Thursday, 2017-03-09

*** zz_dimtruck is now known as dimtruck00:01
alex_xumriedem: will check that patch in the morning00:03
alex_xujaypipes: something looking for your feedback https://review.openstack.org/441829, cdent didn't want the cache :)00:03
jaypipesalex_xu: yeah, it's next on my list. currently reviewing https://review.openstack.org/#/c/435954/6/specs/pike/approved/enable-sriov-nic-features.rst00:04
alex_xujaypipes: thanks :)00:05
*** salv-orlando has quit IRC00:05
*** tbachman has joined #openstack-nova00:07
*** david-lyle has joined #openstack-nova00:08
*** yingjun has joined #openstack-nova00:09
dansmithwhat kind of monster doesn't want a cache?00:09
*** claudiub has quit IRC00:11
*** gfhellma has joined #openstack-nova00:12
*** hongbin has quit IRC00:14
*** david-lyle has quit IRC00:15
*** kylek3h has joined #openstack-nova00:18
*** browne has quit IRC00:18
*** gfhellma has quit IRC00:20
*** haplo37 has quit IRC00:20
dansmithalex_xu: can you drop your -1 from this now? https://review.openstack.org/#/c/441388/900:21
*** thorst has joined #openstack-nova00:25
mriedemdansmith: just setup amazon household for prime, finally00:25
mriedemoh so sweet00:25
mriedemnow i don't have to secretly log into laura's account to order things00:25
dansmithyou mean you just got prime, or just got setup so laura can order stuff without you?00:25
dansmithoh heh00:25
mriedemshe's had prime for years00:25
dansmithwelcome to 2009!00:26
mriedembut has always been my proxy00:26
*** dave-mccowan has joined #openstack-nova00:26
mriedem\o/00:26
jaypipesmriedem: can you record the Grand Tour and let me stream from your house? :)00:26
dansmithjaypipes: I know you're old, but... amazon prime video is streaming and generally "recording" it is hard to do00:27
jaypipesdansmith: I was expecting mriedem to set up a camcorder, tape it, upload the tape to his personal server and let me stream it from there.00:27
* dansmith headdesks00:28
jaypipesrofl00:28
jaypipesalex_xu: ok, sorry, I goofed. you were actually third in the queue, right behind my current review on https://review.openstack.org/#/c/364468/11/specs/pike/approved/reserve-numa-with-pci.rst00:29
*** haplo37 has joined #openstack-nova00:29
*** jamesdenton has joined #openstack-nova00:29
*** thorst has quit IRC00:29
openstackgerritDan Smith proposed openstack/nova master: Change MQ targeting to honor only what is in the context  https://review.openstack.org/43802300:43
openstackgerritDan Smith proposed openstack/nova master: Teach os-migrations about cells  https://review.openstack.org/44268500:43
openstackgerritDan Smith proposed openstack/nova master: Add cell field to Destination object  https://review.openstack.org/43802400:43
openstackgerritDan Smith proposed openstack/nova master: Teach simple_tenant_usage about cells  https://review.openstack.org/44138800:43
openstackgerritDan Smith proposed openstack/nova master: Teach os-aggregates about cells  https://review.openstack.org/44220000:43
mriedemwho wants to volunteer to run the meeting in the morning?00:44
*** dtp has quit IRC00:45
*** catintheroof has quit IRC00:46
*** zhurong has joined #openstack-nova00:47
*** rfolco has quit IRC00:48
*** iceyao has joined #openstack-nova00:50
*** Shunli has joined #openstack-nova00:50
*** david-lyle has joined #openstack-nova00:54
*** ducttape_ has quit IRC00:54
*** david-lyle has quit IRC00:54
openstackgerritDan Smith proposed openstack/nova master: Change MQ targeting to honor only what is in the context  https://review.openstack.org/43802300:57
openstackgerritDan Smith proposed openstack/nova master: Teach os-migrations about cells  https://review.openstack.org/44268500:57
openstackgerritDan Smith proposed openstack/nova master: Add cell field to Destination object  https://review.openstack.org/43802400:57
openstackgerritDan Smith proposed openstack/nova master: Teach simple_tenant_usage about cells  https://review.openstack.org/44138800:57
openstackgerritDan Smith proposed openstack/nova master: Teach os-aggregates about cells  https://review.openstack.org/44220000:57
* dansmith shakes his fist at pep800:59
*** yuntongjin has joined #openstack-nova01:00
*** jamesdenton has quit IRC01:00
*** Jeffrey4l_ has quit IRC01:00
*** dave-mccowan has quit IRC01:02
*** thorst has joined #openstack-nova01:02
*** thorst has quit IRC01:02
*** ssurana has joined #openstack-nova01:03
*** mlavalle has quit IRC01:10
*** iceyao has quit IRC01:10
*** esberglu has joined #openstack-nova01:10
*** iceyao has joined #openstack-nova01:10
*** esberglu has quit IRC01:11
*** ducttape_ has joined #openstack-nova01:13
dansmithgfdi01:14
dansmithjaypipes: your requests have thrown off the whole cosmic balance of my patches01:14
jaypipesdansmith: :(01:14
jaypipessorry duder.01:14
melwittjaypipes: I've got the transactional safety change in my working dir. just finishing up some changes to the quota count patch01:15
jaypipesmelwitt: coolio.01:16
*** iceyao_ has joined #openstack-nova01:20
*** namnh has joined #openstack-nova01:21
*** iceyao has quit IRC01:24
openstackgerritDan Smith proposed openstack/nova master: Change MQ targeting to honor only what is in the context  https://review.openstack.org/43802301:25
openstackgerritDan Smith proposed openstack/nova master: Teach os-migrations about cells  https://review.openstack.org/44268501:25
openstackgerritDan Smith proposed openstack/nova master: Add cell field to Destination object  https://review.openstack.org/43802401:25
openstackgerritDan Smith proposed openstack/nova master: Teach simple_tenant_usage about cells  https://review.openstack.org/44138801:25
openstackgerritDan Smith proposed openstack/nova master: Teach os-aggregates about cells  https://review.openstack.org/44220001:25
*** Lambert has joined #openstack-nova01:26
*** unicell has quit IRC01:34
*** ducttape_ has quit IRC01:35
*** mtanino has quit IRC01:35
*** salv-orlando has joined #openstack-nova01:37
*** gjayavelu has quit IRC01:38
*** ducttape_ has joined #openstack-nova01:38
*** youzp_ has joined #openstack-nova01:39
*** tuanluong has joined #openstack-nova01:40
*** salv-orlando has quit IRC01:41
*** adrian_otto has quit IRC01:46
*** ducttape_ has quit IRC01:49
*** dave-mcc_ has joined #openstack-nova01:50
*** yuntongjin has quit IRC01:52
*** yuntongjin has joined #openstack-nova01:52
*** david-lyle has joined #openstack-nova01:52
*** kevinz has joined #openstack-nova01:57
*** tommylikehu has joined #openstack-nova02:01
*** thorst has joined #openstack-nova02:03
*** ijw has quit IRC02:06
*** tovin07_ has joined #openstack-nova02:06
*** ijw has joined #openstack-nova02:07
*** thorst has quit IRC02:08
*** ijw has quit IRC02:09
*** tommylikehu has quit IRC02:12
*** haplo37 has quit IRC02:12
*** trinaths1 has left #openstack-nova02:17
*** Kevin_Zheng has joined #openstack-nova02:17
Kevin_ZhengHi, is this a known issue about quotas and cells? https://bugs.launchpad.net/nova/+bug/167062702:18
openstackLaunchpad bug 1670627 in OpenStack Compute (nova) "quota is always in-use after delete the ERROR instances " [Undecided,In progress] - Assigned to huangtianhua (huangtianhua)02:18
Kevin_ZhengLooks critical02:19
*** gjayavelu has joined #openstack-nova02:19
*** haplo37 has joined #openstack-nova02:21
melwittKevin_Zheng: we have seen some quota bugs on the multicell patches that are currently in progress, but I hadn't heard about the specific issue with ERROR instances until now02:23
*** markvoelker has quit IRC02:23
*** gjayavelu has quit IRC02:23
Kevin_Zhengmelwitt: Thanks for the reply02:23
*** tbachman has quit IRC02:24
melwittKevin_Zheng: and I think your analysis is right. we've been working on quotas in regard to cells right now, which will address the issues you're seeing. just didn't realize there were already problems in the wild with the single cell case02:24
Kevin_Zhengmelwitt: shouldn't all quota calculated from nova_api db?02:24
melwittKevin_Zheng: yes, see this series https://review.openstack.org/#/c/410945/02:24
*** erlon has quit IRC02:25
Kevin_Zhengmelwiit: yeah, I know you are working on quota related issues so I ask here to see if it is already known :)02:25
melwittit isn't in nova_api db yet, that series moves it there and gets rid of "usages" and "reservations"02:25
Kevin_ZhengOk, thanks, I will have a look02:26
Kevin_Zhengwill it fix the problem we meet?02:27
melwittah, cool. yeah, we didn't know about that one yet and I agree it sounds critical. our intention was that things should be okay in the single cell deployment02:27
mriedemKevin_Zheng: we'll have to fix the bug separately02:28
*** gongysh has joined #openstack-nova02:28
melwittyes, it will fix that problem. but I think we'll have to do a small different fix for now, that we can backport02:28
mriedembecause we have to backport the fix02:28
mriedemjinx02:28
Kevin_Zhengyeah sounds reasonable02:28
mriedemso when we bury the instance in cell0,02:28
melwitthah, yeah02:28
mriedemwe have to remove the reservation from cell002:28
mriedemi think?02:28
mriedemwe must be making a claim in cell0 or how does that work?02:29
mriedems/claim/reservation/02:29
*** yuntongjin has quit IRC02:29
*** yuntongjin has joined #openstack-nova02:30
melwittthinking ... we're not doing anything to target the reservation/usage/etc02:30
melwittso it should be in cell0 I think. and that's not getting removed ... so I think you're right02:30
mriedemthe quota reservation happens in nova-api here https://github.com/openstack/nova/blob/master/nova/compute/api.py#L31202:33
mriedembut since that's not targeted to cell0, wouldn't that be in the main cell, i.e. nova database?02:33
*** Apoorva has quit IRC02:34
mriedemthe commit is here https://github.com/openstack/nova/blob/master/nova/compute/api.py#L105902:34
mriedemthere is no rollback of the quota in _bury_in_cell002:35
mriedemhttps://github.com/openstack/nova/blob/master/nova/conductor/manager.py#L83602:35
mriedemwhich probably makes sense because don't we want to keep the reservation until the instance is actually deleted?02:35
melwittthere shouldn't be because ERROR instances occupy quota02:35
melwittbut when the user deletes it, we need to do a reverse reserve and commit and somehow that's going to the wrong db02:36
mriedemwe put the instance in cell0 here https://github.com/openstack/nova/blob/master/nova/conductor/manager.py#L88102:37
melwittit's probably something like, we target cell0 to delete the instance and the quota reserve/commit for the freeing of quota goes to cell0 in the same target and we didn't mean to02:37
melwittor something02:37
*** salv-orlando has joined #openstack-nova02:38
mriedemyeah, so we made the reservation in the nova db, but put the instance in the nova_cell0 db, and right when we delete i bet we find it in cell0 via the instance mapping02:38
*** Jeffrey4l_ has joined #openstack-nova02:39
mriedemhttps://github.com/openstack/nova/blob/master/nova/compute/api.py#L170802:40
melwittyeah, I'm looking in compute/api.py and I'm not seeing any freeing of quota for local deletes02:40
mriedemi think deleted would be False there02:40
mriedembecause conductor deleted the build request02:40
mriedemwe delete the build request in _bury_in_cell0 here https://github.com/openstack/nova/blob/master/nova/conductor/manager.py#L88802:40
melwittyeah. so far it's looking to me like we're missing a quotas.commit() for the local delete cases02:41
*** jamesdenton has joined #openstack-nova02:41
melwittlike if we're deleting from here https://github.com/openstack/nova/blob/master/nova/compute/api.py#L179902:41
melwittthis would be if it never landed on a host02:42
*** salv-orlando has quit IRC02:42
melwittwe're not doing anything with quotas until L1842 but you have to have reserve and commit for a delete too02:43
mriedemyeah i was looking at _lookup_instance https://github.com/openstack/nova/blob/master/nova/compute/api.py#L166002:43
*** zerda4 has quit IRC02:43
mriedemand we pull it from cell0 here https://github.com/openstack/nova/blob/master/nova/compute/api.py#L169002:43
mriedemlooks like we need something like this https://github.com/openstack/nova/blob/master/nova/compute/api.py#L1729-L175602:44
melwittI see that _delete_while_booting takes care of quotas correctly, it's just we're missing any quota handling at all at L179902:44
melwittyes, agree02:45
*** phuongnh has joined #openstack-nova02:45
*** jamesdenton has quit IRC02:45
*** rmk has joined #openstack-nova02:45
*** rmk has quit IRC02:45
*** rmk has joined #openstack-nova02:45
mriedemshould be relatively straightforward02:47
melwittyeah, I think so too02:48
*** rfolco has joined #openstack-nova02:49
*** rfolco has quit IRC02:50
*** ducttape_ has joined #openstack-nova02:50
melwittthanks Kevin_Zheng for alerting about that02:50
Kevin_Zhengmelwitt: your are welcome, still try to catch all you guys discussed :)02:51
*** raunak has quit IRC02:51
melwittKevin_Zheng: mriedem added a comment to the lp bug to explain what was concluded02:52
melwittit turned out to be something pretty simple02:52
melwittthat particular local delete case is missing quota cleanup02:53
openstackgerritMatt Riedemann proposed openstack/nova master: WIP: Decrement quota when deleting an instance in cell0  https://review.openstack.org/44339502:53
mriedemmelwitt: Kevin_Zheng: ^ w/o any tests run or anything02:53
*** dave-mcc_ has quit IRC02:54
*** thorst has joined #openstack-nova02:54
*** amrith has quit IRC02:54
*** yuntongjin has quit IRC02:54
*** armax has joined #openstack-nova02:54
*** thorst has quit IRC02:54
*** namnh_ has joined #openstack-nova02:54
*** ducttape_ has quit IRC02:55
*** amrith has joined #openstack-nova02:55
*** namnh has quit IRC02:56
Kevin_Zhenghttps://review.openstack.org/#/c/44300302:56
Kevin_Zheng^ seems we have already tried this but didnt work?02:56
*** amotoki has joined #openstack-nova02:57
melwittgah, I didn't look at the patch first02:58
*** yuntongjin has joined #openstack-nova02:58
Kevin_Zhengwe tested manually02:59
*** huangtianhua has joined #openstack-nova02:59
Kevin_Zhengfix like this actually decrease usage in cell002:59
melwittI'm trying to understand why that code is going to cell002:59
*** hongbin has joined #openstack-nova03:00
melwittmriedem: ^03:01
mriedemyeah idk03:01
mriedemare we mutating the context somewhere?03:01
melwittI thought probably not. I'm not sure if we yet merged the thing method that targets a context and lets it continue on targeted03:03
melwittand also I didn't think it would be used in this case03:03
mriedemthat definitely came later03:08
mriedemset_target_cell03:08
