proposed openstack/nova: Fix up non-cells-aware context managers in test_db_api https://review.openstack.org/406379 | 14:45 |
*** eharney has quit IRC | 14:46 | |
coreycb | hi all, is placement api valid to specify on enabled_apis in nova.conf? | 14:46 |
*** kaisers has joined #openstack-nova | 14:46 | |
*** parora has quit IRC | 14:47 | |
openstackgerrit | Vladik Romanovsky proposed openstack/nova: tests: avoid starting compute service twice in sriov functional test https://review.openstack.org/407572 | 14:47 |
openstackgerrit | Vladik Romanovsky proposed openstack/nova: tests: generate correct pci addresses for fake pci devices https://review.openstack.org/407730 | 14:47 |
openstackgerrit | Dan Smith proposed openstack/nova: Make servers api view load instance fault from proper cell https://review.openstack.org/407238 | 14:47 |
openstackgerrit | Dan Smith proposed openstack/nova: Add schedule_and_build_instances conductor method https://review.openstack.org/367557 | 14:47 |
*** ducttape_ has joined #openstack-nova | 14:47 | |
openstackgerrit | Dan Smith proposed openstack/nova: WIP Multicell support for instance listing https://review.openstack.org/396775 | 14:47 |
openstackgerrit | Dan Smith proposed openstack/nova: Make RPCFixture support multiple connections https://review.openstack.org/396417 | 14:47 |
openstackgerrit | Dan Smith proposed openstack/nova: Setup CellsV2 environment in base test https://review.openstack.org/406380 | 14:47 |
openstackgerrit | Dan Smith proposed openstack/nova: Add SecurityGroup.identifier to prefer uuid over name https://review.openstack.org/406260 | 14:47 |
openstackgerrit | Dan Smith proposed openstack/nova: Add SingleCellSimple fixture https://review.openstack.org/405684 | 14:47 |
openstackgerrit | Dan Smith proposed openstack/nova: WIP Move instance creation to conductor https://review.openstack.org/319379 | 14:47 |
openstackgerrit | Dan Smith proposed openstack/nova: Fix up non-cells-aware context managers in test_db_api https://review.openstack.org/406379 | 14:47 |
openstackgerrit | Evgeny Antyshev proposed openstack/nova: WIP: Convert volume format name from 'parallels' to 'ploop' https://review.openstack.org/408149 | 14:47 |
*** markus_z has quit IRC | 14:49 | |
*** shangxdy has quit IRC | 14:50 | |
tdurakov | sdague, snikitin: let's compare time between installing all test-requirements and hacking only, is that profit that valuable to worry about it? | 14:52 |
openstackgerrit | Roman Podoliaka proposed openstack/nova-specs: WIP: Ironic: add support for instance RAID configuration https://review.openstack.org/408151 | 14:52 |
*** rama_y has joined #openstack-nova | 14:53 | |
openstackgerrit | Matthew Booth proposed openstack/nova: libvirt: Fix initialising of LVM ephemeral disks https://review.openstack.org/408152 | 14:53 |
mdbooth | ftersin: ^^^ | 14:53 |
*** jamespag` is now known as jamespage | 14:53 | |
*** tongli has joined #openstack-nova | 14:53 | |
*** cdent has quit IRC | 14:55 | |
*** mriedem has joined #openstack-nova | 14:55 | |
snikitin | sdague tdurakov: installing of 'hacking' takes 2 sec. installting of test-req.txt takes 35 sec | 14:56 |
*** irenab has quit IRC | 14:56 | |
openstackgerrit | Ed Leafe proposed openstack/nova: Cascade deletes of RP aggregate associations https://review.openstack.org/407707 | 14:56 |
snikitin | sdague, tdurakov I restored 'fast' solution. so you can comment is too https://review.openstack.org/#/c/408055/ | 14:56 |
*** smatzek has joined #openstack-nova | 14:57 | |
edleafe | cdent: ^^ tests added. Have at it. | 14:57 |
*** hfu has quit IRC | 14:57 | |
*** jamesdenton has joined #openstack-nova | 14:57 | |
openstackgerrit | Matthew Booth proposed openstack/nova: libvirt: Add create_from_image and create_from_func to Backend https://review.openstack.org/333244 | 14:58 |
*** markus_z has joined #openstack-nova | 14:58 | |
*** irenab has joined #openstack-nova | 14:59 | |
sdague | snikitin: oh, so you were going to change deps to just test-requirements, and not install requirements? | 15:01 |
sdague | yeh, that's probably not too bad | 15:01 |
sdague | it's when numpy gets installed in requirements.txt that's really terrible | 15:01 |
*** links has joined #openstack-nova | 15:03 | |
moshele | jaypipes:hi kindly reminder to review https://review.openstack.org/#/c/199488/ | 15:04 |
moshele | jaypipes: hi kindly reminder to review https://review.openstack.org/#/c/199488/ | 15:04 |
*** voelzmo has quit IRC | 15:04 | |
*** liverpooler has quit IRC | 15:05 | |
*** liverpooler has joined #openstack-nova | 15:05 | |
*** irenab has quit IRC | 15:06 | |
*** tbachman has joined #openstack-nova | 15:07 | |
*** voelzmo_ has joined #openstack-nova | 15:07 | |
*** Amit_M has quit IRC | 15:07 | |
*** hongbin has joined #openstack-nova | 15:07 | |
*** rama_y has quit IRC | 15:09 | |
*** TMM has joined #openstack-nova | 15:11 | |
TMM | Hi all, can someone point me in the right direction for trying to debug "Filter NUMATopologyFilter returned 0 hosts" errors when trying to deploy flavors with numa node extra specs? | 15:11 |
TMM | I'm kind of lost | 15:12 |
mriedem | pkholkin: what do you think about replacing that negative logic with positive logic using max_version here? https://review.openstack.org/#/c/405461/7/nova/api/openstack/compute/create_backup.py@62 | 15:12 |
openstackgerrit | Eric Berglund proposed openstack/nova: Introduce PowerVM driver skeleton https://review.openstack.org/391288 | 15:12 |
openstackgerrit | Pavel Kholkin proposed openstack/python-novaclient: [proxy-api] microversion 2.39 deprecates image-metadata proxy API https://review.openstack.org/408166 | 15:12 |
mriedem | if request_version.is_supported(max_version='2.38'): | 15:12 |
mriedem | that seems cleaner to me | 15:12 |
pkholkin | mriedem: I had this idea, too | 15:13 |
mriedem | TMM: cfriesen might be able to help but he's not around it looks like | 15:13 |
pkholkin | alex_xu: mriedem: started novaclient patch for 2.39 https://review.openstack.org/#/c/408166/ | 15:14 |
dansmith | mriedem: thanks for pushing that stuff through last night | 15:14 |
dansmith | mriedem: I realized I forgot to remove the back from my patch anyway, which I've now done | 15:14 |
*** irenab has joined #openstack-nova | 15:14 | |
pkholkin | mriedem: I think I can change that for positive logic | 15:14 |
mriedem | s/back/hack/ right? | 15:14 |
mriedem | pkholkin: ok | 15:14 |
TMM | mriedem, thanks, I'll wait :) | 15:14 |
*** armax has joined #openstack-nova | 15:14 | |
*** mdrabe has quit IRC | 15:18 | |
snikitin | sdague, Yes. In fact by removing 'deps = hacking' we replace it with 'deps = -r{toxinidir}/test-requirements.txt' from [testenv] | 15:19 |
*** cdent has joined #openstack-nova | 15:20 | |
*** anush has quit IRC | 15:21 | |
*** irenab has quit IRC | 15:21 | |
sdague | snikitin: yeh, it still isn't great that we're installing postgresql c deps for that venv, but such it is I guess | 15:21 |
sdague | mriedem: why don't you want to just stick that one line in there? | 15:22 |
*** cdent_ has joined #openstack-nova | 15:22 | |
sdague | mriedem: this really is only changing about once a year now | 15:22 |
mriedem | i didn't -1 the idea | 15:22 |
mriedem | and said basically the same | 15:23 |
mriedem | but then you +2'ed the revert | 15:23 |
mriedem | so i went with that | 15:23 |
mriedem | can we move the postgresql dep out of test-requirements? | 15:23 |
mriedem | i guess it's needed for the opportunistic db tests | 15:23 |
*** daemontool has joined #openstack-nova | 15:23 | |
openstackgerrit | Pavel Kholkin proposed openstack/nova: [proxy-api] microversion 2.39 deprecates image-metadata proxy API https://review.openstack.org/405461 | 15:23 |
*** dave-mccowan has quit IRC | 15:23 | |
pkholkin | mriedem: done ^ | 15:23 |
pkholkin | also started novaclient patch | 15:23 |
sdague | mriedem: we can't, for those reasons | 15:23 |
mriedem | pkholkin: you'll be pleased to find my -1 on your novaclient patch already :)( | 15:24 |
pkholkin | mriedem: got comments | 15:24 |
pkholkin | yes :) | 15:24 |
mriedem | i'm +2 on https://review.openstack.org/#/c/405461/ now though, thanks | 15:24 |
pkholkin | mriedem: that's the first patch :) | 15:24 |
mriedem | really fast turnaround on that one is helpful | 15:24 |
pkholkin | sounds great! ) | 15:24 |
mriedem | alex_xu: are you still happy with this? https://review.openstack.org/#/c/405461/ | 15:25 |
mriedem | sdague: maybe we should use extras in setup.cfg more | 15:25 |
openstackgerrit | Sam Betts proposed openstack/nova: [WIP] Support Ironic interface attach/detach in nova virt https://review.openstack.org/364413 | 15:25 |
mriedem | but that would require changes to the gate jobs that run those i suppose | 15:25 |
*** cdent has quit IRC | 15:25 | |
*** cdent_ is now known as cdent | 15:25 | |
pkholkin | mriedem: will work on novaclient patch now | 15:25 |
mriedem | i've got to run to a meeting | 15:26 |
pkholkin | mriedem: after that I can do smth else | 15:26 |
pkholkin | ok | 15:26 |
*** catintheroof has quit IRC | 15:26 | |
*** catintheroof has joined #openstack-nova | 15:26 | |
snikitin | mriedem, sdague Oh, I forgot! we need to merge this patch first https://review.openstack.org/#/c/407870/ It fixes 'bad' assetEquals | 15:26 |
*** pradk has quit IRC | 15:26 | |
*** Jeffrey4l has quit IRC | 15:27 | |
*** Jeffrey4l has joined #openstack-nova | 15:27 | |
*** pradk has joined #openstack-nova | 15:27 | |
*** owalsh has quit IRC | 15:27 | |
*** rama_y has joined #openstack-nova | 15:27 | |
*** irenab has joined #openstack-nova | 15:28 | |
*** anush has joined #openstack-nova | 15:29 | |
*** mdnadeem has quit IRC | 15:30 | |
*** mdrabe has joined #openstack-nova | 15:30 | |
*** catintheroof has quit IRC | 15:31 | |
openstackgerrit | Matthew Booth proposed openstack/nova: libvirt: Pass is_block_dev explicitly from create_from_func https://review.openstack.org/333263 | 15:32 |
*** baoli has quit IRC | 15:32 | |
mdbooth | ftersin: ^^^ Updated with a test | 15:32 |
openstackgerrit | Diana Clarke proposed openstack/python-novaclient: Microversion 2.39 - Simple tenant usage pagination https://review.openstack.org/407825 | 15:32 |
*** shangxdy has joined #openstack-nova | 15:32 | |
*** annegentle has quit IRC | 15:33 | |
*** jwcroppe has joined #openstack-nova | 15:33 | |
*** Guest78372 has quit IRC | 15:35 | |
*** irenab has quit IRC | 15:35 | |
*** owalsh has joined #openstack-nova | 15:37 | |
*** yassine has joined #openstack-nova | 15:37 | |
*** moshele has quit IRC | 15:37 | |
*** yassine is now known as Guest78677 | 15:38 | |
*** portdirect is now known as portdirect_lunch | 15:39 | |
*** irenab has joined #openstack-nova | 15:39 | |
openstackgerrit | Matthew Booth proposed openstack/nova: libvirt: Refactor _create_image into storagepolicy https://review.openstack.org/356651 | 15:41 |