melwittgrr this doesn't make sense03:09
*** amotoki has quit IRC03:09
*** nic has quit IRC03:10
*** nicolasbock has joined #openstack-nova03:10
mriedemi would have to recreate locally and tinker with this,03:12
mriedemwhich i don't have time tonight03:12
mriedemKevin_Zheng: the instance is in cell0 though correct?03:12
mriedemand the build_request is deleted?03:12
*** sudipto has joined #openstack-nova03:13
*** sudipto_ has joined #openstack-nova03:14
Kevin_Zhengyes, we are have meeting now, but I will try to dig more if you guys didn't have tien03:15
Kevin_ZhengTime03:15
mriedemmelwitt: i think i found the problem03:15
melwittdo tell03:16
mriedemgather round children03:16
melwittheh03:16
mriedemso when the REST API looks up the instance via id03:16
mriedemit goes down to compute.API.get()03:16
*** thorst has joined #openstack-nova03:16
mriedemeventually to _get_instance03:16
mriedemwe get the cell0 instance mapping here https://github.com/openstack/nova/blob/fa2b4a82648101826566da68dd56d204e269853f/nova/compute/api.py#L229203:17
Kevin_ZhengAh03:17
*** thorst has quit IRC03:17
mriedemand then we mutate the context here https://github.com/openstack/nova/blob/fa2b4a82648101826566da68dd56d204e269853f/nova/compute/api.py#L229403:17
mriedemand forever after, that context object is targeted to cell003:17
*** bkopilov_ has quit IRC03:17
melwittargh, so we are already using set_target_cell03:17
Kevin_ZhengSeems right03:18
*** ducttape_ has joined #openstack-nova03:18
*** nicolasbock has quit IRC03:19
*** ducttape_ has quit IRC03:19
mriedemdansmith: ^ set_target_cell might be screwing us here03:20
mriedemmelwitt: on master yes03:20
mriedembut not ocata03:20
*** ducttape_ has joined #openstack-nova03:20
*** ducttape_ has joined #openstack-nova03:20
melwittI wonder if we should avoid using it until everything that needs to move to api db moves, or try to just patch around it for now?03:20
mriedemwe still have the bug in ocata03:21
mriedembut i think that's the reason why the guys patch didn't work03:21
mriedemand decremented from cell003:21
dansmiththe only time I've experienced this is if the main db for the control services is set to cell003:21
dansmithit shouldn't happen with a single cell and database set to that db right?03:21
mriedemi believe this is single node devstack03:22
Kevin_ZhengWe meet this in devstack env, didn't test in other cases yet03:22
melwittdansmith: this is the quota_usages and reservations tables, so it's updating usages in cell003:22
mriedemdansmith: yeah see https://review.openstack.org/44339503:23
melwittafter set_target_cell targets the context to cell003:23
mriedemwhich doesn't work because it tries to decrement from cell0 because the context is mutated to use cell0 when we lookup the instance here https://github.com/openstack/nova/blob/fa2b4a82648101826566da68dd56d204e269853f/nova/compute/api.py#L229403:23
*** markvoelker has joined #openstack-nova03:24
dansmithI'm confused, nothing ever moves between cell0 and another cell..03:24
mriedemwe make the reservation on quota in the nova db in the api03:24
mriedemthen conductor fails to schedule and creates the instance in cell003:25
mriedemand deletes the build request03:25
dansmithah03:25
mriedemwhen you go to delete the instance, we dont find the build request,03:25
mriedembut we find the instance mapping to cell003:25
melwittdansmith: this is a delete of an ERROR instance in cell0. the REST API does a get of the instance, which set_target_cell to cell0 on the context, then that same context is passed along to quotas.commit() and it decrements usage in the wrong db03:25
mriedemand we delete the instance from the cell0 db03:25
mriedembut we don't decrement the quota03:25
mriedemmelwitt: well, w/o https://review.openstack.org/#/c/443395/ we don't even attempt to decrement quota03:25
melwittdansmith: (after the person tried to add the missing quota cleanup)03:25
dansmithand this is because we don't have the reservations in the api I guess?03:25
mriedemdansmith: yup03:25
*** raunak has joined #openstack-nova03:25
dansmithawesome03:26
*** ssurana has quit IRC03:26
melwittmriedem, dansmith: right. so 1) we're missing quota cleanup 2) when we add quota cleanup, it cleans the wrong db because the context was targeted to cell0 early on03:27
dansmithhas melwitt pointed out that counting fixes all this nonsense?03:27
mriedemyes03:27
dansmithokay good03:27
melwittdansmith: yeah. but we need a backportable fix03:27
mriedembut we can't backprot that to ocata03:27
openstackgerritTony Breeds proposed openstack/os-vif master: Add Constraints support  https://review.openstack.org/41332503:27
mriedemso we need 2 fixes03:27
dansmithyeah, so we probably need a hacky thing to just go find the cell a thing is in (which will only ever be two places to check) right?03:27
mriedemthe set_target_cell one for master only,03:27
dansmithfor ocata03:27
mriedemand then https://review.openstack.org/#/c/443395/ or the other patch for ocata03:27
mriedemwe know what cell it's in already03:27
mriedemand we delete the instance, that's all fine03:27
mriedemwe just never decrement quota in that path, which https://review.openstack.org/#/c/443395/ does03:27
mriedemit just doesn't work on master b/c of set_target_cell03:28
dansmithokay03:28
dansmithyeah03:28
mriedemwas there a reason we needed https://github.com/openstack/nova/blob/fa2b4a82648101826566da68dd56d204e269853f/nova/compute/api.py#L2294 ?03:28
melwittmriedem: yeah, I was thinking consider taking the first patch for ocata https://review.openstack.org/#/c/44300303:28
mriedemlike, can't we just do the context manager there?03:28
dansmithmriedem: correct,03:28
dansmithmriedem: because the caller will need to do things, like later it might need to lazy-load fault for example03:29
dansmithwhich happens deep in the view03:29
*** markvoelker has quit IRC03:29
dansmithlong after we've returned from here and have lost all cell scope03:29
dansmithor instance_info_cache sometimes03:29
mriedemok, so i guess we have to hacky the quota decrement in the delete path03:30
mriedemsuch that it's untargeted03:30
dansmithand get is used to look up a thing which is later passed to something else that will make an rpc call, etc03:30
dansmithanyway03:30
dansmithyeah,03:30
dansmithI have a hard stop now, but happy to work on that thing tomorrow03:30
mriedemi'll roll something on top of the  original fix and he can see if it works03:31
mriedembecause i want to stop too03:31
dansmithmelwitt: this is a reason to unblock your bottom two patches right?03:31
melwittdansmith: yes, I think so03:32
melwittwell, no03:32
*** amotoki has joined #openstack-nova03:32
melwittdansmith: my patches don't move the usages and reservations tables. just leaves them out03:32
dansmithoh right03:32
melwittdansmith: sorry, doesn't move the data03:32
dansmithokay, anyway .. tomorrow03:33
melwittit might still work because we have the tables and the usage might be addressed by the quota resyncing stuff03:33
*** yamahata has quit IRC03:34
*** vladikr has quit IRC03:35
melwittlike, data migration was not supposed to be needed for the usages. have to read over the original spec again03:35
*** karthiks has quit IRC03:36
*** sudipto has quit IRC03:37
*** sudipto_ has quit IRC03:37
*** salv-orlando has joined #openstack-nova03:39
*** jamesdenton has joined #openstack-nova03:41
*** salv-orlando has quit IRC03:43
*** gjayavelu has joined #openstack-nova03:44
*** jamesden_ has joined #openstack-nova03:44
*** jamesdenton has quit IRC03:45
melwittyeah, nevermind. that wouldn't work at all bc no methods added to read/write usages and reservations in api db03:45
*** iceyao_ has quit IRC03:46
*** amotoki has quit IRC03:46
openstackgerritMatt Riedemann proposed openstack/nova master: WIP: Decrement quota when deleting an instance in cell0  https://review.openstack.org/44339503:48
openstackgerritMatt Riedemann proposed openstack/nova master: WIP: Temporarily untarget context when deleting from cell0  https://review.openstack.org/44340303:48
mriedemKevin_Zheng: melwitt: ^ this is pretty ugly, but it's what i'm thinking before going to bed03:48
*** fragatina has quit IRC03:48
*** gjayavelu has quit IRC03:48
*** karthiks has joined #openstack-nova03:49
Kevin_Zhengmriedem: Thanks I will test it :)03:50
Kevin_Zhengmriedem: have a nice sleep03:50
mriedemo/03:50
*** mriedem is now known as mriedem_away03:51
melwittmriedem_away: okay, looking03:52
*** diga has joined #openstack-nova03:55
openstackgerritMatt Riedemann proposed openstack/nova master: WIP: Start removing _lookup_instance  https://review.openstack.org/44340503:57
mriedem_awaymelwitt: one last thing that i thought of while looking through all of this ^03:57
mriedem_awaywe can start unwinding the super complicated spaghetti code that was there for the transition03:57
melwittokay, cool03:57
mriedem_awaynow i'm really leaving03:58
mriedem_awayo/03:58
melwittgnight o/03:58
*** namnh_ has quit IRC03:59
*** hongbin has quit IRC03:59
*** udesale has joined #openstack-nova03:59
*** amotoki has joined #openstack-nova04:02
*** jamesden_ has quit IRC04:04
*** ijw has joined #openstack-nova04:07
*** iceyao has joined #openstack-nova04:07
*** amotoki has quit IRC04:10
*** ratailor has joined #openstack-nova04:10
*** iceyao has quit IRC04:11
*** jamielennox is now known as jamielennox|away04:13
*** KeithMnemonic has quit IRC04:14
*** bkopilov has joined #openstack-nova04:17
*** links has joined #openstack-nova04:17
*** thorst has joined #openstack-nova04:17
*** crushil has joined #openstack-nova04:20
*** bkopilov has quit IRC04:22
*** thorst has quit IRC04:22
*** amotoki has joined #openstack-nova04:24
*** sree has joined #openstack-nova04:24
*** markvoelker has joined #openstack-nova04:25
*** mdnadeem has joined #openstack-nova04:28
*** ducttape_ has quit IRC04:28
*** markvoelker has quit IRC04:29
*** moshele has joined #openstack-nova04:32
*** bkopilov has joined #openstack-nova04:33
*** salv-orlando has joined #openstack-nova04:39
*** moshele has quit IRC04:42
*** salv-orlando has quit IRC04:44
*** gouthamr has quit IRC04:46
*** denaitre has quit IRC04:47
*** iceyao has joined #openstack-nova04:47
*** jamielennox|away is now known as jamielennox04:47
*** iceyao_ has joined #openstack-nova04:51
*** iceyao has quit IRC04:51
*** yuntongjin has quit IRC04:52
*** psachin has joined #openstack-nova04:54
*** iceyao has joined #openstack-nova04:54
*** sam_nowitzki has joined #openstack-nova04:55
*** iceyao has quit IRC04:55
*** iceyao_ has quit IRC04:56
*** iceyao has joined #openstack-nova04:56
*** iceyao_ has joined #openstack-nova04:57
*** iceyao has quit IRC04:57
*** iceyao_ has quit IRC05:00
*** iceyao has joined #openstack-nova05:00
*** gjayavelu has joined #openstack-nova05:00
*** iceyao_ has joined #openstack-nova05:03
*** iceyao has quit IRC05:05
*** iceyao_ has quit IRC05:06
*** iceyao has joined #openstack-nova05:06
*** iceyao_ has joined #openstack-nova05:07
*** iceyao has quit IRC05:07
*** ijw has quit IRC05:08
*** gjayavelu has quit IRC05:09
*** iceyao_ has quit IRC05:09
*** iceyao has joined #openstack-nova05:10
*** diga has quit IRC05:10
*** iceyao_ has joined #openstack-nova05:13
*** iceyao has quit IRC05:14
*** iceyao_ has quit IRC05:14
*** sridharg has joined #openstack-nova05:14
*** iceyao has joined #openstack-nova05:14
*** unicell has joined #openstack-nova05:14
*** ssurana has joined #openstack-nova05:16
*** gongysh has quit IRC05:16
*** iceyao_ has joined #openstack-nova05:17
*** unicell1 has joined #openstack-nova05:18
*** fragatina has joined #openstack-nova05:18
*** thorst has joined #openstack-nova05:18
*** iceyao__ has joined #openstack-nova05:18
*** unicell has quit IRC05:19
*** iceyao has quit IRC05:19
*** ayogi has joined #openstack-nova05:20
*** fragatin_ has joined #openstack-nova05:20
openstackgerritChangBo Guo(gcb) proposed openstack/nova master: Enable more virt.xenapi tests on Python 3  https://review.openstack.org/41442105:21
*** iceyao__ has quit IRC05:22
*** iceyao has joined #openstack-nova05:22
*** iceyao_ has quit IRC05:22
*** fragatina has quit IRC05:22
*** thorst has quit IRC05:23
*** fragatin_ has quit IRC05:25
*** fragatina has joined #openstack-nova05:25
*** sudipto has joined #openstack-nova05:25
*** iceyao_ has joined #openstack-nova05:25
*** markvoelker has joined #openstack-nova05:25
*** iceyao has quit IRC05:26
*** iceyao has joined #openstack-nova05:27
*** ducttape_ has joined #openstack-nova05:28
*** fragatina has quit IRC05:29
*** iceyao_ has quit IRC05:29
*** jose-phillips has quit IRC05:30
*** gjayavelu has joined #openstack-nova05:30
*** iceyao has quit IRC05:30
*** markvoelker has quit IRC05:30
*** iceyao has joined #openstack-nova05:31
*** fragatina has joined #openstack-nova05:32
*** iceyao has quit IRC05:32
*** iceyao has joined #openstack-nova05:32
*** ducttape_ has quit IRC05:32
*** fragatina has quit IRC05:33
*** eliqiao has quit IRC05:33
*** iceyao_ has joined #openstack-nova05:34
*** gjayavelu has quit IRC05:35
*** yamahata has joined #openstack-nova05:36
*** fragatina has joined #openstack-nova05:36
*** eliqiao has joined #openstack-nova05:36
*** iceyao has quit IRC05:37
*** iceyao__ has joined #openstack-nova05:37
*** prateek has joined #openstack-nova05:38
*** iceyao__ has quit IRC05:38
*** iceyao has joined #openstack-nova05:38
*** iceyao_ has quit IRC05:38
*** iceyao has quit IRC05:40
*** iceyao has joined #openstack-nova05:40
*** salv-orlando has joined #openstack-nova05:40
*** fragatina has quit IRC05:41
*** fragatina has joined #openstack-nova05:42
*** iceyao_ has joined #openstack-nova05:43
*** iceyao has quit IRC05:45
*** salv-orlando has quit IRC05:45
openstackgerritEdu Alcaniz proposed openstack/nova master: follow-up  https://review.openstack.org/44342405:45
openstackgerritjichenjc proposed openstack/nova master: Trivial Fix  https://review.openstack.org/44231005:46
*** iceyao_ has quit IRC05:48
*** armax has quit IRC05:50
*** iceyao has joined #openstack-nova05:51
*** iceyao_ has joined #openstack-nova05:53
*** iceyao__ has joined #openstack-nova05:55
*** iceyao has quit IRC05:55
*** iceyao has joined #openstack-nova05:57
*** iceyao_ has quit IRC05:57
*** iceyao__ has quit IRC05:57
*** trinaths has joined #openstack-nova05:58
*** nkrinner_afk is now known as nkrinner05:59
*** gongysh has joined #openstack-nova06:01
*** avolkov has joined #openstack-nova06:02
*** links has quit IRC06:04
*** gjayavelu has joined #openstack-nova06:04
openstackgerritjichenjc proposed openstack/nova master: Add description for Image location in snapshot  https://review.openstack.org/44037906:06
*** gjayavelu has quit IRC06:09
*** crushil has quit IRC06:10
*** edmondsw has joined #openstack-nova06:13
openstackgerritjichenjc proposed openstack/nova master: check hw:mem_page_size when change extra_spec  https://review.openstack.org/40861006:14
*** karthiks has quit IRC06:14
openstackgerritjichenjc proposed openstack/nova master: no instance info cache update if instance deleted  https://review.openstack.org/36358506:14
*** links has joined #openstack-nova06:16
*** edmondsw has quit IRC06:17
*** yuntongjin has joined #openstack-nova06:18
*** thorst has joined #openstack-nova06:19
*** satyar has joined #openstack-nova06:22
*** thorst has quit IRC06:24
*** moshele has joined #openstack-nova06:27
*** ducttape_ has joined #openstack-nova06:29
*** trinaths1 has joined #openstack-nova06:29
*** trinaths has quit IRC06:29
*** raunak has quit IRC06:30
*** diga has joined #openstack-nova06:31
*** ducttape_ has quit IRC06:33
*** salv-orlando has joined #openstack-nova06:33
*** yuntongjin1 has joined #openstack-nova06:36
*** yuntongjin has quit IRC06:36
*** gjayavelu has joined #openstack-nova06:39
*** jose-phillips has joined #openstack-nova06:39
*** armax has joined #openstack-nova06:39
*** ssurana has quit IRC06:42
*** rcernin has joined #openstack-nova06:44
*** ssurana has joined #openstack-nova06:45
*** gjayavelu has quit IRC06:48
*** Guest34860 has quit IRC06:51
openstackgerritChangBo Guo(gcb) proposed openstack/nova master: Add volume status check when deleting instances  https://review.openstack.org/26033906:53
*** links has quit IRC06:58
*** amotoki_ has joined #openstack-nova07:01
*** lpetrut has joined #openstack-nova07:02
*** amotoki has quit IRC07:03
*** ssurana has quit IRC07:06
*** armax has quit IRC07:06
*** guchihiro has joined #openstack-nova07:07
*** jamielennox is now known as jamielennox|away07:09
*** ducnc has joined #openstack-nova07:10
*** sandanar has joined #openstack-nova07:12
*** jose-phillips has quit IRC07:12
*** diga has quit IRC07:13
*** yamamoto has quit IRC07:14
*** links has joined #openstack-nova07:15
*** jose-phillips has joined #openstack-nova07:20
*** thorst has joined #openstack-nova07:20
*** arne_r has joined #openstack-nova07:21
*** andreas_s has joined #openstack-nova07:22
*** salv-orl_ has joined #openstack-nova07:22
*** tesseract has joined #openstack-nova07:22
*** yuntongjin1 has quit IRC07:22
*** CristinaPauna has quit IRC07:23
*** thorst has quit IRC07:24
*** yuntongjin has joined #openstack-nova07:24
*** salv-orlando has quit IRC07:25
*** karthiks has joined #openstack-nova07:27
*** jamielennox|away is now known as jamielennox07:27
*** CristinaPauna has joined #openstack-nova07:29
*** ducttape_ has joined #openstack-nova07:30
*** databus23_ has joined #openstack-nova07:34
*** mkoderer_ has joined #openstack-nova07:34
*** david_1 has joined #openstack-nova07:34
*** ducttape_ has quit IRC07:35
*** databus23_ has quit IRC07:36
*** mkoderer_ has quit IRC07:36
*** david_1 has quit IRC07:36
*** dimtruck is now known as zz_dimtruck07:37
*** tbachman has joined #openstack-nova07:39
*** tbachman has quit IRC07:44
*** markus_z has joined #openstack-nova07:44
*** yuntongjin1 has joined #openstack-nova07:48
*** yuntongjin has quit IRC07:48
arne_rmarkus_z: Hi, care to give this https://review.openstack.org/#/c/436868/ your blessing?07:49
markus_zarne_r: I'm going to check it within the next hours07:50
arne_rthanks =)07:52
*** kaisers_ has joined #openstack-nova07:54
*** ralonsoh has joined #openstack-nova07:54
*** Jack_I has joined #openstack-nova07:58
*** lpetrut has quit IRC08:02
*** guchihiro has quit IRC08:05
*** jpena|off is now known as jpena08:05
*** ltomasbo|away is now known as ltomasbo08:11
*** mlakat has joined #openstack-nova08:13
*** mlakat has quit IRC08:13
*** mlakat has joined #openstack-nova08:13
*** diga has joined #openstack-nova08:14
*** mlakat_ has joined #openstack-nova08:14
*** lpetrut has joined #openstack-nova08:17
*** mlakat has quit IRC08:18
*** iceyao has quit IRC08:18
*** sofie_ has joined #openstack-nova08:18
*** jpena is now known as jpena|off08:19
*** thorst has joined #openstack-nova08:21
*** fragatina has quit IRC08:22
*** thorst has quit IRC08:25
*** jpena|off is now known as jpena08:26
Kevin_Zhengmelwitt, mriedem maybe it is good to add tempest test for this: create instance -> error -> delete -> check quota  after this fixed, to avoid same thing happen again08:26
Kevin_Zheng?08:26
*** yassine has joined #openstack-nova08:27
*** jaosorior has joined #openstack-nova08:28
*** abalutoiu_ has quit IRC08:28
*** mnestratov has joined #openstack-nova08:30
*** ducttape_ has joined #openstack-nova08:30
*** mlakat_ has quit IRC08:30
*** mlakat has joined #openstack-nova08:31
markus_zarne_r: It needs a small change, IMO08:33
*** irenab has quit IRC08:33
*** mlakat has quit IRC08:34
*** abalutoiu has joined #openstack-nova08:34
*** mlakat has joined #openstack-nova08:34
*** psachin has quit IRC08:35
*** mlakat has quit IRC08:35
*** ducttape_ has quit IRC08:35
openstackgerritmelanie witt proposed openstack/nova master: Count resources to check quota for cells  https://review.openstack.org/41652108:41
openstackgerritmelanie witt proposed openstack/nova master: Make Quotas object favor the API database  https://review.openstack.org/41094508:41
openstackgerritmelanie witt proposed openstack/nova master: Add online migration to move quotas to API database  https://review.openstack.org/41094608:41
*** irenab_ has joined #openstack-nova08:42
*** markvoelker has joined #openstack-nova08:46
*** lpetrut has quit IRC08:47
*** psachin has joined #openstack-nova08:49
*** jahsis has joined #openstack-nova08:50
openstackgerritQiming Teng proposed openstack/nova master: Fix API doc about server attributes (2.3 API)  https://review.openstack.org/44300008:52
*** gjayavelu has joined #openstack-nova08:58
*** iceyao has joined #openstack-nova08:58
*** sree_ has joined #openstack-nova08:59
openstackgerritDave Chen proposed openstack/nova master: WIP - Recover instance's task state after exception  https://review.openstack.org/44350908:59
*** zzzeek has quit IRC09:00
*** sree_ is now known as Guest9948409:00
*** zzzeek has joined #openstack-nova09:02
*** gjayavelu has quit IRC09:02
*** sree has quit IRC09:03
*** amoralej|off is now known as amoralej09:07
*** mlakat has joined #openstack-nova09:07
*** lucas-afk is now known as lucasagomes09:09
*** yuntongjin1 has quit IRC09:10
*** yuntongjin has joined #openstack-nova09:10
*** karimb has joined #openstack-nova09:14
*** hshiina has joined #openstack-nova09:18
*** alexpilotti has quit IRC09:19
*** jogo has quit IRC09:19
*** alexpilotti has joined #openstack-nova09:20
*** thorst has joined #openstack-nova09:22
openstackgerritZhenyu Zheng proposed openstack/nova master: Don't delete neutron port when attach failed  https://review.openstack.org/40356809:22
*** abalutoiu has quit IRC09:22
*** jogo has joined #openstack-nova09:24
*** Dave____ has joined #openstack-nova09:24
*** namnh has joined #openstack-nova09:26
*** haplo37 has quit IRC09:28
*** Dave has quit IRC09:29
*** Dave____ is now known as Dave09:29
*** thorst has quit IRC09:31
*** ducttape_ has joined #openstack-nova09:31
*** Shunli has quit IRC09:31
*** ducttape_ has quit IRC09:36
*** yingjun has quit IRC09:37
*** haplo37 has joined #openstack-nova09:38
*** kevinz has quit IRC09:39
*** abalutoiu has joined #openstack-nova09:41
*** yamahata has quit IRC09:48
*** edmondsw has joined #openstack-nova09:50
*** youzp_ is now known as youzp09:52
*** gszasz has joined #openstack-nova09:53
*** psachin has quit IRC09:54
*** edmondsw has quit IRC09:54
openstackgerritStephen Finucane proposed openstack/nova-specs master: Reserve NUMA nodes with PCI devices attached  https://review.openstack.org/36446809:57
*** efoley has joined #openstack-nova10:00
asettlejohnthetubaguy: morning! You around yet? :)10:04
*** psachin has joined #openstack-nova10:06
*** namnh has quit IRC10:06
johnthetubaguyasettle: yup10:07
johnthetubaguyasettle: hows things?10:07
asettlejohnthetubaguy: less than awesome rn. I've just had to move up one of our nova-related bugs to critical. I will be sending out to the ML the details but I was just going to see if you could help me get all hands on deck.10:07
johnthetubaguyasettle: you got a link?10:08
asettlehttps://bugs.launchpad.net/openstack-manuals/+bug/166348510:08
openstackLaunchpad bug 1663485 in openstack-manuals "Document Nova v2 Cells and Placement API" [Critical,In progress] - Assigned to Brian Moss (bmoss)10:08
asettleRelevant patch: https://review.openstack.org/#/c/438328/10:08
asettleIt still doesn't work.10:08
asettleWe need to branch, but we can't because the guide is broken at the moment as a result of this.10:08
johnthetubaguyasettle: doesn't work after you try with https://review.openstack.org/#/c/438328/ added?10:10
asettlejohnthetubaguy: My understanding is that it does not.10:10
*** hshiina has quit IRC10:10
johnthetubaguyah, I see the comment from bmoss now10:11
bauzasasettle: see the comment from Dimitry10:11
bauzasoops10:11
johnthetubaguyasettle: the people you want for this are melwitt and dansmith really, both west coast10:11
*** tovin07 has quit IRC10:11
bauzass/Dimitry/Georgios10:11
bauzasI can try to help10:12
johnthetubaguybauzas: cool10:12
johnthetubaguyit is RDO related10:12
*** tovin07_ has quit IRC10:12
asettlebauzas: would really appreciate that. You understand it doesn't really matter right now who did what, the fact is it doesn't work and we need all the assistance we can get - so, TL;DR - would super appreciate that.10:12
asettlebauzas: where is the commetn from GEorgios?10:13
bauzasasettle: https://bugs.launchpad.net/openstack-manuals/+bug/1663485/comments/2010:13
openstackLaunchpad bug 1663485 in openstack-manuals "Document Nova v2 Cells and Placement API" [Critical,In progress] - Assigned to Brian Moss (bmoss)10:13
asettleOh in the bug ;)10:13
*** kiseok7 has joined #openstack-nova10:13
bauzasyeah10:13
asettlebauzas: yep I had seen that. Fortunately, Georgios error turned out well.10:14
bauzasasettle: the RDO issue is tracked by https://bugzilla.redhat.com/show_bug.cgi?id=140509810:14
openstackbugzilla.redhat.com bug 1405098 in openstack-tripleo-heat-templates "Ocata requires additional cellvs2 setup" [High,On_qa] - Assigned to owalsh10:14
asettleHowever, Thiago's comments bring up another set of issues in the docs and Brian's one too.10:14
bauzasowalsh: around ?10:14
asettlebauzas: yep I'd been made aware of that :)10:14
bauzasowalsh is one of our top-notch engineers responsible for integrating Nova features into TripleO10:14
johnthetubaguyI don't see any logs about what is going wrong, I guess10:15
bauzasthe problem I have is that I'm blind of the left problems10:15
johnthetubaguyyeah, same10:15
asettlebauzas: awesome :)10:15
bauzasI would love to see what's missing or still having problems so we could discuss on what's wrong10:15
johnthetubaguythere is no info to help debug the issues, apart from go reproduce it I guess10:15
johnthetubaguyand hope you see the same problem10:16
asettlejohnthetubaguy: pretty much. I'm still going to reach out to the ML to get some reviews and help outside of the nova team.10:16
asettleWhat else can I do?10:16
johnthetubaguyif we could get logs from bmoss, that would really help us10:16
asettleBrian has gone to bed, and Amy (spotz) who is the other person experiencing the issue won't be up for a bit (Austin)10:16
bauzasupstream documentation is one thing, deployment procedures for a specific tool is another10:16
asettlejohnthetubaguy: let me try and contact him. I know in theory he's awake.10:16
johnthetubaguyOK, I don't want to get him out of bed, just saying we don't really have anything to work on here10:17
asettlebauzas: very much so. andymccr offered to help too as he got it working in OSA. But I'm not too sure how mcuh it would help or not (I mean, I ain't knocking hte offer back)10:17
bauzasso, here, we could try to see what are the discrepancies between upstream docs and how TripleO gates nova10:17
asettleNah, it's only 8pm there10:17
asettleUno momenot10:17
bauzasasettle: you're EU based ?10:17
asettlebauzas: yes, London :)10:17
bauzasokay10:17
bauzasowalsh is Irish and I'm French10:17
asettleExcellent :D10:17
bauzasso we can somehow sort that out10:17
andymccryeah if we compare it to OSA or TripleO we should be able to get it going10:17
bauzasnot sure it's excellent to be French, but meh :p10:18
johnthetubaguywell, can compare it to the in tree docs too10:18
asettleI am happy to spin up an instance and start replicating if I can't get bmoss to give me his error logs rn10:18
*** tuanluong has quit IRC10:18
asettlebauzas: it's pretty great right now :P10:18
bauzasso, the question I have is10:19
johnthetubaguyandymccr: did we get that ordering issue sorted?10:19
bauzaswithout using a deployment tool, are we able to spin up an instance by following the upstream documentation ?10:19
bauzasand if not, what is the error ?10:19
asettlebauzas: that's basically the problem. No, we cannot.10:19
asettlespotz experienced this, and so did bmoss10:19
johnthetubaguyusing the install guide, it doesn't work right now, right?10:20
bauzasasettle: do you have ways to reproduce that so you could throw me the exception ?10:20
asettleI've just text Brian now to see if he can shoot me his logs (I'm sure his wife will be hatin')10:20
andymccrjohnthetubaguy: yes sir - we now do a "api_db version" before attempting the sync - if it's 0 --> sync, if it's not --> don't sync just create cell.10:20
andymccr(and sync later)10:20
asettlebauzas: not at the moment, no. Will work on getting you more information. I don't have the instance up, Brian does. Let me get back to you :)10:20
asettle(I hope?)10:20
bauzasBrian is ET-based ?10:20
asettlebauzas: Australia10:20
bauzasmeh okay10:21
asettleSo, complete opposite timezone.10:21
bauzasnot sure we could get him10:21
johnthetubaguyandymccr: I wonder if the docs have that issue, they seem to not do api_db sync10:21
asettleIt's after 8pm - not too late.10:21
andymccrjohnthetubaguy: hmm maybe10:21
andymccrjohnthetubaguy: on a new build you need it first10:21
*** iceyao has quit IRC10:21
*** edmondsw has joined #openstack-nova10:22
asettlebauzas: he's awake ;) he's just turning on his computer to send me the logs for you10:23
*** satyar has quit IRC10:23
*** karimb has quit IRC10:23
asettle"I updated the patch (with the comments from mriedem_away) but haven't tested the revision yet. I'll send you my nova logs."10:23