*** dave-mccowan has joined #openstack-nova | 15:41 | |
*** irenab has quit IRC | 15:42 | |
openstackgerrit | Matthew Booth proposed openstack/nova: libvirt: don't test cache() in test_driver https://review.openstack.org/406869 | 15:43 |
*** mlavalle has joined #openstack-nova | 15:45 | |
openstackgerrit | Matthew Booth proposed openstack/nova: libvirt: Introduce ImageCacheLocalDir https://review.openstack.org/279669 | 15:45 |
*** NightKhaos has quit IRC | 15:46 | |
*** annegentle has joined #openstack-nova | 15:46 | |
openstackgerrit | Matthew Booth proposed openstack/nova: libvirt: Resize an existing disk https://review.openstack.org/406870 | 15:47 |
dansmith | melwitt: mriedem: also thanks for the reviews | 15:48 |
*** NightKhaos has joined #openstack-nova | 15:48 | |
*** heroux has quit IRC | 15:48 | |
*** irenab has joined #openstack-nova | 15:49 | |
openstackgerrit | Matthew Booth proposed openstack/nova: libvirt: Add ensure_backing_from_* https://review.openstack.org/406871 | 15:50 |
*** hogepodge has joined #openstack-nova | 15:50 | |
*** Sukhdev_ has joined #openstack-nova | 15:51 | |
*** rama_y has quit IRC | 15:51 | |
openstackgerrit | Matthew Booth proposed openstack/nova: libvirt: Don't call _create_image from finish_migration https://review.openstack.org/337160 | 15:52 |
mdbooth | Looking for a second +2 at the head of my queue here, btw: https://review.openstack.org/#/c/400871/ | 15:52 |
*** portdirect_lunch is now known as portdirect | 15:52 | |
*** jheroux has joined #openstack-nova | 15:52 | |
mdbooth | That's the last test-only change before the real stuff starts | 15:52 |
*** zenoway has quit IRC | 15:53 | |
mriedem | i don't think untestably is a word | 15:53 |
mdbooth | mriedem: It is now | 15:53 |
mriedem | touche | 15:53 |
*** zenoway has joined #openstack-nova | 15:53 | |
* mdbooth scribbles it into his dictionary | 15:53 | |
openstackgerrit | Matthew Booth proposed openstack/nova: libvirt: Remove _create_image from driver https://review.openstack.org/406872 | 15:54 |
mdbooth | ^^^ One of my favourites | 15:54 |
*** kevinz has quit IRC | 15:55 | |
mdbooth | 2 files changed, 27 insertions(+), 92 deletions(-) | 15:55 |
*** jdurgin1 has joined #openstack-nova | 15:56 | |
ftersin_ | mdbooth: you need to rebase the series again :/ | 15:56 |
mdbooth | ftersin_: Really? | 15:57 |
*** ducttape_ has quit IRC | 15:57 | |
*** ducttape_ has joined #openstack-nova | 15:57 | |
* mdbooth slumps | 15:57 | |
*** catintheroof has joined #openstack-nova | 15:57 | |
*** irenab has quit IRC | 15:57 | |
mdbooth | ftersin_: Did you look at my Lvm ephemeral fix, btw? | 15:58 |
ftersin_ | as i see gerrit writes 'merge conflict'. i do see no reason to not trust it | 15:58 |
*** john-davidge has quit IRC | 15:58 | |
*** zenoway has quit IRC | 15:58 | |
mdbooth | I think it covers your ScaleIO use case, and addresses your concerns on the next 2 patches | 15:58 |
mriedem | mdbooth: question inline for a mock.ANY in there https://review.openstack.org/#/c/400871/ | 15:58 |
ftersin_ | i did. got stuck on test, being switched to something another... | 15:58 |
openstackgerrit | Stephen Finucane proposed openstack/nova: tests: Replace use of CONF with monkey patching https://review.openstack.org/408191 | 15:59 |
openstackgerrit | Stephen Finucane proposed openstack/nova: console: Move proxies to 'console/proxy' https://review.openstack.org/408192 | 15:59 |
openstackgerrit | Stephen Finucane proposed openstack/nova: conf: Deprecate 'console_driver' https://review.openstack.org/408193 | 15:59 |
dansmith | sdague: libvirt test and fixture fixes you would love to +W: https://review.openstack.org/#/c/407730/3 | 15:59 |
dansmith | sdague: one on top of that too | 15:59 |
sdague | dansmith: ok, let me look | 15:59 |
*** parora has joined #openstack-nova | 16:00 | |
*** diga has joined #openstack-nova | 16:00 | |
sdague | L191 here - https://review.openstack.org/#/c/407730/3/nova/tests/unit/virt/libvirt/fakelibvirt.py can't be right can it? That doesn't look like valid format | 16:00 |
*** jheroux has quit IRC | 16:00 | |
*** prateek_ has quit IRC | 16:01 | |
sdague | actually, in general, the new formats in the template seem... odd | 16:01 |
dansmith | sdague: it passes the test, so it has to be right? | 16:01 |
sdague | %(slot)02x doesn't seem like a thing | 16:01 |
*** prateek_ has joined #openstack-nova | 16:01 | |
dansmith | sdague: sfinucan made a point about having checked them, surprised they worked | 16:01 |
dansmith | sdague: sure it is, why not? | 16:01 |
dansmith | sdague: >>> '%(foo)02x' % {'foo': 1} | 16:02 |
dansmith | '01' | 16:02 |
sdague | oh... sorry | 16:02 |
*** voelzmo_ has quit IRC | 16:03 | |
*** mpaolino_ has quit IRC | 16:03 | |
sfinucan | sdague: Wonder if it would be a little easier to read if we used '.format' instead? | 16:03 |
*** tongli has quit IRC | 16:03 | |
dansmith | please no | 16:04 |
* dansmith hates .format | 16:04 | |
dansmith | :) | 16:04 |
*** gmann_ has quit IRC | 16:04 | |
*** parora has quit IRC | 16:04 | |
mdbooth | Isn't there a third string formatting convention in python 3 now? | 16:04 |
sdague | sfinucan: it might be, I think the advanced format strings aren't obvious at least to me, but now that I've got that in my head, I can look at the rest of it | 16:04 |
mdbooth | If they keep this up, I'lll start referring to it as perlython | 16:04 |
sfinucan | mdbooth: Careful - you'll end up sounding like Zed Shaw :P | 16:05 |
mdbooth | Heh, I read that | 16:05 |
mdbooth | I disagree with his reasons, but agree with his conclusions | 16:06 |
mdbooth | I think arguing that a language is bad because it's hard for beginners is stupid | 16:06 |
*** jheroux has joined #openstack-nova | 16:06 | |
mdbooth | It's like saying F1 drivers should all use Nissan Micras, because that's what they learned in | 16:06 |
markus_z | would make the races more interesting | 16:06 |
mdbooth | markus_z: lol | 16:07 |
sdague | sounds like a top gear episode | 16:07 |
dansmith | sdague: fyi, these two fixes are needed for me to make my cellsv2 change to the base test.py.. these started rando-failing when I made my change, so they're probably already unstable | 16:07 |
dansmith | mostly the second patch I think, but.. | 16:07 |
*** rama_y has joined #openstack-nova | 16:07 | |
sdague | ok | 16:07 |
*** irenab has joined #openstack-nova | 16:08 | |
*** edand has quit IRC | 16:08 | |
openstackgerrit | Diana Clarke proposed openstack/python-novaclient: Microversion 2.39 - Simple tenant usage pagination https://review.openstack.org/407825 | 16:09 |
*** prometheanfire has joined #openstack-nova | 16:09 | |
*** jgriffith_away is now known as jgriffith | 16:09 | |
sdague | dansmith: so... dev is still going to be potentially weird in that template right, as you can't get a 0xa | 16:09 |
prometheanfire | anyone use/test the zfs storage backend in libvirt with nova? https://libvirt.org/storage.html#StorageBackendZFS | 16:09 |
dansmith | sdague: not sure what you mean | 16:09 |
sdague | 0x%(dev)d | 16:10 |
mdbooth | prometheanfire: I doubt Nova supports it | 16:10 |
dansmith | sdague: oh you mean if we had a large device number and didn't actually output hex/ | 16:10 |
prometheanfire | mdbooth: me too, just sounds 'neat' is all (since right now I'm doing nova-compute on zfs) | 16:11 |
sdague | dansmith: if dev is passed into the function > 9 the template will be all wrong | 16:11 |
prometheanfire | it'd be nice not to double-cow... | 16:11 |
sdague | as it does similar conversions to slots | 16:11 |
dansmith | sdague: that's unchanged in this patch so I dunno what the impact is | 16:11 |
*** rmart04 has quit IRC | 16:11 | |
sdague | dansmith: right, agreed | 16:11 |
sdague | just making sure we note it somewhere while staring at this | 16:11 |
dansmith | vladikr: ^ | 16:11 |
*** irenab has quit IRC | 16:12 | |
dansmith | definitely feels wrong to format something with 0x%d for sure | 16:12 |
*** lpetrut has quit IRC | 16:12 | |
vladikr | dansmith, on a call, just a sec | 16:12 |
*** armax has quit IRC | 16:12 | |
*** lpetrut has joined #openstack-nova | 16:12 | |
*** eharney has joined #openstack-nova | 16:13 | |
*** gszasz has quit IRC | 16:13 | |
*** links has quit IRC | 16:13 | |
*** armax has joined #openstack-nova | 16:14 | |
*** ihrachys has quit IRC | 16:15 | |
*** cfriesen has joined #openstack-nova | 16:16 | |
*** kaisers_ has quit IRC | 16:16 | |
*** armax has quit IRC | 16:17 | |
ftersin_ | mdbooth: i commented the test on https://review.openstack.org/#/c/408152 | 16:18 |
mdbooth | ftersin_: Looking | 16:18 |
*** pcaruana has quit IRC | 16:18 | |
*** rcernin has quit IRC | 16:19 | |
*** armax has joined #openstack-nova | 16:19 | |
sdague | dansmith: I wonder if we should also make start_service explode if there are 2 services started with same exe and hostname | 16:20 |
*** owalsh has quit IRC | 16:20 | |
dansmith | sdague: we do as of my patch that depends on this | 16:20 |
dansmith | sdague: because we create hostmappings for the service in the main thread | 16:21 |
sdague | dansmith: where is that patch? | 16:21 |
sdague | both of these seem good, +A on them | 16:21 |
dansmith | sdague: https://review.openstack.org/#/c/406380/8/nova/test.py@335 is where we will 'splode now | 16:22 |
sdague | I left the comment on the dev formatting in the approved patch as well, hopefully vladikr can circle around on that one as a follow up | 16:22 |
dansmith | sdague: see that the second patch is embedded in that one ^ | 16:22 |
dansmith | sdague: yep | 16:22 |
sdague | dansmith: ok, cool | 16:22 |
dansmith | sdague: thanks for hitting those | 16:23 |
dansmith | and thanks to vladikr for jumping on them | 16:23 |
dansmith | once I convinced him they were broken :P | 16:23 |
*** irenab has joined #openstack-nova | 16:23 | |
*** mjura has quit IRC | 16:23 | |
markus_z | kashyap: I have to review it tomorrow. I have to leave for today. | 16:24 |
mriedem | dansmith: this isn't going to pass is it? https://review.openstack.org/#/c/396417/ | 16:24 |
mriedem | the depends-on isn't on the g-r bump | 16:24 |
*** roeyc has quit IRC | 16:24 | |
mriedem | oh snap https://review.openstack.org/#/c/407794/ | 16:24 |
mriedem | nevm | 16:24 |
dansmith | mriedem: the g-r bump landed | 16:24 |
dansmith | right | 16:24 |
dansmith | guess I should have removed the depends, but.. I didn't. | 16:24 |
prometheanfire | mdbooth: gr bump? | 16:25 |
mriedem | meh | 16:25 |
mriedem | global-requirements | 16:25 |
mriedem | prometheanfire: aren't you reqs core? | 16:25 |
mriedem | :) | 16:25 |
prometheanfire | :D | 16:25 |
prometheanfire | what's needed? | 16:25 |
*** armax has quit IRC | 16:25 | |
mriedem | nothing | 16:26 |
mriedem | i just wasn't up to date yet on over night activity | 16:26 |
prometheanfire | then my job here is done :P | 16:26 |
mriedem | zfs on out of here then! | 16:26 |
*** psachin has quit IRC | 16:26 | |
prometheanfire | lol | 16:26 |
* prometheanfire would like that :P | 16:26 | |