asettleSo, hopefully we can see what the original problem was, and move from there.10:24
bauzas\o/10:24
bauzasI'll have to bail out for a couple of hours10:24
bauzasbut I'll be back around 1pm UTC10:24
asettlebauzas: no problem :) thank you for offering ot help. I've sent out to the ML too. Hopefully we can get some people looking at it too.10:24
asettleSorry andymccr and johnthetubaguy - not ignoring you both :P10:26
asettleGetting the error logs from bmoss rn10:26
*** edmondsw has quit IRC10:26
* johnthetubaguy nods10:28
*** Jack_I has quit IRC10:28
asettleUgh *head desk*10:28
* asettle waves hand outside10:28
asettleAnd it was such a nice day too10:28
*** Jack_I has joined #openstack-nova10:29
bauzasalex_xu: around ?10:29
bauzasalex_xu: not understood your point in https://review.openstack.org/#/c/440580/3/specs/pike/approved/scheduler-hints-in-server-details.rst@3410:30
johnthetubaguyasettle: do we tell people to always install the controller first, and then the compute node?10:30
*** yuntongjin has quit IRC10:30
bauzas"controller" is a pure non-nova thing :)10:31
asettlejohnthetubaguy: good question, lemme just check historically.10:31
asettleYes, we do.10:31
johnthetubaguyOK, thats good10:32
bauzasbut yeah I remember the old docs, and it's the case10:32
* asettle is going back10:32
asettleYep10:32
asettleController -> Compute10:32
bauzasasettle: is the problem due to an upgrade or a greenfields install ?10:32
asettlebauzas: greenfields, I believe.10:32
bauzasbecause instructions are fairly different10:32
asettleWe just test the install instructions, we don't test upgrades.10:32
bauzasfor cells10:32
*** efoley has quit IRC10:33
asettlebauzas: that documentation for upgrades should belong in the upgrade docs, no?10:33
bauzasasettle: yup10:33
asettleYes, okay, I'm not going crazy then.10:33
*** udesale has quit IRC10:33
bauzasjust wanted to verify that bmoss is following the fresh install procedure, that's it10:33
bauzasbecause he's pointing to some internal BZ that mentions upgrades and TripleO hence my confusion10:33
bauzasboth are totally separate things10:34
*** lpetrut has joined #openstack-nova10:34
asettlebauzas: yes he should be. The whole idea behind doing the install guide testing ist hat you have to do a full manual install (which is also why we don't have people racing to do it)10:34
*** sambetts|afk is now known as sambetts10:35
bauzasyeah, I followed it in 2012 :p10:36
bauzasusing .deb packages :blushes:10:36
bauzasglad to see it continuing even if not a lot of people are crazy pants enough to install OpenStack by hand :)10:37
bauzasoh10:37
lyarwoodhttps://review.openstack.org/#/c/391597/ - pretty simple ~1k LOC removal change in favor of os-brick code if anyone has time this morning10:37
bauzasasettle: my reply to your email was bounced10:37
bauzasasettle: because openstack-docs@ is moderated10:37
asettlebauzas: by me!10:37
asettleLet me add you in10:38
bauzasstrange, why are you moderating it ?10:38
johnthetubaguyoh, did not know there was a docs ML10:38
asettlebauzas: cause i'm the PTL :p10:38
bauzaswell, in general, our lists are open10:38
asettlejohnthetubaguy: yeah we're weird like that. I dunno why we have it.10:38
asettleHence why I added the dev list into it, and the docs for others. I don't know the historical reason we have a separate ML10:38
asettleI think it was because we were not a part of the dev cycle for a while there10:38
asettlenow we are10:38
asettleSoooo ya know10:38
bauzashaving a separate ML is fine by me10:39
johnthetubaguythat does explain why I don't see any docs chatter on the ML10:39
asettlejohnthetubaguy: yep :) feel free to sign up.10:39
bauzashaving it moderated is what I'm worried about10:39
johnthetubaguywell, not much anyways10:39
asettleI occasionally post to it10:39
asettlebauzas: we get a *Lot* of spam, unforunatley10:39
bauzasanyway, /me needs to drop off, school pick-up10:39
asettleEverything is let through from people unless it's a clear spam email or bot (which, is unbelievably common)10:39
bauzassoooo \o see ya in a couple of hours10:39
owalshasettle: hi10:40
johnthetubaguyasettle: not sure, that kind of thing can make you accidentally make you a docs CPL10:40
asettleowalsh: hi!10:40
asettlebauzas: have fun!10:40
asettlejohnthetubaguy: *screams*10:40
* owalsh looks at https://review.openstack.org/43832810:40
asettlethank you owalsh !!10:41
*** Lambert has quit IRC10:42
* bauzas waves to owalsh and throws the baby to him :p10:42
*** Lambert has joined #openstack-nova10:42
bauzaskid's time, so \o10:42
asettlebauzas: hey woah it's hardly a baby :P10:42
*** gongysh has quit IRC10:44
owalshasettle: AFAIK need to run api_db sync before map_cell0 or create_cell10:45
asettleowalsh: the person to best answer that is bmoss as he has been testing. I'm just waiting for him to send me his error logs so can I appropriately address questions.10:46
asettleI am currently just escalating to ensure we have eyes on it :)10:46
*** ratailor has quit IRC10:46
*** mvk has quit IRC10:48
*** Lambert has quit IRC10:49
*** cdent has joined #openstack-nova10:49
*** Lambert has joined #openstack-nova10:49
*** sudipto has quit IRC10:51
*** derekh has joined #openstack-nova10:52
*** sam_nowitzki has quit IRC10:54
*** yuntongjin has joined #openstack-nova10:55
asettleowalsh and johnthetubaguy - Brian confirmed that he initially used the dev docs in-tree to install, and those didn't work either *head desk*10:57
gibihi nova! there is a high severity bug fix that needs core attention https://review.openstack.org/#/c/437222/ . Could you check it please?10:57
johnthetubaguyOK, that has worked for some people, at least apparently10:57
johnthetubaguygibi: its open in my browser, just chasing down this docs issue right now10:58
gibijohnthetubaguy: thanks!10:58
jaosoriorgibi: thanks for the fix; that actually breaks us.10:58
*** yuntongjin has quit IRC10:59
*** yuntongjin has joined #openstack-nova10:59
gibijaosorior: hi! breaks me too :)11:00
johnthetubaguygibi: there are related quota issues to that fix, I believe11:00
*** salv-orl_ has quit IRC11:00
johnthetubaguynot that your fix breaks quotas, just that you are pointing at a wider bug / lack of code sharing11:01
*** salv-orlando has joined #openstack-nova11:01
*** salv-orlando has quit IRC11:01
*** salv-orlando has joined #openstack-nova11:02
gibijohnthetubaguy: you mean this fix needs to be extended with some quota fix as well?11:02
johnthetubaguygibi: no, just there are other problems in that area I think11:02
johnthetubaguythere is a quota bug about that, someone was working on it11:02
johnthetubaguygibi: I would love dansmith or melwitt to take a look at that bug fix, they have better context around that code11:03
*** sree has joined #openstack-nova11:03
gibijohnthetubaguy: OK, I will ping them today11:03
johnthetubaguygibi: when I last checked it I was a bit worried about the cut and paste -ness of it, but not sure how to restructure it11:03
gibijohnthetubaguy: yeah, that code path is ugly. I was affraid to touch it as I felt I will break something if I try to factor out the duplication11:04
gibijohnthetubaguy: do you have the bug number for the quota bug? I would review that patch to learn something new11:05
*** Guest99484 has quit IRC11:05
johnthetubaguygibi: I don't was in chat previously, something about hitting over quota errors, and the quota never being removed11:06
asettlejohnthetubaguy: I see you pushed a patch! What's up?11:06
asettleJust replied to the ML with brian's logs11:06
asettlebauzas: ^^11:06
cdentasettle, johnthetubaguy: do we have a more clear picture of the context that leads to "wrong" and whatever "wrong" is with that docs stuff? It feels a bit like we're all guessing11:06
cdentoh good11:06
* gibi diggin launchpad11:06
cdentthat'll probably answer my questions, thank you asettle11:06
asettlecdent: yep take a look.11:06
johnthetubaguygibi: maybe just a function that does destroy and does the notification11:06
*** yuntongjin has quit IRC11:07
johnthetubaguyasettle: I was worried about the order not being very clear, but also very, very important, so I refactored the docs patch while we were waiting for the logs11:07
asettlecdent: the in-tree docs you guys have were also tested and although I don't have the error logs to give you, my understanding is that those instructions immediately error with "placement API is not available, you must configure the placement API."11:07
asettlejohnthetubaguy: you're a gem ;)11:07
gibijohnthetubaguy: this could be the quota bug https://bugs.launchpad.net/nova/+bug/167062711:07
openstackLaunchpad bug 1670627 in OpenStack Compute (nova) "quota is always in-use after delete the ERROR instances " [Critical,In progress] - Assigned to Matt Riedemann (mriedem)11:07
sean-k-mooneytonyb: hi you about? sorry my rely to your comment on https://review.openstack.org/#/c/413325/2/tools/tox_install.sh is a bit ranty11:08
johnthetubaguyasettle: did you include the dev list?11:08
asettlejohnthetubaguy: sure did11:08
johnthetubaguyhmm, not getting the email, will try the archive11:08
cdentasettle: I suspect that is because of this of the <Directory> thing (I added a link to an os-dev posting on the review)11:08
gibijohnthetubaguy: yeah that could be done. If that helps then I'm happy to do that11:09
sean-k-mooneytonyb: that said the fact my cygwin workflow for running tox is now broken really annoyed me when i disscovered it11:09
johnthetubaguygibi: makes it easier not to forget next time I think11:09
cdentthe reason things happen to work in devstack but not elsewhere is because keystone stuff is taking care of adding a <Directory> entry that covers placement too11:09
gibijohnthetubaguy: OK11:09
asettlejohnthetubaguy: http://lists.openstack.org/pipermail/openstack-dev/2017-March/113608.html11:09
asettlecdent: in the in-tree docs or the docs.o.o ones?11:10
cdentasettle: I added a link on the docs.o.o review, just as a reference point11:10
johnthetubaguyoh... maybe the docs don't talk about creating any flavors any more11:10
johnthetubaguythats quite a recent change, we no longer install any flavors by default, you have to create your own11:11
gibijohnthetubaguy: I have to rewoke that it would be easy to do see the block change in https://review.openstack.org/#/c/437222/5/nova/compute/api.py11:11
*** yuntongjin has joined #openstack-nova11:11
cdentasettle, johnthetubaguy: I need to coffee-up and read the logs, but I'll be around to try and help more soon11:11
johnthetubaguycdent: sweet11:11
asettlecdent: thanks dude :)11:11
johnthetubaguycdent: it sounds very familiar11:11
gibijohnthetubaguy: also it seems there are different error handling around instance.destroy11:12
*** kevinz has joined #openstack-nova11:12
gibijohnthetubaguy: sometimes we retry after a refresh, sometimes we don'tt11:12
gibijohnthetubaguy: here https://review.openstack.org/#/c/437222/5/nova/conductor/manager.py11:12
johnthetubaguyasettle: I am not seeing that placement related error anywhere in the logs11:13
johnthetubaguyasettle: do you know where you saw it?11:13
johnthetubaguyI see "libvirtError: Failed to connect socket to '/var/run/libvirt/libvirt-sock': Connection refused"11:14
asettlejohnthetubaguy: that placement error was specifically for the in-tree docs, unreferenced11:14
gibijohnthetubaguy: anyhow I will ping dansmith and melwitt later with this patch11:14
asettleThe error logs I sent you were from the instructions that bmoss tried in the patch11:14
* gibi runs for lunch11:14
asettlejohnthetubaguy: meant to be different things, sorry for confusing you11:14
*** karimb has joined #openstack-nova11:14
johnthetubaguyasettle: so the bug he is hitting seems to be listed here: https://github.com/mseknibilel/OpenStack-Grizzly-Install-Guide/issues/7111:15
johnthetubaguyasettle: seems unrelated to placement or cells11:15
openstackgerritLucian Petrut proposed openstack/nova master: Avoid using fdatasync() when fetching images  https://review.openstack.org/44358311:16
owalshasettle: I think the current instructions should work, I've added some nits but nothing significant11:16
asettleOkay, this is good. I'd like to get a few more eyes on it.11:17
asettleOne of my other concerns is that nobody seemed to agree in that patch set.11:17
asettleThere was a lot of differing opinions.11:17
asettleI might have to spin something up and test it, Brian has just gone to bed.11:17
asettleI'm sorry I can't answer all the questions - I'm speaking on behalf of someone.11:18
*** karimb has quit IRC11:18
cdentasettle: you're encountering that sad reality that seems to be coming up alot lately: Half of us making this stuff haven't got enough time to actually use it and test it :(11:18
*** kevinz has quit IRC11:19
*** sree has quit IRC11:19
johnthetubaguyasettle: afraid the logs don't help here, so someone is going to have to walk through the guide I guess?11:19
*** kevinz has joined #openstack-nova11:20
openstackgerritLucian Petrut proposed openstack/nova master: Avoid using fdatasync() when fetching images  https://review.openstack.org/44358311:20
*** hshiina has joined #openstack-nova11:21
*** JoseMello has joined #openstack-nova11:21
*** sofie_ has quit IRC11:21
asettlejohnthetubaguy: volunteering? :D11:22
asettleYou can do it and I can do it and i'll meet you in the middle11:22
johnthetubaguyasettle: I am curious what your take is on the doc change after this: https://review.openstack.org/#/c/438328/7/doc/install-guide/source/nova-controller-install.rst11:22
asettlecdent: unfortunately, yep. I've been chatting to mriedem_away about getting a dedicated CPL for this stuff. Nova is one of the only projects that don't have one.11:22
asettlejohnthetubaguy: where specifically?11:23
johnthetubaguyasettle: I don't think I can, I am blocking way too many people to do none of my other work I had planned, bauzas sounded like he might11:23
johnthetubaguyasettle: I basically moved all the new instructions into the existing flow11:23
asettlejohnthetubaguy: okay that's a good start. My day is relatively clear :)11:24
johnthetubaguyasettle: this is the bulk of the change https://review.openstack.org/#/c/438328/4..7/doc/install-guide/source/nova-controller-install.rst11:24
asettlejohnthetubaguy: okay, lemme do the thing.11:25
asettleLet's see if I can get lucky.11:26
*** aloga_ has joined #openstack-nova11:27
*** mvk has joined #openstack-nova11:27
* johnthetubaguy crosses fingers11:28