vladikr | sdague, dansmith, back, sorry | 16:26 |
*** annegentle has quit IRC | 16:26 | |
prometheanfire | uc-update should be merged in the next couple of hours as well | 16:27 |
dansmith | vladikr: np, sdague left a comment on the bottom patch | 16:27 |
vladikr | sdague, ah, yep, thanks, I'll fix it | 16:27 |
dansmith | cool | 16:27 |
vladikr | :) | 16:27 |
*** tbachman has quit IRC | 16:29 | |
mriedem | dansmith: +2 on https://review.openstack.org/#/c/396417/ - looks like easy fixes in the next one in the series | 16:29 |
mdbooth | ftersin_: I think that's ok, btw | 16:29 |
sfinucan | johnthetubaguy: went for a wander round nova to figure out what the console_driver CONF opt was for. Seems it's only applicable to XVP - not websocketproxy | 16:29 |
*** ayogi has joined #openstack-nova | 16:29 | |
sfinucan | johnthetubaguy: I don't know if we still care about XVP (i.e. has websocketproxy usurped it?), but I cleaned it up a small bit to clarify the intent anyway https://review.openstack.org/#/q/project:openstack/nova+topic:feat/deprecate-console_driver-opt | 16:30 |
* mdbooth waits for https://review.openstack.org/408152 to clear the gate before rebasing... again | 16:30 | |
dansmith | mriedem: yeah, already fixed locally, running tests | 16:30 |
*** nkrinner is now known as nkrinner_afk | 16:30 | |
*** irenab has quit IRC | 16:30 | |
*** yamahata has joined #openstack-nova | 16:31 | |
*** mvk has quit IRC | 16:31 | |
*** andymaier has quit IRC | 16:32 | |
*** markus_z has quit IRC | 16:33 | |
raj_singh | sfinucan: johnthetubaguy is on his honeymoon. I don't think he cares about any of this ATM :) | 16:33 |
*** mdnadeem has joined #openstack-nova | 16:33 | |
sfinucan | raj_singh: I did not know that. Congrats to him, heh | 16:34 |
*** edand has joined #openstack-nova | 16:34 | |
*** bvanhav__ has joined #openstack-nova | 16:34 | |
*** bvanhav_ has quit IRC | 16:34 | |
*** thorst_ is now known as thorst_afk | 16:35 | |
robcresswell | o/ hey all, how do I go about generating a policy file? Or is there on in the docs somewhere that I can reference quickly? | 16:36 |
*** armax has joined #openstack-nova | 16:37 | |
openstackgerrit | Dan Smith proposed openstack/nova: Add SingleCellSimple fixture https://review.openstack.org/405684 | 16:37 |
mriedem | robcresswell: http://docs.openstack.org/developer/nova/sample_policy.html | 16:37 |
mriedem | http://docs.openstack.org/developer/nova/_static/nova.policy.yaml.sample | 16:37 |
*** ducttape_ has quit IRC | 16:37 | |
mriedem | robcresswell: tox -r -e genpolicy | 16:37 |
*** TMM has quit IRC | 16:37 | |
*** paul-carlton2 has joined #openstack-nova | 16:38 | |
*** ayogi has quit IRC | 16:38 | |
*** prometheanfire has left #openstack-nova | 16:38 | |
*** irenab has joined #openstack-nova | 16:38 | |
mriedem | sfinucan: xen cares about xvp i think | 16:38 |
mriedem | so maybe ask BobBall | 16:38 |
robcresswell | Ah fantastic, thanks mriedem | 16:38 |
mriedem | i wanted to kill it awhile back but found it was maybe still in use by xen folk | 16:39 |
*** tbachman has joined #openstack-nova | 16:39 | |
sdague | dansmith: on https://review.openstack.org/#/c/405684/10/nova/tests/fixtures.py is there are reason those are instance methods instead of defined inside the setUp scope? | 16:39 |
mriedem | btw, i think we should all go in on a rice cooker for johnthetubaguy | 16:39 |
*** edand has quit IRC | 16:39 | |
sfinucan | mriedem: But do they /still/ care about it, I wonder. BobBall sounds as good a person to ask as anyone | 16:40 |
dansmith | sdague: no technical reason, as you can see, but.. is that a pattern? | 16:41 |
sdague | dansmith: mostly just curious whenever I see a thing which takes self, which doesn't need it | 16:41 |
*** paw has quit IRC | 16:42 | |
mriedem | dansmith: left a couple of nits in there | 16:42 |
dansmith | sdague: that wouldn't address the self-ness as setUp scoped things would still have access, but yes those could probably be staticmethods, I just usually don't because mock can be weird on some of those things | 16:42 |
dansmith | like the trouble we had around those staticmethod db handlers in the api objects | 16:43 |
robcresswell | Another question; what's the difference between "os-keypairs", "os-keypairs:discoverable" and "os-keypairs:index"? My assumption is that ":index" is the list call, but I'm a little confused by the other two. | 16:43 |
sdague | dansmith: ok, no prob | 16:43 |
dansmith | mriedem: okay | 16:43 |
mriedem | robcresswell: i think os-keypairs:discoverable makes it show up in the extension list | 16:43 |
mriedem | index is list | 16:44 |
*** umbSublime has quit IRC | 16:44 | |
mriedem | robcresswell: looks like the base one is used for show/detail https://github.com/openstack/nova/blob/master/nova/api/openstack/compute/keypairs.py#L304 | 16:45 |
*** dtp has joined #openstack-nova | 16:45 | |
mriedem | which is then checked in more detail here https://github.com/openstack/nova/blob/master/nova/api/openstack/compute/keypairs.py#L211 | 16:46 |
mriedem | i'm not exactly sure the difference | 16:46 |
mriedem | adding descriptions in the yaml sample is a TODO for the policies | 16:46 |
robcresswell | mriedem: Gotcha, thats enough for me to work on. Thanks again | 16:46 |
dansmith | sdague: mriedem: I'm in the middle of a big '-x tox -epy27' run but will do that docstring thing when it's done | 16:47 |
mriedem | robcresswell: oh https://github.com/openstack/nova/blob/master/nova/api/openstack/compute/keypairs.py#L304 is on the server extension | 16:48 |
mriedem | so the ability to show keypairs for a given server | 16:48 |
mriedem | rather than show details for a given keypair | 16:48 |
mriedem | that Controller class should probably be renamed to be something like ServerKeypairController or something more obvious | 16:48 |
mriedem | sdague: ^ would you agree? | 16:49 |
robcresswell | AH so "os_compute_api:os-keypairs:show" is the show for a given keypair then | 16:49 |
mriedem | yes | 16:49 |
*** irenab has quit IRC | 16:49 | |
mriedem | dansmith: ok, i'm going to be heading to lunch in 10 minutes so can peruse when i'm back | 16:50 |
*** diga has quit IRC | 16:50 | |
dansmith | mriedem: so no cells meeting for you eh/ | 16:51 |
dansmith | guess it will just be me and melwitt :) | 16:51 |
dtp | i'll be there | 16:51 |
openstackgerrit | Ed Leafe proposed openstack/nova: Cascade deletes of RP aggregate associations https://review.openstack.org/407707 | 16:51 |
edleafe | cdent: ^^ moved the test | 16:51 |
dansmith | guess it will just be me and melwitt and dtp :) | 16:51 |
mriedem | dansmith: oh didn't have it on my calendar - what about bauzas?! | 16:51 |
*** armax has quit IRC | 16:51 | |
melwitt | heh | 16:51 |
dansmith | if he shows | 16:51 |
cdent | thanks edleafe will go looksee in a mo | 16:51 |
*** andreas_s has quit IRC | 16:52 | |
*** tbachman has quit IRC | 16:52 | |
mriedem | dansmith: i think you just need to add a big irc nick reminder list for the start of your cells meeting | 16:52 |
mriedem | so i don't need to be responsible | 16:52 |
dansmith | eff that | 16:53 |
*** edand has joined #openstack-nova | 16:53 | |
mriedem | ok it's on my calendar now for the next 2 months, but i'm still not coming, so there | 16:53 |
dansmith | heh | 16:54 |
dansmith | I have little to say anyway | 16:54 |
*** browne has joined #openstack-nova | 16:58 | |
* mdbooth notes that there are currently only 14 patches in the imagebackend rebase stack | 16:58 | |
mdbooth | 11 files changed, 3032 insertions(+), 1080 deletions(-) | 16:59 |
andreaf | mriedem: this issue https://bugs.launchpad.net/tempest/+bug/1646779 is happening relatively frequently in the gate (54 hits in the past 7dd) - I cannot find anything on tempest side really nor in nova/libvirt logs - so I added nova to it now - I'm looking for nova folks help to look into it | 16:59 |
openstack | Launchpad bug 1646779 in tempest "Cannot connect to libvirt" [Undecided,New] | 16:59 |
*** kfox1111 has quit IRC | 17:01 | |
ftersin_ | mdbooth: great work. but is there a final state somewhere? | 17:01 |
*** dtp has left #openstack-nova | 17:01 | |
*** lpetrut has quit IRC | 17:02 | |
*** ducttape_ has joined #openstack-nova | 17:04 | |
*** tbachman has joined #openstack-nova | 17:04 | |
mdbooth | ftersin_: I was just about to push it, but I have to wait until the patch which is currently in the gate merges | 17:04 |
mdbooth | Otherwise I'll end up kicking it out | 17:04 |
mdbooth | ftersin_: Anyway, I need to get the end of the queue updated in gerrit, so I'm going to do a full push tomorrow morning (UK time) when most americans are asleep | 17:05 |
mdbooth | Hopefully it won't stress CI too much | 17:06 |
*** Kevin_Zheng has quit IRC | 17:07 | |
cdent | mdbooth: if you ping me when you get it all up to date tomorrow morning I'll go back through what remains | 17:07 |
ftersin_ | mdbooth: i'm waiting for this with interest | 17:08 |
mdbooth | Thanks | 17:08 |
mdbooth | Resolves-Bug or Closes-Bug, btw? | 17:09 |
mdbooth | Suspect it's Closes-Bug, but Resolves-Bug seemed to work and there are loads of them in the history | 17:09 |
*** moshele has joined #openstack-nova | 17:09 | |
mdbooth | mriedem: Given that I'm rebasing anyway, is it worth changing the above? | 17:09 |
*** Sukhdev_ has quit IRC | 17:10 | |
mdbooth | https://wiki.openstack.org/wiki/GitCommitMessages says Closes | 17:11 |
*** rcernin has joined #openstack-nova | 17:11 | |
mdbooth | diana_clarke: Hey, do you have a list of all the backend patches you abandoned? | 17:12 |
mdbooth | I can't find them locally | 17:13 |
*** rama_y has quit IRC | 17:13 | |
diana_clarke | mdbooth: Is this good enough? https://review.openstack.org/#/q/owner:diana.joan.clarke%2540gmail.com+status:abandoned I could filter it more. | 17:14 |
*** lpetrut has joined #openstack-nova | 17:14 | |
mdbooth | diana_clarke: Yes! Your gerrit-fu is superior to mine. | 17:15 |
mdbooth | Thanks :) | 17:15 |
*** moshele has quit IRC | 17:17 | |
*** sacharya has joined #openstack-nova | 17:20 | |
openstackgerrit | Diana Clarke proposed openstack/python-novaclient: CONF.osapi_max_limit -> CONF.api.max_limit https://review.openstack.org/408224 | 17:20 |
*** ihrachys has joined #openstack-nova | 17:22 | |
*** mvk has joined #openstack-nova | 17:24 | |
*** tbachman_ has joined #openstack-nova | 17:25 | |
*** tbachman has quit IRC | 17:25 | |
*** tbachman_ is now known as tbachman | 17:25 | |
*** Zer0Byte__ has joined #openstack-nova | 17:27 | |
openstackgerrit | Diana Clarke proposed openstack/python-novaclient: Usage missing from generated docs https://review.openstack.org/408227 | 17:28 |
*** baoli has joined #openstack-nova | 17:28 | |
*** paul-carlton2 has quit IRC | 17:30 | |
*** kristia__ has joined #openstack-nova | 17:30 | |
BobBall | mriedem, sfinucan: I don't think we use xvp ATM | 17:30 |
sfinucan | BobBall: If the docs are to be believed, something else has replaced it. Any idea what that might be? http://docs.openstack.org/admin-guide/compute-remote-console-access.html#about-nova-consoleauth | 17:32 |
BobBall | We just use novnc I think? | 17:32 |
*** dhellmann has quit IRC | 17:33 | |
sfinucan | Hmm... mriedem, how would one start with deprecating such a feature like XVP if it might not be used anymore? Mail to the operators list? | 17:33 |
*** mdnadeem has quit IRC | 17:34 | |
*** efoley__ has joined #openstack-nova | 17:34 | |
*** lucasagomes is now known as lucas-afk | 17:35 | |
*** daemontool has quit IRC | 17:35 | |
*** kristian__ has joined #openstack-nova | 17:36 | |
*** kristia__ has quit IRC | 17:36 | |
*** gouthamr has quit IRC | 17:37 | |
*** efoley_ has quit IRC | 17:38 | |
*** Apoorva has joined #openstack-nova | 17:38 | |
*** gouthamr has joined #openstack-nova | 17:38 | |
*** eglynn has quit IRC | 17:40 | |
*** xiaoyang has joined #openstack-nova | 17:40 | |
*** shangxdy has quit IRC | 17:41 | |
*** sudipto_ has quit IRC | 17:42 | |
*** sudipto has quit IRC | 17:42 | |
*** magic has quit IRC | 17:43 | |
*** paw has joined #openstack-nova | 17:44 | |
cfriesen | is there a straightforward way to get stats on where nova services are spending their time? does cProfile work with nova? | 17:46 |
jaypipes | sdague, mriedem, dansmith: bauzas' patch here that adds ability to filter resource provider (objects) via requested resources is ready IMHO: https://review.openstack.org/#/c/386242/. reviews welcome. | 17:46 |
jaypipes | cdent, edleafe: you guys too, though you've each reviewed prior patches (thanks!) | 17:47 |
openstackgerrit | sahid proposed openstack/nova: objects: remove cpu_topology from __init__ of InstanceNUMATopology https://review.openstack.org/408240 | 17:48 |
openstackgerrit | sahid proposed openstack/nova: objects: remove pagesize from __init__ of InstanceNUMATopology https://review.openstack.org/408241 | 17:48 |
openstackgerrit | sahid proposed openstack/nova: objects: remove related pining from __init__ of InstanceNUMATopology https://review.openstack.org/408242 | 17:48 |
*** paul-carlton2 has joined #openstack-nova | 17:48 | |
sahid | dansmith: wip of what you are looking for about the object InstanceNUMATopology ^ | 17:49 |
*** efoley__ has quit IRC | 17:51 | |
*** bvanhav_ has joined #openstack-nova | 17:52 | |
*** bvanhav__ has quit IRC | 17:52 | |
*** sacharya_ has joined #openstack-nova | 17:56 | |
*** derekh has quit IRC | 17:56 | |
*** rama_y has joined #openstack-nova | 17:56 | |
*** sacharya has quit IRC | 17:57 | |
*** alaski has quit IRC | 17:58 | |
*** dhellmann has joined #openstack-nova | 18:00 | |
*** sahid has quit IRC | 18:01 | |
*** yamahata has quit IRC | 18:01 | |
*** david-lyle_ is now known as david-lyle | 18:02 | |
*** fragatina has joined #openstack-nova | 18:03 | |
*** fragatin_ has joined #openstack-nova | 18:04 | |
*** ralonsoh has quit IRC | 18:04 | |
*** cdent has quit IRC | 18:05 | |
*** krtaylor has quit IRC | 18:06 | |
*** cleong has joined #openstack-nova | 18:06 | |
*** alaski has joined #openstack-nova | 18:06 | |
*** sacharya_ has quit IRC | 18:07 | |
BobBall | sfinucan, mriedem: I'm going to look into the xvp proxy actually - gimme a couple of days? | 18:07 |
*** fragatina has quit IRC | 18:08 | |
*** sacharya has joined #openstack-nova | 18:08 | |
openstackgerrit | Anusha Unnam proposed openstack/nova: Do not merge: Add post hook script to run tempest with old n-api https://review.openstack.org/401395 | 18:08 |
*** gjayavelu has joined #openstack-nova | 18:09 | |
*** sapcc-bot2 has quit IRC | 18:10 | |
*** carthaca_1 has quit IRC | 18:10 | |
*** carthaca_ has joined #openstack-nova | 18:11 | |
*** kaisers_ has joined #openstack-nova | 18:12 | |
*** sapcc-bot has joined #openstack-nova | 18:12 | |
*** parora has joined #openstack-nova | 18:12 | |
*** tbachman has quit IRC | 18:13 | |
*** prateek_ has quit IRC | 18:13 | |
*** gjayavelu has quit IRC | 18:13 | |
*** slaweq_ has quit IRC | 18:14 | |
*** dhellmann has quit IRC | 18:15 | |
*** slaweq_ has joined #openstack-nova | 18:16 | |
openstackgerrit | Dan Smith proposed openstack/nova: Make servers api view load instance fault from proper cell https://review.openstack.org/407238 | 18:17 |
openstackgerrit | Dan Smith proposed openstack/nova: Add schedule_and_build_instances conductor method https://review.openstack.org/367557 | 18:17 |
openstackgerrit | Dan Smith proposed openstack/nova: WIP Multicell support for instance listing https://review.openstack.org/396775 | 18:17 |
openstackgerrit | Dan Smith proposed openstack/nova: Setup CellsV2 environment in base test https://review.openstack.org/406380 | 18:17 |
openstackgerrit | Dan Smith proposed openstack/nova: Add SecurityGroup.identifier to prefer uuid over name https://review.openstack.org/406260 | 18:17 |
openstackgerrit | Dan Smith proposed openstack/nova: Add SingleCellSimple fixture https://review.openstack.org/405684 | 18:18 |
openstackgerrit | Dan Smith proposed openstack/nova: WIP Move instance creation to conductor https://review.openstack.org/319379 | 18:18 |
openstackgerrit | Dan Smith proposed openstack/nova: Fix up non-cells-aware context managers in test_db_api https://review.openstack.org/406379 | 18:18 |
dansmith | melwitt: you don't have a strong opinion about it, and neither do I aside from not wanting to go back and do all those again... is there anything better about the proposed approach other than that it's arguably a little nicer to look at? | 18:19 |
dansmith | if so, we should change it | 18:19 |
dansmith | if not, meh | 18:19 |
*** thorst_afk is now known as thorst_ | 18:19 | |
*** anush has quit IRC | 18:20 | |
*** gjayavelu has joined #openstack-nova | 18:20 | |
*** slaweq_ has quit IRC | 18:21 | |
*** prateek has joined #openstack-nova | 18:21 | |
*** slaweq_ has joined #openstack-nova | 18:23 | |
edleafe | jaypipes: can you clear up what the 'can_host' column indicates? Is it simply to say that a given RP is a compute node? | 18:24 |
*** parora has quit IRC | 18:24 | |
*** lpetrut has quit IRC | 18:24 | |
*** amoralej is now known as amoralej|off | 18:25 | |
*** nic has joined #openstack-nova | 18:27 | |
*** slaweq_ has quit IRC | 18:27 | |
*** tbachman has joined #openstack-nova | 18:28 | |
mriedem | mdbooth: should be closes-bug but not worth changing in an already approved change probably | 18:30 |
*** mat128 has quit IRC | 18:31 | |
mriedem | sfinucan: i want to say i emailed the operators list about xvp once awhile back, but it might have been a dream | 18:31 |
mriedem | cfriesen: os-profiler? | 18:32 |
mriedem | cfriesen: https://review.openstack.org/#/c/254703/ + rally? | 18:32 |
*** tbachman has quit IRC | 18:33 | |
*** tbachman has joined #openstack-nova | 18:33 | |
openstackgerrit | Sujitha proposed openstack/nova: Adding tags to InstancePayload https://review.openstack.org/407228 | 18:34 |
jaypipes | edleafe: yes. | 18:37 |
jaypipes | edleafe: we aren't yet using it correctly though. | 18:37 |
*** mat128 has joined #openstack-nova | 18:37 | |
jaypipes | edleafe: we will eventually use it but meh. | 18:38 |
edleafe | jaypipes: why should we even need it? | 18:38 |
jaypipes | edleafe: when joining against resource_provider_aggregates, we look for the providers that are *not* compute_nodes to determine if those providers can supply resources for a given consumer. | 18:38 |
jaypipes | edleafe: for example, without that flag, there's no way to differentiate between a compute node providing local disk and a shared storage pool providing shared disk. | 18:39 |
*** paul-carlton2 has quit IRC | 18:39 | |
jaypipes | edleafe: and when trying to determine appropriate providers that match a given request for resources, we need to query for compute nodes that have all the local resources we need or that have some of those resources but are associated with a provider of shared resources. | 18:40 |
*** anush has joined #openstack-nova | 18:41 | |
edleafe | jaypipes: I'm still not seeing it. If we are already filtering on VCPU/RAM, that will remove the shared disk providers, no? | 18:41 |
edleafe | jaypipes: it also seems like it's making placement and RPs very nova-centric | 18:43 |
*** b_rake has joined #openstack-nova | 18:43 | |
*** prateek has quit IRC | 18:43 | |
mriedem | dansmith: were you going to make the fakes in here static methods? https://review.openstack.org/#/c/405684/10..11/nova/tests/fixtures.py | 18:44 |
mriedem | dansmith: i think you said no because mocks/fixtures with staticmethods are wonky sometimes | 18:45 |
mriedem | but just checking | 18:45 |
*** irenab has joined #openstack-nova | 18:45 | |
*** voelzmo has joined #openstack-nova | 18:45 | |
jaypipes | edleafe: so, the user just requests DISK_GB = 100. let's say we have 3 compute nodes. all of them are in the same aggregate. 2 of them have no local disk. one has local disk. the aggregate is associated with a provider of shared storage. if I query for providers that have 2 VCPU, 1024 MEMORY_MB and 100 DISK_GB, I need to find providers that *either* have all of the resources locally OR that have | 18:45 |
jaypipes | some of those resources locally but are connected to a resource provider via an aggregate that satisfies the resources that are *not* on the compute node locally. | 18:45 |
*** armax has joined #openstack-nova | 18:46 | |
*** mkoderer has quit IRC | 18:46 | |
*** nicolasbock has quit IRC | 18:47 | |
dansmith | mriedem: right, I wasn't | 18:47 |
*** nicolasbock has joined #openstack-nova | 18:47 | |
*** nicolasbock has quit IRC | 18:47 | |
jaypipes | edleafe: in that second query, I have two compute nodes that have local VCPU and MEMORY_MB but not DISK_GB, so I must join to resource_provider_aggregates to determine the providers that are associated with any aggregate those nodes are in that has an inventory of DISK_GB. in the case described above, that join would return *BOTH* the shared storage pool resource provider *AND* the compute node | 18:48 |
*** nicolasbock has joined #openstack-nova | 18:48 | |
jaypipes | with local disk (because it's in the same aggregate). That isn't what we want, because we can't have the instance consume VCPU and RAM on one host and DISK_GB on another compute host. | 18:48 |
*** Nakato has quit IRC | 18:48 | |
*** Nakato has joined #openstack-nova | 18:48 | |
edleafe | jaypipes: I understand the use case, and the logic for shared storage. | 18:48 |
*** liusheng has quit IRC | 18:49 | |
edleafe | jaypipes: It just seems so nova-centric. When (not if :) the placement engine is used for all sorts of resource placement, it seems like we'll need to do the same for other "main" types of providers. | 18:50 |
*** liusheng has joined #openstack-nova | 18:50 | |
edleafe | jaypipes: i.e., "main" vs. "shared" | 18:51 |
jaypipes | edleafe: do you have a suggestion to make it more generic? | 18:51 |