*** thorst has joined #openstack-nova11:28
*** yamamoto has joined #openstack-nova11:29
*** vishwanathj has quit IRC11:29
*** vishwanathj has joined #openstack-nova11:29
*** kornicameister has quit IRC11:30
*** kornicameister has joined #openstack-nova11:31
*** nicolasbock has joined #openstack-nova11:32
*** sean-k-mooney has quit IRC11:32
*** aarefiev_afk has quit IRC11:32
*** markvoelker has quit IRC11:33
*** thorst has quit IRC11:33
*** vsaienk0 has quit IRC11:35
*** rmart04 has joined #openstack-nova11:39
*** yuntongjin has quit IRC11:43
*** yuntongjin has joined #openstack-nova11:44
*** purplerbot has quit IRC11:44
*** purplerbot has joined #openstack-nova11:44
alex_xubauzas: so currently our API is free dict for hints field?11:45
*** purplerbot has quit IRC11:46
alex_xubauzas: oops, I saw you said i have valid comment11:46
*** rfolco has joined #openstack-nova11:47
*** purplerbot has joined #openstack-nova11:49
*** purplerbot has quit IRC11:49
*** rfolco_ has joined #openstack-nova11:50
*** rmart04 has quit IRC11:51
*** rmart04 has joined #openstack-nova11:52
*** rfolco has quit IRC11:53
*** sdague has joined #openstack-nova11:53
*** purplerbot has joined #openstack-nova11:55
*** Matias_ has quit IRC11:56
*** Matias_ has joined #openstack-nova11:57
*** rmart04 has quit IRC11:57
*** kaisers_ has quit IRC11:57
*** rfolco_ has quit IRC11:58
*** rfolco has joined #openstack-nova11:59
*** kaisers_ has joined #openstack-nova11:59
*** kaisers_ has quit IRC11:59
*** kaisers_ has joined #openstack-nova12:00
*** moshele has quit IRC12:01
*** moshele has joined #openstack-nova12:01
*** purplerbot has quit IRC12:02
*** purplerbot has joined #openstack-nova12:02
*** aysyd has joined #openstack-nova12:03
*** jpena is now known as jpena|lunch12:05
*** tuanluong has joined #openstack-nova12:06
*** mkoderer_ has joined #openstack-nova12:06
*** databus23_ has joined #openstack-nova12:06
*** seife_ has joined #openstack-nova12:06
*** david_1 has joined #openstack-nova12:06
*** amotoki_ has quit IRC12:07
*** zhurong has quit IRC12:08
*** amotoki has joined #openstack-nova12:08
*** seife_ has quit IRC12:08
*** databus23_ has quit IRC12:08
*** mkoderer_ has quit IRC12:08
*** david_1 has quit IRC12:08
*** crusio has quit IRC12:08
*** crusio has joined #openstack-nova12:09
*** claudiub has joined #openstack-nova12:10
*** aarefiev_afk has joined #openstack-nova12:12
*** lucasagomes is now known as lucas-hungry12:12
*** amotoki has quit IRC12:13
*** cdent has quit IRC12:16
*** rdo has joined #openstack-nova12:20
*** yuntongjin1 has joined #openstack-nova12:20
*** yuntongjin has quit IRC12:21
*** tuanluong has quit IRC12:21
*** yamamoto has quit IRC12:21
*** yamamoto has joined #openstack-nova12:21
*** rdo_ has quit IRC12:22
arne_rmarkus_z: I addressed the issues in https://review.openstack.org/#/c/436868/ and Jenkins verified12:22
*** cdent has joined #openstack-nova12:23
markus_zarne_r: The zkvm CI should also pass12:24
*** aloga__ has joined #openstack-nova12:24
*** aloga___ has joined #openstack-nova12:24
*** 21WAAQZ9J has joined #openstack-nova12:24
*** aloga____ has joined #openstack-nova12:24
*** aloga___1 has joined #openstack-nova12:24
*** aloga___2 has joined #openstack-nova12:24
*** aloga___3 has joined #openstack-nova12:24
*** aloga___4 has joined #openstack-nova12:24
*** aloga___5 has joined #openstack-nova12:24
*** aloga___6 has joined #openstack-nova12:24
*** aloga___7 has joined #openstack-nova12:24
*** yuntongjin1 has quit IRC12:25
arne_rmriedem_away lyearwood: Is this good enough for a workflow now? https://review.openstack.org/#/c/436868/12:25
*** amotoki has joined #openstack-nova12:26
*** sudipto has joined #openstack-nova12:29
*** bkopilov has quit IRC12:30
*** smatzek has joined #openstack-nova12:31
*** smatzek_ has joined #openstack-nova12:32
*** Lambert has quit IRC12:33
*** jvarlamova__ has quit IRC12:33
*** smatzek has quit IRC12:36
*** hferenc has joined #openstack-nova12:36
*** sdague has quit IRC12:36
*** thorst has joined #openstack-nova12:37
*** aloga___1 has quit IRC12:38
*** aloga___6 has quit IRC12:38
*** aloga_ has quit IRC12:38
*** aloga___5 has quit IRC12:38
*** 21WAAQZ9J has quit IRC12:38
*** aloga__ has quit IRC12:38
*** aloga___7 has quit IRC12:38
*** aloga___ has quit IRC12:38
*** aloga___4 has quit IRC12:38
*** aloga___2 has quit IRC12:38
*** aloga___3 has quit IRC12:38
*** aloga____ has quit IRC12:38
*** salv-orlando has quit IRC12:38
*** moshele has quit IRC12:38
*** moshele has joined #openstack-nova12:38
*** phuongnh has quit IRC12:40
*** eglynn has joined #openstack-nova12:42
*** phuongnh has joined #openstack-nova12:43
*** gongysh has joined #openstack-nova12:43
*** sree has joined #openstack-nova12:44
*** sridharg has quit IRC12:44
*** sree_ has joined #openstack-nova12:47
*** sree_ is now known as Guest4153112:48
*** lucas-hungry is now known as lucasagomes12:48
*** kaisers_ has quit IRC12:48
*** sree has quit IRC12:49
alex_xubauzas: johnthetubaguy, looks like we have no quota for scheduler_hints....so is that something user can input any number of hints?12:50
*** ayogi has quit IRC12:50
alex_xubut I guess that limited by the max length of request body12:53
*** phuongnh has quit IRC12:54
*** priteau has joined #openstack-nova12:55
*** mriedem_away is now known as mriedem12:55
*** dave-mccowan has joined #openstack-nova12:56
mriedemasettle: we had docs CPL volunteers in the mailing list12:56
asettlemriedem: yes I was waiting to hear back from you regarding confirmation of that.12:56
asettleI saw there were two volunteers - let me know who is available to provide the time12:57
mriedemasettle: Kevin_Zheng and nicolas bock12:57
asettlemriedem: terrific: https://wiki.openstack.org/wiki/CrossProjectLiaisons#Documentation12:57
asettleIt would be very much appreciated if you could update that table :)12:57
openstackgerritAlex Xu proposed openstack/nova master: placement: Add Traits API to placement service  https://review.openstack.org/37620012:57
openstackgerritAlex Xu proposed openstack/nova master: placement: Move the common code out of ResourceClassCache  https://review.openstack.org/41600712:57
openstackgerritAlex Xu proposed openstack/nova master: placement: Add Trait and TraitList objects  https://review.openstack.org/37619912:57
openstackgerritAlex Xu proposed openstack/nova master: placement: Add simple cache object for traits  https://review.openstack.org/44182912:57
*** trinaths1 has left #openstack-nova12:58
mriedemwill do12:58
mriedemnicolasbock: fyi ^12:59
asettleMuchos appreciated :) and now, back to the manual install that'll kill my will to live12:59
nicolasbockmriedem, thanks12:59
*** tblakes has joined #openstack-nova12:59
mriedemasettle: updated13:00
asettlenicolasbock: while you're here ;) if you haven't seen my email to the dev ML, I've had to update a nova-related bug to critical. https://bugs.launchpad.net/openstack-manuals/+bug/166348513:01
openstackLaunchpad bug 1663485 in openstack-manuals "Document Nova v2 Cells and Placement API" [Critical,In progress] - Assigned to John Garbutt (johngarbutt)13:01
asettleWe've gotta hit the ground running13:01
mriedemjohnthetubaguy: bauzas: cdent: sfinucan: i'm not going to be able to run the full meeting this morning, got single parenting before school duties - can one of you pick up running the meeting?13:01
asettleI need as many people as possible to help test: https://review.openstack.org/#/c/438328/13:01
cdentasettle: i've started a run through of the current doc13:01
cdentmriedem: yeah, I'm sure one of us can get it13:01
* asettle sends cdent flowers 13:01
nicolasbockasettle, I'll have a look13:02
johnthetubaguymriedem: +1 for cdent to take the helm13:02
mriedemheh, i see what you did there13:02
asettleUgh thank you! I'm on it as well. Starting completely fresh.13:02
asettleReally appreciate it cdent and nicolasbock :)13:02
*** jpena|lunch is now known as jpena13:02
* cdent shakes tiny fist at johnthetubaguy 13:02
mriedemthe nova meeting agenda is up to date as of last night, so should be good to go there13:03
cdentbut yeah, sure13:03
asettleIn theory, the instructions should be right now. But I need to make sure :)13:03
* johnthetubaguy smiles at cdent13:03
*** jaosorior is now known as jaosorior_brb13:03
johnthetubaguycdent: I am going to try join a TC vision thing thats happening in boston today, somehow13:03
openstackgerritLee Yarwood proposed openstack/nova master: compute: Move detach logic from manager into driver BDM  https://review.openstack.org/43952013:05
openstackgerritLee Yarwood proposed openstack/nova master: objects: Add attachment_id to BlockDeviceMapping  https://review.openstack.org/43766513:05
openstackgerritLee Yarwood proposed openstack/nova master: compute: Only destroy BDMs after successful detach call  https://review.openstack.org/44069313:05
openstackgerritLee Yarwood proposed openstack/nova master: db: Add attachment_id to block_device_mapping  https://review.openstack.org/43759713:05
cdentjohnthetubaguy: good luck with that, remote vision is hard13:05
johnthetubaguycdent: yeah13:06
*** esberglu has joined #openstack-nova13:06
*** youzp has quit IRC13:06
*** yamamoto has quit IRC13:07
bauzasjohnthetubaguy: asettle: saw the ML replies, looks like the problem is related to libvirt so13:08
*** sean-k-mooney has joined #openstack-nova13:08
johnthetubaguyasettle: stupid question, but do we know how the packaging folks tested the packages?13:08
*** markus_z has quit IRC13:09
bauzasmriedem: okay, I can run it13:09
johnthetubaguyasettle: I think I always assumed they would update the install docs based on the release notes and upstream notes, maybe that was never a thing13:09
mriedembauzas: i think cdent is on the hook13:11
cdentlike a fish13:11
asettlejohnthetubaguy: 'how' - as in, physically how they test it, or how do we know if they did it?13:11
asettlebauzas: yes, I hope that's all that is. But we still need to test to ensure that.13:11
johnthetubaguyasettle: physically really13:11
bauzasmriedem: cdent: okay, no worries, good with me13:12
*** edmondsw has joined #openstack-nova13:12
*** yamamoto has joined #openstack-nova13:13
asettlejohnthetubaguy: that I can't answer. We have people who test it for us. We track the work here: https://wiki.openstack.org/wiki/Documentation/OcataDocTesting13:14
asettleAs you can see, we're a bit blocked.13:14
johnthetubaguyasettle: no worries, just missing bits of the loop here13:14
openstackgerritsahid proposed openstack/nova master: libvirt: return a CPU overhead if isolate emulator threads requested  https://review.openstack.org/38536413:15
openstackgerritsahid proposed openstack/nova master: numa: fit instance NUMA node with cpus reserved onto host NUMA node  https://review.openstack.org/40087513:15
openstackgerritsahid proposed openstack/nova master: virt: update overhead to take into account vCPUs  https://review.openstack.org/38502413:15
asettleWell, we get access to the packages, we then test. Let me poke around and get some holes fixed up for you. I admit entirely I haven't been a part of the install guide process for quite some time.13:16
openstackgerritsahid proposed openstack/nova master: libvirt: make emulator threads to run on the reserved pCPU  https://review.openstack.org/40087813:16
openstackgerritsahid proposed openstack/nova master: numa: take into account cpus reserved  https://review.openstack.org/40087613:16
openstackgerritsahid proposed openstack/nova master: numa: update numa usage incording the cpu reserved  https://review.openstack.org/40087713:16
*** yamamoto has quit IRC13:16
johnthetubaguyasettle: no worries, same here13:17
*** yamamoto has joined #openstack-nova13:17
*** yamamoto has quit IRC13:21
openstackgerritSivasathurappan Radhakrishnan proposed openstack/nova master: Port binding based on events during live migration  https://review.openstack.org/43487013:21
openstackgerritDao Cong Tien proposed openstack/nova master: Stop using mox in unit/virt/disk/test_api.py  https://review.openstack.org/34206113:24
openstackgerritAlex Xu proposed openstack/nova master: placement: Add Traits API to placement service  https://review.openstack.org/37620013:24
openstackgerritMatthew Booth proposed openstack/nova master: WIP: Replace target_cell and set_target_cell with context.for_cell()  https://review.openstack.org/44362713:24
cdentasettle: as a mildly irrelevant aside: a) I'm able to follow these docs in general quite well so: awesome!, b) the nova endpoints are being created with tenant_id in them, has the "naw, that's icky" been communicated (to you and others) in any fashion?13:24
*** prateek has quit IRC13:24
*** yamamoto has joined #openstack-nova13:25
*** markus_z has joined #openstack-nova13:25
mdboothdansmith: If you're alive yet, I starting writing this: https://review.openstack.org/#/c/443627/13:25
asettlecdent: a) SWISH! Terrific. How are you testing? Greenfields? b) the nova docs have not been updated in a *long* time I'm assuming. Hence, need for liaison blah blah you get it. Tl;DR - no.13:25
cdent:)13:26
mdboothIt's in response to my review comment on https://review.openstack.org/#/c/442685/13:26
cdentasettle: I'm using some bare vms, ubuntu13:26
mdboothdansmith: If it's architecturally sound it removes a ton of edge cases and boilerplate13:26
mdboothmelwitt: ^^^ https://review.openstack.org/#/c/443627/13:27
asettlecdent: okay ubuntu is good. That was one of the ones with an issue. I'm trying to do CentOS (key word: trying - screw netowrking rn)13:27
*** salv-orlando has joined #openstack-nova13:28
mdboothdansmith melwitt: It's a bunch of donkey work, so I'm going to leave it like that until we validate/invalidate the approach13:29
*** kaisers_ has joined #openstack-nova13:29
*** diga has quit IRC13:30
*** cleong has joined #openstack-nova13:31
openstackgerritEric Fried proposed openstack/nova master: PowerVM Driver: spawn/delete #1: no-ops  https://review.openstack.org/43811913:31
openstackgerritEric Fried proposed openstack/nova master: PowerVM Driver: spawn/destroy #2: functional  https://review.openstack.org/43859813:33
owalshasettle: is there another doc for upgrades, or do we just care about greenfield installs?13:33
openstackgerritEric Fried proposed openstack/nova master: PowerVM Driver: power_on/off and reboot  https://review.openstack.org/42738013:34
*** kaisers_ has quit IRC13:34
openstackgerritEric Fried proposed openstack/nova master: PowerVM Driver: spawn/destroy #3: TaskFlow  https://review.openstack.org/43872913:35
openstackgerritEric Fried proposed openstack/nova master: PowerVM Driver: spawn/destroy #4: full flavor  https://review.openstack.org/39128813:36
asettleowalsh: we only care about greenfields with regards to this doc13:36