edleafe | jaypipes: dunno... off the top of my head, "shared" seems like a candidate | 18:52 |
*** ducttape_ has quit IRC | 18:52 | |
edleafe | jaypipes: of course, that reverses the logic for 'can_host' | 18:52 |
jaypipes | edleafe: sure, I could totes go for that. | 18:52 |
edleafe | jaypipes: so maybe 'primary' | 18:52 |
jaypipes | edleafe: can_host isn't currently used anywhere AFAIK. | 18:52 |
jaypipes | edleafe: so should be an easy change if you want to tackle it. | 18:53 |
edleafe | jaypipes: I saw it in reviewing bauzas's patch: https://review.openstack.org/#/c/386242/11/nova/objects/resource_provider.py@562 | 18:54 |
edleafe | jaypipes: just made me think that I wasn't really clear on its purpose | 18:54 |
edleafe | jaypipes: but yeah, I can do a quick name change. | 18:54 |
edleafe | jaypipes: your preference for the name? | 18:55 |
*** nicolasbock has quit IRC | 18:56 | |
jaypipes | edleafe: shared is good I think? | 18:56 |
*** nicolasbock has joined #openstack-nova | 18:56 | |
edleafe | jaypipes: sounds good | 18:57 |
*** ljxiash has joined #openstack-nova | 19:00 | |
*** ihrachys has quit IRC | 19:01 | |
mriedem | sdague: if you're around, alex was +2 on this earlier, i'm +2 now, it's pretty straight forward pattern of proxy api deprecation which i know you love https://review.openstack.org/#/c/405461/ | 19:02 |
mriedem | would close a bp which you know i love | 19:03 |
*** armax has quit IRC | 19:03 | |
*** ljxiash has quit IRC | 19:05 | |
*** armax has joined #openstack-nova | 19:06 | |
*** ducttape_ has joined #openstack-nova | 19:08 | |
mriedem | kashyap: any idea why libvirt dies here? http://logs.openstack.org/66/401366/4/gate/gate-tempest-dsvm-neutron-full-ubuntu-xenial/3deacc5/logs/libvirt/libvirtd.txt.gz#_2016-12-01_18_16_04_907 | 19:10 |
mriedem | seeing quite a few virtlogd i/o errors in syslog http://logs.openstack.org/66/401366/4/gate/gate-tempest-dsvm-neutron-full-ubuntu-xenial/3deacc5/logs/syslog.txt.gz | 19:12 |
*** topol has left #openstack-nova | 19:13 | |
*** hemna_ has joined #openstack-nova | 19:13 | |
*** hemna_ has quit IRC | 19:15 | |
*** fragatin_ has quit IRC | 19:16 | |
*** krtaylor has joined #openstack-nova | 19:17 | |
mriedem | we're not using the recently merged virtlogd console stuff though b/c that requires libvirt>=1.3.3 and we test with 1.3.1 | 19:17 |
*** dinobot_ has quit IRC | 19:19 | |
*** ijw has joined #openstack-nova | 19:19 | |
*** rama_y has quit IRC | 19:22 | |
*** ociuhandu has quit IRC | 19:24 | |
*** Sukhdev has joined #openstack-nova | 19:25 | |
*** irenab has quit IRC | 19:28 | |
*** fragatina has joined #openstack-nova | 19:29 | |
*** irenab has joined #openstack-nova | 19:30 | |
*** gouthamr has quit IRC | 19:30 | |
*** s1061123 has quit IRC | 19:31 | |
*** s1061123 has joined #openstack-nova | 19:31 | |
*** jbernard has quit IRC | 19:32 | |
*** gouthamr has joined #openstack-nova | 19:33 | |
*** ihrachys has joined #openstack-nova | 19:35 | |
openstackgerrit | Diana Clarke proposed openstack/python-novaclient: Microversion 2.39 - Simple tenant usage pagination https://review.openstack.org/407825 | 19:36 |
openstackgerrit | Matt Riedemann proposed openstack/nova: api-ref: note that os-virtual-interfaces is nova-network only https://review.openstack.org/408270 | 19:37 |
*** armax has quit IRC | 19:37 | |
*** jbernard has joined #openstack-nova | 19:38 | |
*** fragatina has quit IRC | 19:40 | |
*** ljxiash has joined #openstack-nova | 19:42 | |
*** irenab has quit IRC | 19:44 | |
*** ljxiash has quit IRC | 19:46 | |
sdague | mriedem: just got back to my desk, will look now | 19:47 |
*** voelzmo has quit IRC | 19:48 | |
*** voelzmo has joined #openstack-nova | 19:48 | |
*** ihrachys has quit IRC | 19:49 | |
*** lamt has quit IRC | 19:49 | |
openstackgerrit | Maciej Szankin proposed openstack/nova: conf: add warning for vm's max delete attempts https://review.openstack.org/408274 | 19:51 |
*** irenab has joined #openstack-nova | 19:51 | |
*** dhellmann_ has joined #openstack-nova | 19:52 | |
*** voelzmo has quit IRC | 19:53 | |
*** jwcroppe has quit IRC | 19:53 | |
openstackgerrit | Maciej Szankin proposed openstack/nova: conf: add warning for vm's max delete attempts https://review.openstack.org/408274 | 19:54 |
*** kristian__ has quit IRC | 19:54 | |
*** jwcroppe has joined #openstack-nova | 19:54 | |
*** voelzmo has joined #openstack-nova | 19:54 | |
*** ihrachys has joined #openstack-nova | 19:55 | |
*** ihrachys has quit IRC | 19:55 | |
*** ihrachys has joined #openstack-nova | 19:56 | |
*** ihrachys has quit IRC | 19:56 | |
*** ihrachys has joined #openstack-nova | 19:57 | |
*** fragatina has joined #openstack-nova | 19:58 | |
*** jwcroppe has quit IRC | 19:58 | |
*** dhellmann_ is now known as dhellmann | 19:59 | |
*** cdent has joined #openstack-nova | 19:59 | |
*** irenab has quit IRC | 20:00 | |
*** clenimar has quit IRC | 20:01 | |
*** ihrachys has quit IRC | 20:02 | |
openstackgerrit | Maciej Szankin proposed openstack/nova: conf: remove deprecated barbican options https://review.openstack.org/397774 | 20:03 |
mriedem | kashyap: mdbooth: know anything about an -9 exit code from qemu-img info? | 20:07 |
mriedem | https://bugs.launchpad.net/nova/+bug/1646181 | 20:07 |
openstack | Launchpad bug 1646181 in OpenStack Compute (nova) "Fail to boot VM out of large snapshots (30GB+)" [Undecided,New] | 20:07 |
sdague | mriedem: +A | 20:08 |
mriedem | sdague: thanks | 20:08 |
*** irenab has joined #openstack-nova | 20:08 | |
mriedem | btw, nfs causes all of the problems | 20:08 |
mriedem | https://bugs.launchpad.net/nova/+bug/1646896 | 20:08 |
openstack | Launchpad bug 1646896 in OpenStack Compute (nova) "System hangs when using NFS storage backend with loopback mounts" [Undecided,New] | 20:08 |
*** anush has quit IRC | 20:10 | |
sdague | dansmith: can you comment on this bug - https://bugs.launchpad.net/nova/+bug/1644186 | 20:11 |
openstack | Launchpad bug 1644186 in OpenStack Compute (nova) "nova service performs wrong update from version 15 to 9" [Undecided,New] | 20:11 |
*** Nakato has quit IRC | 20:11 | |
dansmith | sdague: yeah | 20:11 |
*** sapcc-bot has quit IRC | 20:12 | |
*** carthaca_ has quit IRC | 20:12 | |
sdague | mriedem: yeh, well NFS is all kinds of bonkers at times | 20:12 |
*** Nakato has joined #openstack-nova | 20:12 | |
*** sapcc-bot has joined #openstack-nova | 20:12 | |
*** carthaca_ has joined #openstack-nova | 20:13 | |
sdague | mriedem: for the NFS bug I would suggest Confirmed / Low and if they push a patch, cool. | 20:13 |
openstackgerrit | Jay Pipes proposed openstack/nova: Extend get_all_by_filters to support resource criteria https://review.openstack.org/386242 | 20:15 |
openstackgerrit | Jay Pipes proposed openstack/nova: placement: API filters providers by resources https://review.openstack.org/408285 | 20:15 |
*** rfolco has quit IRC | 20:18 | |
*** irenab has quit IRC | 20:20 | |
mriedem | booya http://logs.openstack.org/00/352200/14/check/gate-tempest-dsvm-neutron-linuxbridge-ubuntu-xenial/6ba61d3/logs/screen-n-cpu.txt.gz?level=TRACE#_2016-11-17_04_43_55_673 | 20:24 |
*** anush has joined #openstack-nova | 20:24 | |
openstackgerrit | Pavel Kholkin proposed openstack/python-novaclient: WIP: [proxy-api] microversion 2.39 deprecates image-metadata proxy API https://review.openstack.org/408166 | 20:26 |
bauzas | p!spy | 20:26 |
*** irenab has joined #openstack-nova | 20:27 | |
*** Nakato has quit IRC | 20:28 | |
*** Nakato has joined #openstack-nova | 20:29 | |
*** voelzmo has quit IRC | 20:31 | |
*** voelzmo has joined #openstack-nova | 20:32 | |
*** gouthamr has quit IRC | 20:32 | |
*** bvanhav_ has quit IRC | 20:33 | |
*** voelzmo has quit IRC | 20:36 | |
*** irenab has quit IRC | 20:38 | |
*** harlowja has quit IRC | 20:41 | |
openstackgerrit | Pavel Kholkin proposed openstack/python-novaclient: remove variable '_' from tests https://review.openstack.org/408295 | 20:41 |
bauzas | jaypipes: sorry, I was on PTO this afternoon | 20:42 |
bauzas | jaypipes: why are you providing a separate change for providing GET ? | 20:42 |
*** TravT has joined #openstack-nova | 20:42 | |
*** diogogmt has joined #openstack-nova | 20:43 | |
bauzas | jaypipes: I thought we had a agreement between dansmith, mriedem, melwitt, alaski and me having POST /<something> | 20:43 |
bauzas | I know cdent and edleafe disagreed on that | 20:43 |
bauzas | but what I wrote tried to be a trade-off | 20:44 |
bauzas | ie. using POST with a body, but also using the RFC | 20:44 |
bauzas | ie. saying 'yup, we're getting you a 200, not a 201, which means we don't *create* a resource' | 20:44 |
mriedem | i thought we were doing POST + body | 20:44 |
bauzas | mriedem: what I wrote | 20:45 |
edleafe | bauzas: it isn't just POST vs GET, it's creating a brand new endpoint when we already have one for getting a filtered list of resource providers | 20:45 |
bauzas | edleafe: that's not an endpoint, that's a REST resource | 20:45 |
bauzas | we have the same endpoint | 20:45 |
*** irenab has joined #openstack-nova | 20:45 | |
edleafe | bauzas: fine, call it what you like | 20:45 |
edleafe | bauzas: we already have GET /resource_providers?<filters> | 20:46 |
bauzas | edleafe: I know | 20:46 |
edleafe | All we need to do is add more filters | 20:46 |
edleafe | not create an artificial resource | 20:46 |
bauzas | edleafe: and for the exact reason we have the same discussion since 4 weeks now, I don't want to have GET | 20:46 |
*** lamt has joined #openstack-nova | 20:46 | |
bauzas | having a long URL seems very bad to me | 20:47 |
mriedem | bauzas: so how do you do a POST to get info rather than create a new RP? | 20:47 |
mriedem | different body format? | 20:47 |
edleafe | mriedem: he does POST to a new resource called 'request' | 20:47 |
bauzas | mriedem: https://review.openstack.org/#/c/392569 | 20:47 |
edleafe | which means nothing in the placement API | 20:47 |
bauzas | edleafe: no, false | 20:48 |
bauzas | edleafe: again, in the same endpoint | 20:48 |
bauzas | which *is* the placement API | 20:48 |
*** rtheis has quit IRC | 20:48 | |
bauzas | just a separate resource for that placement API | 20:48 |
edleafe | bauzas: There is no concept of 'request' in the placement API | 20:48 |
bauzas | edleafe: and ? | 20:48 |
mriedem | bauzas: i think he means there is no such thing as a 'requests' resource | 20:48 |
mriedem | no db table or object | 20:48 |
mriedem | first, let's all take a deep breath and chill out :) | 20:48 |
bauzas | mriedem: we already have a POST method for the RP resource | 20:49 |
mriedem | b/c you're talking past each other | 20:49 |
mriedem | bauzas: i know, to create the RP | 20:49 |
mriedem | so we need a different route | 20:49 |
bauzas | mriedem: exactly, hence me creating a new resource | 20:49 |
edleafe | we already have a GET method to return a list of filtered RPs | 20:49 |
mriedem | but /requests could be resource providers, or inventories, or allocations, or aggregates, etc etc | 20:49 |