asettle(wow, freud, originally wrote 'shit doc' o.o)13:36
openstackgerritEric Fried proposed openstack/nova master: PowerVM Driver: console  https://review.openstack.org/40940213:36
openstackgerritEric Fried proposed openstack/nova master: WIP: PowerVM Driver: SSP emphemeral disk support  https://review.openstack.org/44318913:38
sean-k-mooneythis is proably a dumb question but if i resize an image it wont loose data will it? assuming it succedded etc13:39
sean-k-mooneysorry not image instance13:39
sean-k-mooneyjust want to double check before i do it to the dev env i use for dpdk13:40
*** ducttape_ has joined #openstack-nova13:41
efriedsean-k-mooney IIRC, resize will detach disks, destroy the instance, create a new instance with the same name and UUID and the new flavor specs, then reattach the same disks.13:41
*** karimb has joined #openstack-nova13:41
efriedsean-k-mooney So no, you shouldn't lose data that's on the disks.13:41
sean-k-mooneyefried: ok that was what i was hoping you would say. its what i assumed it would do but wanted to double check13:42
efriedsean-k-mooney But I could be wrong about all of that ;-)13:42
*** Guest41531 has quit IRC13:42
sean-k-mooneywell worst case i recerate my dev enviornment. i dont currently have anything on it that is not pushed to gerrit bar my local.conf13:42
efriedI assume resize on a live instance will do a graceful shutdown.13:42
*** vladikr has joined #openstack-nova13:43
openstackgerritMatt Riedemann proposed openstack/nova master: WIP: Temporarily untarget context when deleting from cell0  https://review.openstack.org/44340313:43
mriedemmelwitt: Kevin_Zheng: ^ updated that patch13:44
sean-k-mooneywell the cloud its on is at max capasity and i have the instance shut down and some spawned another instance and now i cant start it so trying to see if i can startit with less ram  instead of deleting other insances which is my next option13:44
mriedemi'll work on a functional test for the recreate today13:44
efriedthorst adreznec esberglu Please re-+1 https://review.openstack.org/438598 - had merge conflict all the way up the chain, but this was the only one I had to rebase by hand.  (gerrit carries +1s through auto rebases, it seems.)13:45
*** liverpooler has quit IRC13:45
efriedsean-k-mooney If only you were running PowerVM ;-)13:45
thorstefried: will do13:45
fricklercdent: regarding "tenant_id in endpoint", is there a document somewhere how it should be upgraded/changed properly?13:45
*** liverpooler has joined #openstack-nova13:45
efriedsean-k-mooney Isn't there a memory overcommit in the conf?13:46
*** kfarr has joined #openstack-nova13:46
cdentfrickler: yes, probably, but I'm not sure where at the moment, and buried in reviewing these docs. sdague will know (when returns from tc things)13:46
sean-k-mooneyefried: im not sure intel it would be trilled with me runing power lol :)13:47
efriedsean-k-mooney So you would at worst have to shut down other instances if you're truly out of memory on the host.13:47
mriedemmdbooth: given your comment on the os-migrations patch and the target cell stuff, and corruption, see https://review.openstack.org/44340313:47
sean-k-mooneylibvirt was failing to alloc the memory so ya it was shutdown because i really was out of memory on the host13:47
efriedsean-k-mooney I can never keep straight which is which, but one of shelve, pause, or suspend will save an instance in such a way that it's recoverable, but its resources are freed up.13:49
bauzassfinucan: heh, fun you took it :p13:50
bauzassfinucan: https://review.openstack.org/#/c/235930/13:50
openstackgerritMatt Riedemann proposed openstack/nova master: WIP: Temporarily untarget context when deleting from cell0  https://review.openstack.org/44340313:50
sfinucanbauzas: After 18 months I figured you wouldn't mind :P13:50
sfinucanalso, I was trying to get DevStack in Vagrant going (again)13:51
*** sridharg has joined #openstack-nova13:51
bauzassfinucan: heh, you hit my 18-month acceptance deadline13:51
sean-k-mooneysfinucan: qq. you were doing some struff with adding tags to alias correct? e.g. for pci passthrough?13:51
sfinucansean-k-mooney: Not me personally, but someone was iirc13:52
*** READ10 has joined #openstack-nova13:52
sean-k-mooneysfinucan: do you know if that ever merged?13:53
sfinucansean-k-mooney: I don't, I'm afraid. moshele might be a better person to ask13:53
sean-k-mooneyim asking for the nic feature spec, jay was suggesting moveing the feautre request for the alias to the flavor extra spec13:54
sean-k-mooneywe could move it to the flavor or image mentadata but i was wondering if the device tags would be better but that assumes that work was completed13:55
sean-k-mooneysfinucan: but cool thanks ill take a look and see13:55
*** iceyao has joined #openstack-nova13:56
cdentcoreycb (or anyone else): where's the right place to report an apparent bug in the the packaging (specifically the apache config) for placement api?13:58
mriedemcdent: probably ubuntu/nova13:58
*** takashin has joined #openstack-nova13:58
cdentthanks mriedem13:58
mriedemhttps://bugs.launchpad.net/ubuntu/+source/nova/+bugs ?13:59
coreycbcdent, mriedem, yes that's correct13:59
mriedemi'm not entirely sure, jamespage can probably help too13:59
mriedem\o/13:59
cdentgold star for mriedem13:59
coreycb:)13:59
coreycbcdent, mind forwarding that along once it's open?13:59
cdentyeah, will do,14:00
coreycbcdent, thanks14:00
*** iceyao has quit IRC14:00
*** eharney has joined #openstack-nova14:03
*** yingjun has joined #openstack-nova14:05
*** edleafe_ has joined #openstack-nova14:05
*** udesale has joined #openstack-nova14:06
*** gongysh has quit IRC14:08
*** amotoki has quit IRC14:08
*** gouthamr has joined #openstack-nova14:08
*** efoley has joined #openstack-nova14:12
*** NikhilS has joined #openstack-nova14:13
jaypipessean-k-mooney, sfinucan: fancy a hangout to discuss this?14:14
sfinucanjaypipes: The NIC feature spec, you mean?14:15
*** zz_dimtruck is now known as dimtruck14:15
*** yingjun has quit IRC14:16
*** nshetty has joined #openstack-nova14:16
*** NikhilS has quit IRC14:16
jaypipessfinucan: ya14:18
sfinucanjaypipes: Sure, once the meeting is over14:18
jaypipessfinucan: because, frankly, it's more than just the NIC features.14:18
*** gszasz has quit IRC14:19
*** romainh has joined #openstack-nova14:19
*** ducttape_ has quit IRC14:19
*** yamamoto has quit IRC14:23
*** Kevin_Zheng_Hex has joined #openstack-nova14:25
*** jamesdenton has joined #openstack-nova14:27
*** mriedem has quit IRC14:28
*** bkopilov has joined #openstack-nova14:28
*** nshetty has quit IRC14:29
openstackgerritLee Yarwood proposed openstack/nova master: compute: Move detach logic from manager into driver BDM  https://review.openstack.org/43952014:31
openstackgerritLee Yarwood proposed openstack/nova master: objects: Add attachment_id to BlockDeviceMapping  https://review.openstack.org/43766514:31
openstackgerritLee Yarwood proposed openstack/nova master: compute: Only destroy BDMs after successful detach call  https://review.openstack.org/44069314:31
*** mlavalle has joined #openstack-nova14:32
*** satyar has joined #openstack-nova14:34
sean-k-mooneyjaypipes: sfinucan rodolfo will be tied up for an hour he is presenting in a meeting. but after that sure14:34
jaypipessean-k-mooney: coolio.14:34
sfinucansean-k-mooney: (y)14:35
*** yuntongjin has joined #openstack-nova14:35
*** links has quit IRC14:35
*** smatzek_ has quit IRC14:36
*** markus_z has quit IRC14:37
*** baoli has joined #openstack-nova14:39
*** Jeffrey4l_ has quit IRC14:39
*** takashin has left #openstack-nova14:41
*** iceyao has joined #openstack-nova14:41
*** jaosorior_brb is now known as jaosorior14:43
*** dimtruck is now known as zz_dimtruck14:43
mdboothcdent: Pinging you only because you raised it in the meeting, but fwiw I'm -1 on the patch which implements https://blueprints.launchpad.net/nova/+spec/lvm-thin-pool for 'spec type reasons'14:44
mdboothThere's no spec for me to -1, though14:45
gibijaypipes: hi! Here is a fix to for the orphaning of the instance.flavor we talked about at the PTG. https://review.openstack.org/#/c/440347/ Could you check it please if you have time?14:45
cdentmdbooth: yeah, I'm not up to date on any of that, was just parroting the agenda14:45
*** adrian_otto has joined #openstack-nova14:45
mdboothcdent: Figured14:45
*** kaisers_ has joined #openstack-nova14:45
mdboothmri14:46
mdboothBleargh14:46
mdboothHe's not here14:46
cdenthe has to parent14:46
asettlejohnthetubaguy: in relate to your question earlier about testing packages, there's a bit of info here: https://docs.openstack.org/contributor-guide/release/taskdetail.html14:46
asettleUnder "installation tutorial testing"14:46
mdboothI hate it when that happens14:46
asettleTL;DR - locate pre-release packages for each distribution, and disseminate information about obtaining the packages for testing purposes.14:46
* mdbooth would appreciate a second opinion on whether I'm being an arsehole objecting to https://review.openstack.org/#/c/442126/14:47
*** kevinz has quit IRC14:49
mdboothTL;DR thin pools are very similar to sparse volumes. I don't want to support both. Happy to support thin pools, but only if we deprecate sparse volumes, we have a plan to convert sparse volumes to thin pools automatically (so we don't need to support both), and it's painless for the operator.14:49
lyarwoodmdbooth: I'll go over it again, did you see my comment a few days ago about the current lvm.conf in F25 defaulting to using thinp with the current lvcreate commands for sparse volumes in Nova14:49
mdboothlyarwood: I did, yes.14:49
*** crushil has joined #openstack-nova14:49
mdboothI wonder if there are any implications of mixing them in the same VG.14:50
mdboothi.e. I created these sparse LVs before that change, and these other LVs after that change, so I have a mix in the same VG.14:50
lyarwoodmdbooth: I assume that's fine as they wouldn't change the default otherwise right?14:51
mdboothlyarwood: Anyway, I don't object to thin pools, only supporting 2 very similar ways to do the same thing.14:51
*** Jeffrey4l_ has joined #openstack-nova14:51
mdboothlyarwood: Hard data would be nice.14:52
lyarwoodyup agreed14:52
*** hongbin has joined #openstack-nova14:52
*** markus_z has joined #openstack-nova14:53
cdentcoreycb: that bug https://bugs.launchpad.net/ubuntu/+source/nova/+bug/167150914:53
openstackLaunchpad bug 1671509 in nova (Ubuntu) "placement api configuration in apache configure CustomLog incorrectly" [Undecided,New]14:53
jaypipesedleafe, cdent: your reviews on https://review.openstack.org/#/c/441544/ would be great. thx in advance!14:54
*** nkrinner is now known as nkrinner_afk14:54
cdentjaypipes: noted!14:55
coreycbcdent, thanks we'll get that fixed up asap14:55
edleafejaypipes: queued14:55
jaypipesdanke14:55
*** mriedem has joined #openstack-nova14:56
*** prateek has joined #openstack-nova14:56
*** prateek has quit IRC14:57
*** smatzek_ has joined #openstack-nova14:57
jaypipesmdbooth: you previously reviewed https://review.openstack.org/#/c/399679/. would be great to have a once over from you on that.14:57
*** smatzek_ has quit IRC14:57
mdboothjaypipes: Sure.14:57
jaypipescheers14:58
*** smatzek_ has joined #openstack-nova14:58
*** felipemonteiro_ has joined #openstack-nova14:58
asettlejohnthetubaguy: further explanation, package managers test often with automatic setup and that might include updating config files14:58
mdboothjaypipes: Incidentally, I was looking for validation of this approach for handling RequestContexts for cells: https://review.openstack.org/#/c/443627/14:58
*** eharney has quit IRC14:59
mdboothOr invalidation, obviously, which is why I haven't tried to do the whole job yet.14:59
jaypipesmdbooth: k, queued.14:59
*** yingjun has joined #openstack-nova14:59
mdboothjaypipes: Thanks14:59
*** ujjain has joined #openstack-nova15:00
*** ujjain has quit IRC15:00
*** ujjain has joined #openstack-nova15:00
openstackgerritGábor Antal proposed openstack/nova master: Add json style checking for sample notifications  https://review.openstack.org/44367715:01
dansmithmdbooth: I skimmed it briefly15:03
dansmithmdbooth: I think we can go with that approach for things for object contexts, where we iterate over multiple cells15:03
dansmithmdbooth: I don't like it for clearly single-cell-targeted operations on an instance or a host15:03
dansmithmdbooth: because after getting all of this to work, I feel confident we'll have some leaks where we don't realize that we've now got two copies of the context15:04
mdboothdansmith: Ok. So it's worth adding context.for_cell() and fixing up say get_all() and its users?15:04
dansmithmdbooth: the context _is_ gobal and by reference everywhere15:04
dansmithmdbooth: we get it by reference from the request15:04
*** armax has joined #openstack-nova15:04
dansmithmdbooth: maybe, I also would want to see that change on top of a fully working tempest run of the current stack to prove it doesn't break anything else15:05
*** ducttape_ has joined #openstack-nova15:05
jaypipessfinucan: so https://review.openstack.org/#/c/399653/is going to require the gate to have libvirt 1.2.915:05
mdboothdansmith: That was another reason I didn't want to go all the way down this rabbit hole. I expect there are a bunch of tricky cases.15:06
dansmithmdbooth: there are a ton15:06
mdboothdansmith: However, I was hoping that by removing the whole side-effect of changing the global context there might be significantly less15:06
dansmithmdbooth: I think you might be surprised15:06
*** thorst is now known as thorst_afk15:06
sfinucanjaypipes: I thought the point of those tests was to fake out libvirt?15:07
mdboothdansmith: Ok, I'll limit to get_all() initially and see how it looks.15:07
mdboothThanks for looking.15:07
jaypipessfinucan: see the functional test failure :)15:07
cdentcoreycb: I think I've got another one. Is it considered a bug for the service.conf files to have uncommented config settings?15:07
*** ekuris has quit IRC15:07
coreycbcdent, not necessarily.  some of the packages set defaults.15:08
sfinucanjaypipes: Well that doesn't look right :D15:08
cdentcoreycb: in this case it is that os_region_name in the [placement] section is being set15:08
sfinucanGuess we can mock out that version check too15:08
dansmithmdbooth: fwiw, we already had a discussion about whether or not to fork the context in the objects that are tied to cells and decided on this approach to start15:08
dansmithmdbooth: one thing to remember is that if you send an object over rpc, it resets the context on all the objects in the stack15:09