bauzas | edleafe: and I don't like it | 20:49 |
bauzas | mriedem: that's fine to me | 20:49 |
mriedem | i'm not a fan of passing a giant json blob as a query parameter either | 20:50 |
edleafe | bauzas: I'm sorry you don't like it, but that's the way HTTP works | 20:50 |
bauzas | mriedem: I'm providing a Content-Location header for knowing which exact resource | 20:50 |
edleafe | mriedem: nothing json about it | 20:50 |
edleafe | just key=val query parameters | 20:50 |
mriedem | what happens when we send required vs preferred? | 20:50 |
mriedem | are we going to namespace that? | 20:50 |
edleafe | mriedem: yes | 20:51 |
edleafe | jaypipes and I discussed that yesterday | 20:51 |
mriedem | oh barf | 20:51 |
bauzas | edleafe: again, I'm saying 'if you post that request, you'll get some identifier | 20:51 |
bauzas | https://review.openstack.org/#/c/392569/11/nova/api/openstack/placement/handlers/request.py@58 | 20:51 |
bauzas | aaaaand now, I'm seeing a totally different change against mine | 20:52 |
bauzas | lovely | 20:52 |
*** READ10 has quit IRC | 20:52 | |
mriedem | are there length restrictions on query parameters? | 20:52 |
bauzas | yup | 20:52 |
bauzas | the exact reason why I dislike that direction | 20:52 |
mriedem | "maximum URL length is 2048 characters" | 20:52 |
edleafe | mriedem: yes, a little over 8000 characters | 20:52 |
bauzas | edleafe: no | 20:52 |
bauzas | not over 8000 chars | 20:53 |
mriedem | sorry i'm just looking at http://www.w3schools.com/TAgs/ref_httpmethods.asp | 20:53 |
bauzas | edleafe: the exact answer is "it depends" | 20:53 |
bauzas | which is terrible | 20:53 |
edleafe | The only exception was IE browser, which was very weak - about 2K | 20:53 |
cdent | I'm sorry, there is a lot of misinformation being bounced around here. | 20:53 |
bauzas | so, depending of which WSGI server you used, you gonna don't have the same | 20:53 |
auggy | msg diana_clarke hi! how are you?? | 20:53 |
auggy | ahah whoops | 20:53 |
edleafe | The only limit we care about is the server, not a browser | 20:54 |
* cdent smiiles at auggy | 20:54 | |
diana_clarke | auggy: good! : | 20:54 |
*** irenab has quit IRC | 20:54 | |
mriedem | why can't we do a POST with a 200 response? | 20:54 |
mriedem | for 200 repsonse, "POST an entity describing or containing the result of the action; " | 20:54 |
*** b_rake has quit IRC | 20:54 | |
mriedem | *response | 20:54 |
* cdent raises his hand | 20:55 | |
mriedem | i'm posting a body of required/preferred stuff to process and want a list of available RPs back | 20:55 |
*** Nakato has quit IRC | 20:55 | |
*** READ10 has joined #openstack-nova | 20:55 | |
*** kristian__ has joined #openstack-nova | 20:56 | |
bauzas | mriedem: we can POST and return 200 | 20:56 |
bauzas | mriedem: that's exactly what I did | 20:56 |
bauzas | mriedem: but the route is different | 20:56 |
*** ociuhandu has joined #openstack-nova | 20:56 | |
*** Nakato has joined #openstack-nova | 20:56 | |
bauzas | mriedem: and I would hate having POST /rp returning either 201 or 200 | 20:56 |
mriedem | i don't like POST to /resource_providers having a different meaning for create and get, agree | 20:57 |
bauzas | mriedem: https://review.openstack.org/#/c/392569/11/nova/api/openstack/placement/handlers/request.py@53 | 20:57 |
*** Sukhdev has quit IRC | 20:57 | |
cdent | mriedem, bauzas: Do you guys mind if I step in here? | 20:58 |
*** kristian__ has quit IRC | 20:58 | |
mriedem | bauzas: where is the header stuff in this? | 20:58 |
bauzas | cdent: edleafe already did, but go for it, of course | 20:58 |
*** xyang1 has joined #openstack-nova | 20:58 | |
bauzas | mriedem: https://review.openstack.org/#/c/392569/11/nova/api/openstack/placement/handlers/request.py@61 | 20:58 |
cdent | there should be one and only one url that represents a list of resource providers. what have that already at /resource_providers | 20:59 |
bauzas | cdent: agreed | 20:59 |
cdent | we also already have a mechanism for retrieving a filtered list of resource providers: GET on that url | 20:59 |
bauzas | cdent: the exact reason why I'm returning a Content-Location header | 20:59 |
mriedem | bauzas: sure, i thought you meant it was processed from the request https://review.openstack.org/#/c/392569/11/nova/api/openstack/placement/handlers/request.py@81 | 20:59 |
cdent | (i'll get to the content-location header in a moment) | 21:00 |
mriedem | i.e. i'm requesting some kind of processing on allocations, or inventories, or resource providers | 21:00 |
bauzas | cdent: which means there is no way to get a list of RPs by POST /request | 21:00 |
*** dimtruck is now known as zz_dimtruck | 21:00 | |
bauzas | cdent: you get a result of a processing action | 21:00 |
*** anush has quit IRC | 21:00 | |
bauzas | cdent: and that result is equivalent to a RP list, but that's not a RP list | 21:00 |
cdent | there has been zero evidence that the query string for filtering a list of resource providers will grow without bound when we add resources | 21:00 |
cdent | there seems to be an expectation that it will | 21:01 |
*** irenab has joined #openstack-nova | 21:01 | |
mriedem | i guess i don't like assuming it won't | 21:01 |
bauzas | mriedem: oh no, the /request resource is rather a REST resource for the RequestSpec object | 21:01 |
cdent | except that by the time we get to when we have large "request specs" that we want to send to the placement api we will be doing a POST to /allocations to make a claim | 21:01 |
*** ociuhandu has quit IRC | 21:01 | |
mriedem | like when we get into nested resource providers and such | 21:01 |
mriedem | and relationships | 21:01 |
cdent | the operation that we are doing now is _only_ to limit a list of resource providers | 21:01 |
mriedem | and required/preferred logic | 21:01 |
bauzas | mriedem: ie. we have a ReqSpec object as a resource for describing user wishes | 21:01 |
bauzas | and we return something | 21:02 |
*** kristian__ has joined #openstack-nova | 21:03 | |
* edleafe is starting to get deja vu about POST /servers/<id>/action overloading | 21:03 | |
bauzas | cdent: again, the very main problem I have with GET /rps is that the maximum string size is different between two WSGI servers | 21:03 |
mriedem | cdent: but an allocation requires a resource provider right? | 21:03 |
bauzas | edleafe: that's something already there, so API WG should have a guideline, heh ? :) | 21:03 |
edleafe | bauzas: that's one of the insane APIs that led to the creation of the API-WG | 21:04 |
cdent | mriedem: no, not when making an actual claim (this is some ways off): send placement requirements get back _the_ resource provider you claimed (in one step) | 21:04 |
mriedem | i agree server actions is a bit wonky in places, i.e. the ones that have a body just for the action name but nothing inside the body | 21:04 |
bauzas | like, I'm totally okay, and I think the RFC7231 too, that a POST can be an action | 21:04 |
bauzas | edleafe: ^ | 21:04 |
edleafe | bauzas: *can* be | 21:04 |
cdent | bauzas: there's nothing wrong with using POST for all sorts of things, however it simply doesn't apply in this case on two dimensions: | 21:05 |
bauzas | edleafe: yeah, so what's the problem ? | 21:05 |
bauzas | I'm not returning 201 | 21:05 |
edleafe | that's why the guideline says that in the extreme, POST instead of GET is acceptable | 21:05 |
cdent | one is that the request needs to be against /resource_providers because that's what we want; two is that we are asking to GET a list of resource providers. | 21:05 |
bauzas | I'm returning 200, which is as per the RFC what we should return when we call an action | 21:05 |
mriedem | cdent: "send placement requirements get back _the_ resource provider you claimed (in one step)" | 21:05 |
mriedem | in ^, | 21:05 |
mriedem | if there are multiple resource providers that satisfy the request spec, what's the logic to pick one? just random? | 21:06 |
cdent | mriedem: that's the long term vision when doing a POST to /allocations | 21:06 |
mriedem | sure | 21:06 |
cdent | that's when the placement API _is_ the scheduler | 21:06 |
mriedem | so we start filtering/weighing RPs | 21:06 |
mriedem | and return the one RP that was picked | 21:06 |
cdent | yeah, way long term | 21:06 |
mriedem | for the allocation | 21:06 |
*** Sukhdev has joined #openstack-nova | 21:07 | |
bauzas | mriedem: FWIW, I originally called the REST resource "claim" but that looked overkill, so I preferred naming it "request" which is less confusing and more abstract :) | 21:07 |
cdent | in the meantime we're just doing limiting | 21:07 |
mriedem | why couldn't POST /resource_provider/requests/ resolve #1 above? | 21:07 |
*** kristian__ has quit IRC | 21:07 | |
mriedem | i know it's icky | 21:08 |
mriedem | but it's at least scoped to the RP resource | 21:08 |
bauzas | cdent: edleafe: so I hadn't get an answer to my precise point, don't you think it could defeat interop to get something very different if we use GET with a long query string, depending on your WSGI server ? | 21:08 |
bauzas | mriedem: I'm fine with that | 21:08 |
bauzas | mriedem: I'm not *that* tied to using /request | 21:08 |
mriedem | another thing is, the example request in http://docs-draft.openstack.org/18/385618/3/check/gate-nova-specs-docs-ubuntu-xenial/b089055//doc/build/html/specs/ocata/approved/resource-providers-get-by-request.html#post-resource-providers makes sense to me, | 21:08 |
*** paw has quit IRC | 21:09 | |
bauzas | mriedem: but cdent argued that it would mean that 'requests' could be an RP | 21:09 |
edleafe | bauzas: what is the worst case scenario? IOW, what's the shortest query string any WSGI server supports? | 21:09 |
mriedem | so i think i'd like to see examples of what a similar request would look like with the GET and filter params | 21:09 |
cdent | mriedem: that's hieararchy you are inferring there is not really how URIs work, but that's neither hear nor there. If POST is required (even though I still see no valid reason for why) then POST to /resource_providers (with a different schema than used for create) would the right thing semantically. But I still think GET is (way) more correct. | 21:09 |
mriedem | GET /resource_providers/?VCPU=2&MEMORY_MB=1024 | 21:09 |
cdent | mriedem: that's one of the proposals, jay put something slightly different in his recent change that implements his idea | 21:10 |
edleafe | mriedem: that spec was passed with the caveat that the approach was still to be determined | 21:10 |
cdent | but it is quite similar to that | 21:10 |
bauzas | edleafe: that's not a valid answer, the valid answer is still 'it depends' | 21:10 |
mriedem | edleafe: yes i know | 21:10 |
edleafe | mriedem: I'd be happy to create an alternate implementation | 21:10 |
sdague | is this POST more like a search? (coming up to speed late) | 21:10 |
bauzas | edleafe: if some server changes the max size for the URI, then kaboom | 21:10 |