dansmithmdbooth: so if you make a call over to conductor, for anything like boot or resize, contexts get reset anyway15:09
coreycbcdent, that may not need to be set.  we set defaults so that the service will can start after the package is installed.15:09
openstackgerritDan Smith proposed openstack/nova master: WIP: Temporarily untarget context when deleting from cell0  https://review.openstack.org/44340315:10
cdentcoreycb: the reason I noticed was because the docs I was testing changed it, i made all those changs at the top of the section, and then realized it was going to conflict with something that's set further below15:10
cdentcoreycb: ah, okay, yeah, that makes sense then15:10
cdentit won't start without that15:10
cdent(there's no default)15:10
mriedemdansmith: ftr, i did give you the business, yes15:11
mriedem:)15:11
dansmithmriedem: yeah, I just totally wasn't paying attention when I wrote that :)15:11
dansmithmriedem: see my rev15:11
dansmith?15:12
*** yuntongjin has quit IRC15:12
mriedemyes15:13
mriedemwait, what?15:13
mriedemoh i see15:13
*** hshiina has quit IRC15:13
dansmithactually, I have to fix target_cell too15:13
mriedemdansmith: you missed on15:13
mriedem*one15:13
dansmithoh no15:14
dansmithit works15:14
*** eharney has joined #openstack-nova15:14
mriedemyou missed one in the exception block15:14
dansmithah yeah15:14
dansmithjust a sec15:14
*** zz_dimtruck is now known as dimtruck15:14
openstackgerritDan Smith proposed openstack/nova master: WIP: Temporarily untarget context when deleting from cell0  https://review.openstack.org/44340315:14
openstackgerritBalazs Gibizer proposed openstack/nova master: Using max api version in notificationt sample test  https://review.openstack.org/44368615:14
*** salv-orlando has quit IRC15:14
*** salv-orlando has joined #openstack-nova15:15
*** awaugama has joined #openstack-nova15:15
coreycbcdent, ok.  we're considering loosening the restriction of having to start services by default because it can cause issues particularly with deployment tools.15:15
dansmithmriedem: so we should be able to write a functional test for this that checks the quota in each after delete right?15:15
coreycbcdent, but we haven't done that yet.  if we did i think we'd just use the default config files.15:15
cdentcoreycb: it's a reasonable thing to do for machines being directly used by humans but not so much in this case :)15:16
mriedemdansmith: yes15:16
*** unicell has joined #openstack-nova15:16
mriedemthat's the plan15:16
coreycbcdent, yeah :)15:17
mriedemthe tricky thing is going to be,15:17
mriedemwe only need one of these fixes for the backport to ocata15:17
mriedemso the functional test won't pass until the top patch in pike15:17
mriedembut it will work for the bottom patch in ocata15:17
mriedembut re-ordering the changes in master doesn't make sense to me15:17
dansmithokay15:17
*** unicell1 has quit IRC15:17
mriedemso i'm thinking maybe the functional test is in a separate patch15:18
mriedemnot sure15:18
dansmithyeah15:18
*** tbachman has joined #openstack-nova15:19
*** iceyao has quit IRC15:19
*** iceyao has joined #openstack-nova15:19
*** yamahata has joined #openstack-nova15:21
mriedemso i think the easy thing i can do first is create a functional regression patch that goes at the bottom of the series and recreates the failure, and asserts the error,15:22
mriedemthen the fix patches on top of that,15:22
mriedemand the top fix patch changes the functional test to show it working15:22
mriedemthat was my plan anyway15:22
dansmithI was just looking at the tests15:22
dansmithgive me a few minutes15:22
*** sridharg has quit IRC15:22
*** awaugama has quit IRC15:22
dansmithwe don't really have a way to query into the reservation stuff to look, AFAICT, but I think I can hack something in at least to check15:23
mriedemwell,15:23
mriedemfor a functional test i think this is pretty straightforward,15:23
mriedemwe have cell1 and cell0,15:23
mdboothdansmith: Incidentally, https://review.openstack.org/#/c/443403/ is pretty much what I was talking about with the side-effect thing.15:23
mriedemwe disable nova-compute or do something to make scheduling fail,15:23
*** yamamoto has joined #openstack-nova15:23
mriedemcreate the server and show it goes into ERROR state,15:23
dansmithmdbooth: it's not at all, since this is a problem across multiple requests15:23
mriedemassert we have quota of instances=9 (default-1)15:23
mriedemthen delete the instance,15:23
mriedemand assert that quota of instances=10 (but it won't be until the fix)15:24
mriedemit's basically the same recreate steps as in the bug report15:24
dansmithmriedem: I guess I'm lost in the db api on terminology, knowing which things are returning limits and which are calculated by limit and usage15:24
*** iceyao has quit IRC15:24
dansmithmriedem: but yeah, that's what I'm doing15:24
mriedemthe functional test just hits the rest api15:24
mriedemlike novaclient would15:24
*** Kevin_Zheng_Hex has quit IRC15:25
mriedemi really don't want to be going into the actual databases15:25
mriedemi was going to do this in nova/tests/functional/regressions15:25
dansmithmriedem: well, you're not really validating that exactly what you think is happening really is, but okay15:25
dansmithI'll leave it to you then15:26
mriedemas long as the regression test asserts the failure (the recreate), and the fix has to change the test to show it now passing, i think that's what we care about15:26
dansmithI just want to see them in the wrong database, but okay15:27
dansmiththe quota stuff is so opaque to me that I'd want to see that we're out of sync for the reason we think and not one of the other reasons quotas gets out of sync15:27
dansmithbut that's fine, you're far more familiar, so go for it15:27
*** iceyao has joined #openstack-nova15:27
mriedemwhy did you change set_target_cell?15:28
mriedemwe're calling target_cell and passing None15:28
dansmithtarget_cell calls set_target_cell15:28
mriedemgotcha15:29
*** iceyao has quit IRC15:29
*** yamamoto has quit IRC15:29
*** iceyao has joined #openstack-nova15:30
*** karimb has quit IRC15:31
johnthetubaguymdbooth: I am looking at your comments on migrate_volume_completion15:32
mdboothjohnthetubaguy: I spent half a day or so reading the cinder side code for that, btw15:32
johnthetubaguymdbooth: I don't feel like I am understanding what that is trying to do, and if it should die/get replaced in this new model15:32
mdboothI'll never get that back15:32
johnthetubaguyyeah, I did that about 6-8 months ago, and its wondered out my head again15:32
*** adrian_otto has quit IRC15:32
mdboothjohnthetubaguy: I think it needs to continue to exist, because cinder needs to know that the copy is complete so it can cleanup the old volume.15:33
johnthetubaguymdbooth: yeah, we totally need that bit15:34
mdboothHowever, I suspect it could similarly exist in a new form15:34
johnthetubaguyyeah, that15:34
mdboothjohnthetubaguy: It would be great if the new form included a flag for 'we didn't do the copy'15:34
*** iceyao has quit IRC15:34
mdboothSo that in cases where the driver can't do it, eg because the instance is stopped, cinder knows it needs to do it15:35
mdboothAnd we don't need to fail the operation15:35
*** awaugama has joined #openstack-nova15:35
johnthetubaguymdbooth: my problem is how do we know what to do with the attachments, does it alway delete the correct attachment for us?15:36
mdboothWell cinder will have its own migration record which it hasn't exposed to Nova, which iirc is how it decides which 'mode' it's in15:37
mdboothCan cinder just work it out?15:37
johnthetubaguymdbooth: I think it knows, I just wonder if it always does the same thing from our view of the world15:37
johnthetubaguylike it always deletes the correct attachment15:37
johnthetubaguyso we don't need to delete any attachments15:38
mdboothjohnthetubaguy: We say: we're done with the swap volume of att_A to att_b for instance X, and we did the copy15:38
mdboothThat's enough info for cinder in all cases, right?15:38
mdboothAnd I don't think we need a return value any more15:39
johnthetubaguywhat I mean is, what do we have to do?15:39
mdboothBecause we don't need to do connection_info munging15:39
johnthetubaguycinder gives us the volume-uuid of the finally attached volume to update in the BDM15:39
mdboothI think the only thing Nova has to do is callback cinder with ^^^15:39
johnthetubaguymaybe we just don't delete any attachments15:39
*** karimb has joined #openstack-nova15:39
mdboothIt actually occurred to me that all we really needed to do was delete the attachment15:40
mdbooththe source attachment, that is15:40
mdboothAnd that's actually enough info for cinder15:40
mdboothHowever, I vastly prefer to be explicit15:40
mdboothI think we delete the attachment, then make the callback15:40
gibidansmith, melwitt: Hi! could you please look at the bugfix of the instance.delete regression? https://review.openstack.org/#/c/437222/15:40
johnthetubaguyOh, true, cinder could delete the volume if it knows it has to15:40
johnthetubaguybut something nice about migrate being explicit, in case that gets messed up15:41
mdboothYeah15:41
asettlejaypipes: thanks for reviewing the doc :) were you able to test or just take a look?15:41
johnthetubaguymdbooth: I have no idea how this could work with multi-attached volumes?? who does the delete?15:41
jaypipesasettle: lol, just on a hangout with cdent talking about that right now :)15:41
jaypipesasettle: apparently there *is* an openstack-nova-api-placement package...15:41
asettlejaypipes: wow you guys have all the fun don't you :p15:41
mdboothjohnthetubaguy: Whoa, I hadn't even considered that15:42
mdboothjohnthetubaguy: Who does the copy?15:42
*** mdnadeem has quit IRC15:42
asettlejaypipes: oh. so. what's next?15:42
*** marst has quit IRC15:42
jaypipesasettle: well, other than that one comment from me, I think all my other comments were correct ;)15:42
*** moshele has quit IRC15:42
*** sree has joined #openstack-nova15:42
johnthetubaguymdbooth: everyone? oh dear... yeah.15:43
mdboothjohnthetubaguy: Ok, I think that problem can be solved if we add a 'please do the copy' flag to nova's swap_volume call15:43
mdboothI think cinder then has enough tools in the box to do this15:43
asettlejaypipes: yeah just reading htem all now15:43
mdboothHmm, no, you'd need a quiesce operation too15:44
johnthetubaguymdbooth: well we would also have to call back with the attachment_ids and not the volume ids right?15:44
mdboothjohnthetubaguy: That case requires a lot of thought15:44
mdboothAll attached instances would have to participate15:44
openstackgerritDan Smith proposed openstack/nova master: Change MQ targeting to honor only what is in the context  https://review.openstack.org/43802315:45
openstackgerritDan Smith proposed openstack/nova master: Teach os-migrations about cells  https://review.openstack.org/44268515:45
openstackgerritDan Smith proposed openstack/nova master: Add cell field to Destination object  https://review.openstack.org/43802415:45
openstackgerritDan Smith proposed openstack/nova master: Teach simple_tenant_usage about cells  https://review.openstack.org/44138815:45
openstackgerritDan Smith proposed openstack/nova master: Teach os-aggregates about cells  https://review.openstack.org/44220015:45
mdboothjohnthetubaguy: So attachment_ids are going to make that case simpler, but it's also going to involve coordinating a simultaneous volume switch amongst all attached instances.15:45
johnthetubaguymdbooth: its assume the user ensures only one is writing, but doesn't tell cinder right now, this is messed up, I think we should just fail this for multi-attach, for the moment at least15:45
mdboothThe latter problem is going to be funky15:45
johnthetubaguymdbooth: +1 its hard15:46
*** mriedem1 has joined #openstack-nova15:46
dansmithgibi: was that a regression in ocata?15:46
gibidansmith: yes it was15:46
*** mriedem has quit IRC15:46
gibidansmith: I will backport it15:46
dansmithmriedem1: so we should rebase the quota fix on this one probably15:46
dansmithbecause they will conflict15:46
mdboothjohnthetubaguy: Yeah, I think it would be cinder's job to refuse to retype multi-attach volumes until support for that is added.15:47
*** kaisers_ has quit IRC15:47
*** sree has quit IRC15:47
mdboothNova shouldn't concern itself, just do the swap, and copy if requested15:47
johnthetubaguymdbooth: well, Nova can't do the copy either right, when its Nova initiated15:48
*** tbachman has quit IRC15:49
mdboothI mean, Nova gets an api call with (att_A, att_b, do_copy=True)15:49
mdboothThat call may come direct from a user, or from cinder which has a migration record15:49
*** tbachman has joined #openstack-nova15:49
*** marst has joined #openstack-nova15:49
mdboothEither way, Nova's behaviour is the same:15:49
mdboothAttach to att_b15:49
mdboothCopy if requested15:50
mdboothDetach from att_a15:50
mdboothcall cinder with finished(att_a, att_b, instance_x, did_copy=True)15:50
dansmithgibi: question inthere15:50
mdboothAnd we update the connection_info to be that for att_b, in all cases15:51
*** ducttape_ has quit IRC15:51
dansmithgibi: unfortunately there's more cleanup in that patch than just is necessary for the backport, AFAICT15:51
mdboothjohnthetubaguy: And we trust the caller not to do anything which is going to cause data corruption15:51
dansmithgibi: I'm kinda conflicted on whether to just do that whole thing vs. two patches, one that is a smaller backport, followed by the cleanup15:51
mdboothjohnthetubaguy: I think instance is redundant in that callback, btw, because it's a property of both attachments15:53
*** mriedem1 is now known as mriedem15:53
mriedemdansmith: which one?15:53
mdboothIt shouldn't be there15:53
dansmithmriedem: https://review.openstack.org/#/c/437222/515:53
*** sdague has joined #openstack-nova15:53
mriedemoh yeah i knew about that, just hadn't gone through it yet15:53
gibidansmith: you mean do the minimal thing first, backport that then do some cleanup in a followup on master?15:54
dansmithgibi: yeah15:54
mriedemsmaller backports are much easier and nicer15:54
cdentjaypipes: I had to -1 get_invetory, but it is a very soft -115:55
mriedemdansmith: gibi: what's the cleanup? the context manager?15:55
dansmithmriedem: yeah, and the use of it in a few places that aren't affected, AFAICT15:56
mriedemjaypipes: https://review.openstack.org/#/c/442200/ - that's the bottom of the series and you were +2 yesterday15:57