mriedem | sdague: yes | 21:10 |
cdent | sdague: it's just a filter | 21:10 |
*** irenab has quit IRC | 21:10 | |
cdent | as in it is _literally_ a filter, and is implemented as one | 21:11 |
sdague | I guess, would POST /resource_providers/search be more clear? | 21:11 |
mriedem | edleafe: cdent: i'd like to compare POST vs GET with required/preferred resource classes | 21:11 |
mriedem | because i have a feeling the namespacing is going to get nasty | 21:11 |
bauzas | sdague: that is a reasonable approach which isn't fully RESTful but totally acceptable to me for the god sake of pragmatism | 21:11 |
edleafe | bauzas: unless you have actual numbers for these supposed limits, it's meaningless | 21:11 |
sdague | because this is also about consumers keeping the model in their head | 21:11 |
cdent | here's jay's proposal: http://lists.openstack.org/pipermail/openstack-dev/2016-December/108545.html | 21:11 |
sdague | and even if POST /resource_providers is technically ok | 21:11 |
sdague | it's different than the way we use POST to top level constructs in other places in openstack, and so ends up being conceptual speed bump | 21:12 |
*** ftersin_ has quit IRC | 21:12 | |
cdent | BTW: part of the reason for wanting GET is because of pragmatism: I can create, by typing, a reasonable query to see which compute nodes can give me 4 VPCU and 20GB of disk | 21:12 |
cdent | no json required | 21:12 |
sdague | cdent: yeh, GET seems ideal honestly | 21:13 |
bauzas | cdent: but we could be stuck with that forever | 21:13 |
sdague | bauzas: well, maybe, we have microversioning | 21:13 |
*** ociuhandu has joined #openstack-nova | 21:13 | |
bauzas | cdent: and then, once using traits and large requests, we could regret it | 21:13 |
sdague | if we think there is a good chance that GET works for a while | 21:13 |
bauzas | and that's not a FUD | 21:13 |
edleafe | bauzas: likewise, we would be stuck with POST forever | 21:13 |
sdague | we could have that as a primary interface | 21:13 |
sdague | and we could always add POST /resource_providers/search later in the game for really complex queries | 21:14 |
mriedem | GET /resource_providers?resources=DISK_GB:40,VCPU:2,MEMORY_MB:2048&required=storage:ssd,hw:cpu:x86:avx2&preferred=virt:hyperv:gen2 | 21:14 |
bauzas | sdague: I'd prefer to go on the POST /rp/search way now | 21:14 |
edleafe | jay's example is fairly complex, and is all of 126 characters | 21:14 |
mriedem | ^ is jaypipes' more complicated example with namespaces from the ML | 21:14 |
cdent | it's important to remember (as I said above) that there are two different interfaces under construction: now we are creating one to limit a list of resource providers by some characteristics. later we are creating one to fundamentally claim one | 21:14 |
bauzas | cdent: and ? | 21:15 |
bauzas | cdent: your point being? | 21:15 |
bauzas | cdent: that those two interfaces should be the same ? | 21:15 |
bauzas | or the contrary? | 21:15 |
mriedem | can i try to interpret? | 21:15 |
bauzas | mriedem: your pleasure | 21:16 |
mriedem | i think cdent is saying that by the time we need to deal with large complicated claims like this, or filters/claims, that those are POSTs to /allocations, | 21:16 |
cdent | bauzas: no, that the later interface will of necessity be different (it will involve a write to persistence and will have far more specific semantics) | 21:16 |
cdent | also what mriedem says | 21:16 |
mriedem | but right now we don't have traits, so maybe we don't need to boil the ocean on that requirement | 21:16 |
mriedem | if the simple query param filtering on GET is sufficient | 21:16 |
mriedem | until it's not | 21:16 |
mriedem | at which point, microversions mothertruckers | 21:16 |
sdague | mriedem: and when it is not, we can build an interface that is sufficient | 21:16 |
sdague | but lets not overly complicate the current interface well ahead of time | 21:17 |
sdague | GET is definitely easier to understand and dev hack against in adhoc ways | 21:17 |
bauzas | looks like the consensus moved then | 21:17 |
*** ociuhandu has quit IRC | 21:17 | |
mriedem | when we started talking about this <30 min ago i wasn't in the know about the POST to /allocations (claims) stuff | 21:17 |
bauzas | well | 21:18 |
*** kristia__ has joined #openstack-nova | 21:18 | |
mriedem | if we eventually reserve the complicated request spec stuff for that, i'm fine | 21:18 |
bauzas | on one hand, we say we shouldn't consider traits as a thing | 21:18 |
cdent | mriedem: we talked about that somewhat vaguely in portland, but only vaguely | 21:18 |
mriedem | we have no required/preferred logic built in right now | 21:18 |
bauzas | but on the other hand, we overcomplicate ourselves defining a scheduler claim interface we haven't yet agreed to have | 21:18 |
cdent | even when we do add traits, it will not be hugely complex, nor lengthy | 21:18 |
mriedem | cdent: i lost track of about 60% of what was said in hillsboro wrt RPs | 21:18 |
*** WillUK2010 has joined #openstack-nova | 21:19 | |
mriedem | cdent: except that we'd have 10 nova DBs by queens | 21:19 |
cdent | mriedem: yeah, I think everyone did. Only 10, I'm pretty sure it was 11 | 21:19 |
sdague | and, more importantly, if we do hit a limit, we actually have a mechanism for addressing that | 21:19 |
mriedem | cdent: well i wouldn't assume that, because NFV | 21:19 |
*** cleong has quit IRC | 21:19 | |
sdague | so it's not like GET actually paints us into a corner | 21:19 |
cdent | is 'because NFV' the new 'because docker'? | 21:20 |
mriedem | well, because NFV means, because crazy fucking use case | 21:20 |
*** Nakato has quit IRC | 21:20 | |
sdague | well, in NFV you specify which version of which FPGA firmware you are going to put this on paired with firmware version you need in your nics.... :) | 21:20 |
*** Nakato has joined #openstack-nova | 21:20 | |
*** irenab has joined #openstack-nova | 21:21 | |
mriedem | yeah | 21:21 |
sdague | I do agree this might grow into crazy land, but it still feels better to do the simpler thing first, with the "here is how we can do the more complicated thing if we have to" | 21:21 |
mriedem | so, the request spec gets nice and uber complicated | 21:21 |
mriedem | i think i've come around to that also | 21:22 |
sdague | yeh, it becomes NP complete problem to figure out where a VM lands | 21:22 |
edleafe | Hmmm... my google-fu is lacking. Can't find any documentation on query string limits for WSGI servers | 21:22 |
edleafe | But this shows limits of servers and browsers: https://boutell.com/newfaq/misc/urllength.html | 21:23 |
bauzas | edleafe: for example, I hate that http://serverfault.com/questions/757364/issues-with-url-length-in-apache-mod-wsgi | 21:24 |
bauzas | I mean, I hate the problem it could create | 21:24 |
cdent | I think this is a good problem to have | 21:25 |
edleafe | bauzas: thanks. That notes problems with URLs over 6K. I don't think we'll ever get close to that | 21:25 |
bauzas | I know that for now, CPU, RAM and mem are not lengthy bits | 21:25 |
cdent | if someone is making a multi-thousand char query to filter their list of resource providers, something is pathological and we've made some big mistakes on how we've designed and managed our systems | 21:25 |
bauzas | but I assume we were trying to address PCI requests ? | 21:25 |
bauzas | and NUMA queries ? | 21:25 |
sdague | yeh, 4K or 8K I thought was the issue | 21:25 |
sdague | we actually hit this with ec2 api when it was in tree | 21:26 |
sdague | because that's entirely a GET interface | 21:26 |
sdague | and some EC2 api calls > 8k of query params | 21:26 |
sdague | which apache can't do | 21:26 |
edleafe | sdague: ugh - I remember those | 21:26 |
mriedem | cdent: s/someone/compute/ in NFV | 21:27 |
mriedem | remember | 21:27 |
mriedem | crazy | 21:27 |
cdent | I'll stick to my statement: if the system starts requiring things that are that long, the system blowing up on them is our signal that we need to do something differently or make them stop | 21:27 |
openstackgerrit | Jim Rollenhagen proposed openstack/nova: Update ironic driver get_available_nodes docstring https://review.openstack.org/408308 | 21:27 |
jroll | ^ easy review if someone needs a break from the hard stuff, just a docstring update | 21:27 |
mriedem | jroll: ! | 21:28 |
edleafe | The request will be the resource class and amount. The complexity is in the nested resource providers | 21:28 |
*** smatzek has quit IRC | 21:28 | |
mriedem | you've walked into it now | 21:28 |
edleafe | That isn't part of the URL | 21:28 |
jroll | uh oh | 21:28 |
*** ociuhandu has joined #openstack-nova | 21:28 | |
mriedem | edleafe: yes it's nested RPs that worry me | 21:28 |
edleafe | jroll: wrong place at the wrong time | 21:28 |
edleafe | :) | 21:28 |
bauzas | edleafe: what if I'm adding more custom resource classes ? | 21:28 |
edleafe | mriedem: but they are never part of the API | 21:29 |
bauzas | edleafe: and what if I'm wanting more than just the standard tuple ? | 21:29 |
bauzas | edleafe: sorry? | 21:29 |
sdague | jroll: +2 | 21:29 |
jroll | nice, thanks sdague | 21:29 |
mriedem | edleafe: why not? | 21:29 |
bauzas | edleafe: say I'm adding a custom resource class of FOOBAR | 21:29 |
mriedem | bauzas: you mean CUSTOM_FOOBAR? | 21:29 |
bauzas | edleafe: I should be able to query the placement API against that FOOBAR | 21:29 |
edleafe | mriedem: because the request if for a) amounts of a class and b) traits (eventually) | 21:29 |
bauzas | mriedem: yeah | 21:29 |
mriedem | CUSTOM_BAREMETAL_JROLL_DONUTS | 21:29 |
bauzas | yeah that, loving it | 21:29 |
jroll | ಠ_ಠ | 21:29 |
*** burt has quit IRC | 21:30 | |
edleafe | how those classes are arranged with nested RPs is internal to the placement engine | 21:30 |
sdague | so... before we solve CUSTOM_BAREMETAL_JROLL_DONUTS, I would rather be using this for CPU/MEM :) | 21:30 |
bauzas | or people would want to class French cheeses | 21:30 |
bauzas | sdague: but we agreed on having custom resource classes for Ocata :) | 21:30 |
bauzas | that's a priority too | 21:30 |
bauzas | which means that by Ocata, people could litterally ask for anything they like | 21:31 |
sdague | bauzas: ok, but we still aren't scheduling with it? | 21:31 |
mriedem | i'm less worried about GET /resource_providers/?CUSTOM_BAREMETAL_JROLL_DONUTS=5 than nested RPs | 21:31 |
*** irenab has quit IRC | 21:31 | |
bauzas | sdague: well, we could | 21:31 |
dansmith | sdague: FWIW, I agree that anything beyond reporting and scheduling with the base resources is a distraction at this point | 21:31 |
dansmith | sdague: but I'm in the minority | 21:31 |
mriedem | CUSTOM_BAREMETAL_JROLL_DONUTS=5 just returns me a list of ironic nodes doesn't it? | 21:31 |
sdague | dansmith: ok, I'm in that minority as well then | 21:31 |
cdent | dansmith: i'm in that minority too, so it is starting to sound like not a minority | 21:32 |
edleafe | mriedem: yeah, because adding CUSTOM_BAREMETAL_JROLL_DONUTS=5 increases the length by a whopping 32 characters | 21:32 |
sdague | because, it feels like we often end up arguing about things that can't land in this cycle, instead of the things that can, so less lands in current cycle | 21:32 |
mriedem | edleafe: i'm not talking about length restrictions actually | 21:32 |
mriedem | but anyway | 21:32 |
*** ociuhandu has quit IRC | 21:32 | |
edleafe | cdent: sdague: dansmith: welcome to the minority! It's getting crowded here | 21:32 |
dansmith | sdague: and, it would be nice to prove out what we have and make actual gains, before we get further down the rabbit hole | 21:32 |
bauzas | sdague: cdent: dansmith: do we have a Majority Whip then ? | 21:33 |
edleafe | mriedem: understood. But POST doesn't help that either | 21:33 |
bauzas | call Frank Underwood | 21:33 |
dansmith | sdague: so far we have no gains from all this, just reporting a bunch of stuff we're not using | 21:33 |
sdague | 3 means we're an organization right? | 21:33 |
mriedem | ok, so let's go with GET + filteres | 21:33 |
mriedem | *filters | 21:33 |
sdague | alice's restaurant rules | 21:33 |
*** yamahata has joined #openstack-nova | 21:33 | |
edleafe | mriedem: heh, I thought that was French | 21:33 |
mriedem | if we turn out to be terribly wrong, we pivot with a microversion | 21:33 |
cdent | dansmith: the limiting in the scheduler is very soon, but it required this thing we were debating just now :( | 21:33 |
dansmith | cdent: yeah, I know the query thing is a dep | 21:34 |
cdent | your statement is kind of a qed: "query thing" :) | 21:34 |
mriedem | bauzas: so i think we're doing GET | 21:34 |
jroll | so the downside of not scheduling on custom stuff is it pushes getting rid of the ironic-horribleness out another cycle, to be clear (don't have a strong opinion on whether y'all do that or not) | 21:34 |
bauzas | mriedem: that's what I'm seeing :) | 21:34 |
dansmith | jroll: I know | 21:35 |
dansmith | bauzas: really? | 21:35 |
dansmith | er, | 21:35 |
dansmith | mriedem: really? | 21:35 |
edleafe | jroll: agreed, but we set modest goals for Ocata | 21:35 |
*** Jeffrey4l has quit IRC | 21:35 | |
jroll | right, just making sure that's clear :) | 21:35 |
jroll | I know some people are super hyped about deleting that code | 21:35 |
mriedem | dansmith: that's what we were just talking about | 21:35 |
dansmith | mriedem: seems like the most people prefer POST no? | 21:35 |
mriedem | dansmith: no | 21:35 |
*** Jeffrey4l has joined #openstack-nova | 21:35 | |
mriedem | unless i'm misunderstanding | 21:35 |
dansmith | huh, okay | 21:35 |
mriedem | as i said earlier, i wasn't aware fully about the eventual plan to POST requests to /allocations to make RP claims | 21:36 |
mriedem | traits aren't a thing yet, so the namespace/nesting garbage in the GET query is less of a concern for me right now | 21:36 |
openstackgerrit | Chris Dent proposed openstack/nova: WIP [placement] start a placement_dev doc https://review.openstack.org/408313 | 21:36 |
mriedem | as noted, if we worry about all of that stuff too much we won't get anything done for ocata | 21:36 |
mriedem | and we can always do a POST later if needed | 21:37 |
*** catintheroof has quit IRC | 21:37 | |
dansmith | bummer | 21:37 |
cdent | jaypipes, edleafe, bauzas (and anyone else): that placement_dev doc ^^^ could do with some feedback on direction and style before I get too much further | 21:37 |
*** irenab has joined #openstack-nova | 21:38 | |
*** catintheroof has joined #openstack-nova | 21:38 | |
bauzas | mriedem: okay, Jay wrote a bit of that direction in a separate change, I'll squash it with my change and co-author mr. jaypipes | 21:38 |
bauzas | mriedem: I still think we could need a microversion very soon, but okay, let's merge that | 21:39 |
mriedem | bauzas: that's this right? https://review.openstack.org/#/c/408285/ | 21:40 |
bauzas | I'm just done with discussing over and over this since more than 4 weeks now | 21:40 |
bauzas | so, if the consensus moved and now more people are inclined with GET, then prepare your +2s | 21:40 |
mriedem | sdague: to make sure, you're pro GET now right? | 21:40 |
mriedem | sdague: with the understanding we can change later if needed? | 21:40 |
sdague | mriedem: I am pro GET | 21:40 |
bauzas | mriedem: yup | 21:41 |
mriedem | bauzas: i'm prepared to go home soon | 21:41 |
mriedem | and cry | 21:41 |
bauzas | mriedem: jay made most of the work, hence me merging that | 21:41 |
sdague | yes, with the back pocket move to change once NFV has a turing complete query string to allocate servers | 21:41 |
bauzas | mriedem: that will be tomorrow anyway | 21:41 |
*** harlowja has joined #openstack-nova | 21:41 | |
dansmith | so, | 21:41 |
dansmith | we all know we'll have to have a post, | 21:41 |
dansmith | but meh? | 21:42 |
dansmith | I like this plan | 21:42 |
bauzas | exactly | 21:42 |
cdent | dansmith: [t 4fPX] | 21:42 |
purplerbot | <cdent> except that by the time we get to when we have large "request specs" that we want to send to the placement api we will be doing a POST to /allocations to make a claim [2016-12-07 21:01:35] [n 4fPX] | 21:42 |
bauzas | cdent: wrong assumption | 21:42 |
edleafe | dansmith: the POST is for a completely different action | 21:42 |
dansmith | yeah, I disagree with that | 21:42 |
*** catintheroof has quit IRC | 21:42 | |
cdent | so the idea is pretty much what you were saying: do the basics now, get the win | 21:42 |
dansmith | but doing the thing that we can do with the get via post is equally as trivial | 21:43 |
cdent | update the scheduler to actually get some advantage from the info | 21:43 |
*** thorst_ has quit IRC | 21:43 | |
bauzas | cdent: because scheduler claims are way out of being merged soon, but nested resource providers and custom resource classes are already a thing | 21:43 |
openstackgerrit | Sarafraj Singh proposed openstack/nova: WIP: Objectify requests_and_created_ports https://review.openstack.org/405398 | 21:43 |
dansmith | anyway, whatever, I keep saying I'm going to stop having this conversation | 21:43 |
cdent | looks let's not go back into this. we have a plan | 21:43 |
edleafe | dansmith: another use case for this is for things like Watcher that want to check for possible hosts before live migrating | 21:43 |
*** thorst_ has joined #openstack-nova | 21:44 | |
edleafe | this filtered list of hosts isn't going away | 21:44 |
cdent | sdague's perspective is spot on: [t n8R] | 21:44 |
purplerbot | <sdague> GET is definitely easier to understand and dev hack against in adhoc ways [2016-12-07 21:17:31] [n n8R] | 21:44 |
cdent | that provides us with flexibility | 21:44 |
cdent | and explorability | 21:44 |
*** thorst_ has quit IRC | 21:44 | |
cdent | and doesn't lock us into anything | 21:44 |
bauzas | I'm out of battery and spit | 21:44 |
bauzas | oops | 21:45 |
bauzas | spit is maybe not the right word | 21:45 |
cdent | best quote ever | 21:45 |
cdent | no it's perfect | 21:45 |
cdent | :) | 21:45 |
*** slaweq_ has joined #openstack-nova | 21:46 | |
*** paw has joined #openstack-nova | 21:46 | |
bauzas | could I beg you for changing that to "I'm out of battery and breath" | 21:47 |
bauzas | so I'm gonna leave anyway and do $stuff tomorrow | 21:47 |
openstackgerrit | Diana Clarke proposed openstack/nova: Simple tenant usage pagination https://review.openstack.org/386093 | 21:48 |
*** irenab has quit IRC | 21:48 | |
*** jamielennox|away is now known as jamielennox | 21:49 | |
cdent | yeah, me too, been too long of a day | 21:49 |
*** cdent has quit IRC | 21:51 | |
openstackgerrit | Sujitha proposed openstack/nova: Adding tags to InstancePayload https://review.openstack.org/407228 | 21:53 |
*** takashin has joined #openstack-nova | 21:54 | |
*** krtaylor has quit IRC | 21:58 | |
*** irenab has joined #openstack-nova | 21:59 | |
*** slunkad has quit IRC | 22:05 | |
*** slunkad has joined #openstack-nova | 22:06 | |
*** mriedem has quit IRC | 22:06 | |
*** irenab has quit IRC | 22:07 | |
*** dave-mccowan has quit IRC | 22:09 | |
openstackgerrit | Diana Clarke proposed openstack/nova: Simple tenant usage pagination https://review.openstack.org/386093 | 22:11 |
*** kristia__ has quit IRC | 22:12 | |
*** baoli has quit IRC | 22:12 | |
*** baoli has joined #openstack-nova | 22:13 | |
*** irenab has joined #openstack-nova | 22:14 | |
*** baoli has quit IRC | 22:18 | |
*** kaisers_ has quit IRC | 22:18 | |
*** irenab has quit IRC | 22:19 | |
*** jheroux has quit IRC | 22:23 | |
*** READ10 has quit IRC | 22:25 | |
*** eharney has quit IRC | 22:27 | |
*** kylek3h_away has quit IRC | 22:32 | |
*** armax has joined #openstack-nova | 22:33 | |
*** tblakes has quit IRC | 22:35 | |
*** slaweq_ has quit IRC | 22:37 | |
*** esberglu has quit IRC | 22:37 | |
*** edand has quit IRC | 22:38 | |
*** esberglu has joined #openstack-nova | 22:38 | |
*** krtaylor has joined #openstack-nova | 22:42 | |
*** esberglu has quit IRC | 22:42 | |
*** rcernin has quit IRC | 22:48 | |
*** ayoung has quit IRC | 22:48 | |
*** lamt has quit IRC | 22:48 | |
*** jamesdenton has quit IRC | 22:51 | |
*** browne has quit IRC | 22:53 | |
*** mriedem has joined #openstack-nova | 22:55 | |
*** unicell has joined #openstack-nova | 23:03 | |
*** armax has quit IRC | 23:03 | |
*** mdrabe has quit IRC | 23:06 | |
*** ducttape_ has quit IRC | 23:07 | |
*** WillUK2010 has quit IRC | 23:09 | |
*** gjayavelu has quit IRC | 23:11 | |
*** armax has joined #openstack-nova | 23:12 | |
*** gjayavelu has joined #openstack-nova | 23:14 | |
*** kaisers_ has joined #openstack-nova | 23:14 | |
*** gjayavelu has joined #openstack-nova | 23:14 | |
*** mlavalle has quit IRC | 23:15 | |
*** rfolco has joined #openstack-nova | 23:15 | |
*** rfolco has quit IRC | 23:15 | |
*** harlowja has quit IRC | 23:16 | |
*** david-lyle_ has joined #openstack-nova | 23:20 | |
*** david-lyle_ has quit IRC | 23:20 | |
*** xyang1 has quit IRC | 23:22 | |
*** sacharya has quit IRC | 23:23 | |
*** sacharya has joined #openstack-nova | 23:24 | |
*** zz_dimtruck is now known as dimtruck | 23:26 | |
*** jamielennox is now known as jamielennox|away | 23:27 | |
*** jamielennox|away is now known as jamielennox | 23:28 | |
*** WillUK2010 has joined #openstack-nova | 23:31 | |
*** armax has quit IRC | 23:34 | |
*** browne has joined #openstack-nova | 23:41 | |
*** esberglu has joined #openstack-nova | 23:45 | |
openstackgerrit | Sarafraj Singh proposed openstack/nova: Objectify requests_and_created_ports https://review.openstack.org/405398 | 23:45 |
*** ijw has quit IRC | 23:46 | |
*** ijw has joined #openstack-nova | 23:46 | |
*** esberglu has quit IRC | 23:49 | |
*** tbachman has quit IRC | 23:50 | |
*** erlon-airlong has quit IRC | 23:50 | |
*** ducttape_ has joined #openstack-nova | 23:56 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!