gibimriedem: I think this is the only place where it wasn't needed to restore original behavior https://review.openstack.org/#/c/437222/1/nova/compute/api.py@186815:57
gibimriedem, dansmith: I can put this ^^ to a separate patch15:58
dansmithgibi: that's PS115:58
gibiups sorry15:58
*** rcernin has quit IRC15:58
dansmithgibi: there are only two places we're missing the notification,15:58
dansmithand two places you convert to the new one, right?15:58
*** KeithMnemonic has joined #openstack-nova15:58
jaypipesmriedem: k, on it. just +2'd https://review.openstack.org/#/c/345138 BTW15:59
*** jaosorior has quit IRC15:59
mriedemok, haven't reviewed that yet at all15:59
dansmithI have I'll circle back15:59
gibidansmith: true. I looked at the wrong ps15:59
*** raj_singh_ has joined #openstack-nova16:00
gibidansmith: I can split the patch16:00
dansmithgibi: I think splitting those up would be good for backport cleanliness if you're cool with it16:00
dansmithgibi: sweet, thanks for finding, fixing :)16:00
gibidansmith: cool, I will do it16:00
*** markus_z has quit IRC16:02
*** dtp has joined #openstack-nova16:04
*** links has joined #openstack-nova16:07
*** jahsis has quit IRC16:12
*** edleafe_ has quit IRC16:13
*** ducttape_ has joined #openstack-nova16:15
*** jaosorior has joined #openstack-nova16:19
*** tbachman has quit IRC16:24
*** yamamoto has joined #openstack-nova16:26
*** nikeshm has joined #openstack-nova16:27
nikeshmhi16:27
nikeshmi am frequently facing the quota issue in openstack master16:28
nikeshmhttps://ask.openstack.org/en/question/494/how-to-reset-incorrect-quota-count/16:28
nikeshmis there any bug reported16:28
nikeshmfor this16:28
*** yamamoto has quit IRC16:30
*** adrian_otto has joined #openstack-nova16:31
*** andreas_s has quit IRC16:32
openstackgerritRodolfo Alonso Hernandez proposed openstack/nova-specs master: Enable SR-IOV NIC offload feature discovery  https://review.openstack.org/43595416:33
openstackgerritChris Friesen proposed openstack/nova master: show flavor info in server details  https://review.openstack.org/43451416:34
jaypipessfinucan, sean-k-mooney: time for a hangout? is rodolfo available? (btw, what's rodolfo's IRC nick?)16:35
sfinucanjaypipes: (y)16:35
sfinucanAs for IRC nick - not a clue16:36
sean-k-mooneyjaypipes: ralonsoh16:37
*** ducttape_ has quit IRC16:37
sean-k-mooneyand ya im free ill ask him16:37
jaypipescool.16:37
ralonsohjaypipes, sean-k-mooney: hi, now I'm free16:38
*** thorst_afk is now known as thorst16:38
jaypipessean-k-mooney, sfinucan, ralonsoh: https://hangouts.google.com/call/qqze37pjs5f4lnatg7owxiaklqe16:39
*** efoley_ has joined #openstack-nova16:41
openstackgerritMatt Riedemann proposed openstack/nova master: Add regression test for bug 1670627  https://review.openstack.org/44374516:41
openstackbug 1670627 in OpenStack Compute (nova) "quota is always in-use after delete the ERROR instances " [Critical,In progress] https://launchpad.net/bugs/1670627 - Assigned to Dan Smith (danms)16:41
mriedemdansmith: melwitt: Kevin_Zheng: ^ this is the regression test i was thinking of16:41
*** yassine has quit IRC16:42
dansmithmriedem: ack, and it works?16:42
*** abalutoiu_ has joined #openstack-nova16:42
mriedemit passes16:42
mriedemi'm going to rebase the fixes on top of it16:43
dansmithit passes on master?16:43
*** sudipto has quit IRC16:44
dansmithoh, I see it validates the broken behavior currently16:44
*** efoley has quit IRC16:44
*** karimb has quit IRC16:45
*** abalutoiu has quit IRC16:45
mriedemright, that's generally how these types of tests work,16:47
mriedemwe write the recreate scenario,16:47
*** sandanar has quit IRC16:47
mriedemthen roll the fix on top and change the assertion in the test when it's fixed16:47
*** psachin has quit IRC16:48
mriedemok, first victim is context.db_connection == None blows up the CellDatabases fixture16:49
*** udesale has quit IRC16:49
dansmithmriedem: yeah you'll have to make the change there since that fixture replaces target_cell16:50
*** kaisers_ has joined #openstack-nova16:50
nikeshmsmcginnis:  https://bugs.launchpad.net/nova/+bug/1670627   :)16:51
openstackLaunchpad bug 1670627 in OpenStack Compute (nova) "quota is always in-use after delete the ERROR instances " [Critical,In progress] - Assigned to Dan Smith (danms)16:51
dansmithmriedem: if it's None, just yield without calling real_target_cell16:51
dansmithmriedem: but inside the lock16:51
mriedemi was just going to use self._default_ctxt_mgr16:52
dansmithnot sure what you mean16:52
*** abalutoiu_ has quit IRC16:53
dansmithmriedem: before this, you need to yield if cell_mapping is None and then return https://github.com/openstack/nova/blob/master/nova/tests/fixtures.py#L402-L40216:53
mriedemsec16:53
mriedemhere it comes16:54
openstackgerritMatt Riedemann proposed openstack/nova master: WIP: Decrement quota when deleting an instance in cell0  https://review.openstack.org/44339516:55
openstackgerritMatt Riedemann proposed openstack/nova master: Add regression test for bug 1670627  https://review.openstack.org/44374516:55
openstackbug 1670627 in OpenStack Compute (nova) "quota is always in-use after delete the ERROR instances " [Critical,In progress] https://launchpad.net/bugs/1670627 - Assigned to Dan Smith (danms)16:55
openstackgerritMatt Riedemann proposed openstack/nova master: WIP: Temporarily untarget context when deleting from cell0  https://review.openstack.org/44340316:55
*** browne has joined #openstack-nova16:55
*** kaisers_ has quit IRC16:55
*** efoley__ has joined #openstack-nova16:56
dansmithmriedem: I'd prefer you didn't do that,16:56
*** jamesden_ has joined #openstack-nova16:56
*** annegentle has joined #openstack-nova16:56
mriedemi'm fine changing it16:56
dansmithmriedem: not only because it ends up mutating the context unlike the real live scenario, but it also is database-specific, so when my mq one rolls around, it will target the mq but not the db16:56
*** lucasxu has joined #openstack-nova16:57
*** jamesdenton has quit IRC16:58
mriedemhmm, so that makes the test fail16:58
*** jamesden_ has quit IRC16:58
*** raunak has joined #openstack-nova16:58
*** efoley_ has quit IRC16:59
*** lpetrut has quit IRC17:01
mriedemdansmith: wouldn't the default context be pointing at the main cell db?17:01
mriedemwhich is what we want when we create the reservation17:02
dansmithmriedem: it points at the one that was default=true, but in real life you'll end up taking the global one instead of having your cell targeted at the other one17:02
dansmithmriedem: which means it depends on their config, and won't look like it does in real life17:02
dansmithmriedem: so someone will come along and want to assert that db_connection=None there because that is how target_cell works, but it will be different when using the fixture17:02
*** gfhellma has joined #openstack-nova17:03
arne_rCalling attention to https://review.openstack.org/#/c/436868/ which, if passed, would make our CI folks really happy17:04
mriedemdansmith: ok, in the nova/cinder meeting now, you want to tinker with that locally to change it?17:04
mriedembecause i'm a bit lost in that fixture17:04
dansmithmriedem: sure17:04
dansmithit's complicated17:05
dansmithmriedem: you said the test failed.. which one did you mean? your regression test should fail yes?17:05
mriedemhttps://review.openstack.org/443403 is passing17:06
mriedemwhen i changed to what you wanted in the fixture, it failed17:06
mriedemso to do what you want in the fixture, i need you to take a look at why it's failing17:06
dansmithmriedem: but which test failed, the regression test?17:06
mriedemyes17:06
mriedemi'm just running that one17:06
dansmithoh I see you changed it there, gotcha17:07
dansmithwaiting for a py27 run to finish and then I'll pull it down17:07
*** lucasxu has quit IRC17:08
*** markvoelker has joined #openstack-nova17:08
mriedempy27 for something else right?17:08
dansmithyes17:09
openstackgerritBalazs Gibizer proposed openstack/nova master: fix missing instance.delete notification  https://review.openstack.org/43722217:11
*** mvk has quit IRC17:12
*** hieulq has quit IRC17:12
mriedemok, once i'm done with this meeting i'll fix up the unit tests on the middle patch17:12
*** Madasi has joined #openstack-nova17:14
*** raj_singh_ has quit IRC17:14
*** tbachman has joined #openstack-nova17:15
*** hieulq has joined #openstack-nova17:15
*** arne_r has quit IRC17:16
*** yingjun has quit IRC17:16
*** lucasxu has joined #openstack-nova17:17
*** lucasxu has quit IRC17:17
*** nic has joined #openstack-nova17:17
dansmithmriedem: actually, I changed my mind, this is okay17:18
dansmithI'm not sure it's as obvious, but it's less fiddling, and the context still looks like it should because we call real_target_cell anyway17:19
dansmithmriedem: for completeness let me show you what I was thinking:17:19
dansmithmriedem: http://pastebin.com/raw/isCJMvhG17:19
dansmithmriedem: that works, but it may be more confusing for other people17:20
dansmithif you agree, then just stick with what you've got17:20
mriedem"controllering"?17:20
dansmithhttp://pastebin.com/raw/pNv5smz817:20
mriedemi disagree with that17:20
dansmithis what it looks like applied17:20
dansmithhah I just saw that17:21
*** annegentle has quit IRC17:21
dansmithunintentional, but sounds like me17:21
*** romainh has quit IRC17:21
*** lpetrut has joined #openstack-nova17:21
mriedemso, your change, it's basically just a pass through to the real unwrapped thing, right?17:22
dansmithskipping the global db state switch, yeah17:22
mriedemok i think it makes sense to me, enough ayway17:22
mriedem*anyway17:22
dansmithyours does the switch, but  .. to the main one17:22
openstackgerritBalazs Gibizer proposed openstack/nova master: Transform missing delete notifications  https://review.openstack.org/41029717:23
openstackgerritBalazs Gibizer proposed openstack/nova master: use context mgr in instance.delete  https://review.openstack.org/44376417:23
mriedemok i'll just take your change and update my local copy17:24
gibimriedem, dansmith: I splitted the instance.delete bugfix patch as we discussed https://review.openstack.org/#/c/43722217:24
*** moshele has joined #openstack-nova17:24
dansmithmriedem: totally your call17:24
gibimriedem, dansmith: but I have to run now. thanks for the reviews!17:25
dansmithgibi: I was going to suggest that first patch not adding/using the contextmanager17:25
openstackgerritStephen Finucane proposed openstack/nova master: console: introduce basic framework for security proxying  https://review.openstack.org/34539617:25
openstackgerritStephen Finucane proposed openstack/nova master: console: introduce framework for RFB authentication  https://review.openstack.org/34539717:25
dansmithgibi: but if it's not going to cause conflicts, it's not as much of a problem I guess17:25
openstackgerritStephen Finucane proposed openstack/nova master: console: introduce the VeNCrypt RFB authentication scheme  https://review.openstack.org/34539817:25
openstackgerritStephen Finucane proposed openstack/nova master: console: provide an RFB security proxy implementation  https://review.openstack.org/34539917:25
*** sapcc-bot2 has joined #openstack-nova17:25
*** lpetrut has quit IRC17:25
*** sapcc-bot has quit IRC17:25
gibidansmith: it didn't caused conflict a couple of days ago when I cherry-picked the patch to stable/ocata17:26
*** yamamoto has joined #openstack-nova17:26
dansmithgibi: not now for sure, I just mean in terms of backporting less code. the context manager add is more code, but it's contained17:26
dansmithgibi: if mriedem is cool with it then it's fine17:26
gibidansmith: here is the cherry-pick https://review.openstack.org/#/c/441171/ I will update that as well17:26
gibidansmith: ahh I see17:27
gibidansmith: I agree that the ctx manager is quite self contained solution17:27
dansmithgibi: anyway, I'll look at this in a few, thanks!17:27
gibidansmith: thanks a lot!17:27
* gibi runs away17:28
openstackgerritStephen Finucane proposed openstack/nova master: conf: Fix formatting of network options  https://review.openstack.org/44308617:28
mriedemcatch him!17:28
*** jaosorior has quit IRC17:29
openstackgerritStephen Finucane proposed openstack/nova master: conf: Final cleanups in conf/network  https://review.openstack.org/43102317:29
*** unicell has quit IRC17:29
openstackgerritJohn Garbutt proposed openstack/nova-specs master: Add spec to use cinder's new attachment API  https://review.openstack.org/37320317:30
*** yamamoto has quit IRC17:32
*** mtanino has joined #openstack-nova17:32
*** nikeshm has quit IRC17:34
openstackgerritMatt Riedemann proposed openstack/nova master: WIP: Temporarily untarget context when deleting from cell0  https://review.openstack.org/44340317:34
*** ducttape_ has joined #openstack-nova17:37
*** ducttape_ has quit IRC17:38
*** ducttape_ has joined #openstack-nova17:38
*** kfarr has quit IRC17:40
*** ssurana has joined #openstack-nova17:41
*** snecklifter has joined #openstack-nova17:41
openstackgerritRodolfo Alonso Hernandez proposed openstack/nova-specs master: Enable SR-IOV NIC offload feature discovery  https://review.openstack.org/43595417:41
*** tesseract has quit IRC17:42
openstackgerritSarafraj Singh proposed openstack/nova-specs master: Spec to add per instance timeout for LM  https://review.openstack.org/44280917:43
openstackgerritFeodor Tersin proposed openstack/nova master: Give names to volumes during block device mapping  https://review.openstack.org/21343317:44
*** lucasagomes is now known as lucas-afk17:45
*** mvk has joined #openstack-nova17:47
*** derekh has quit IRC17:47
openstackgerritRafael Folco proposed openstack/nova master: DELETE all inventory for a resource provider  https://review.openstack.org/41666917:47
*** KeithMnemonic has quit IRC17:47
*** ducttape_ has quit IRC17:49
*** browne has quit IRC17:50
dansmithmriedem: I'm going to pull down gibi's patch and fix a test comment of mine17:50
dansmithmriedem: are you cool with the heavier context manager approach for the backport or do you want to do the simplest thing, which is just call the normal methods in the places where we're not?17:51
dansmithI think his approach here ends up getting us a begin/end notification that we wouldn't get otherwise17:51
dansmithwhich is probably good, it's just .. more code17:52
*** moshele has quit IRC17:54
*** slaweq_ has quit IRC17:54
*** moshele has joined #openstack-nova17:55
*** kaisers_ has joined #openstack-nova17:56
*** lpetrut has joined #openstack-nova17:56
*** lucasxu has joined #openstack-nova17:58
*** cdent has quit IRC17:58
mriedemif it's not making the backport a mess i'm fine with it17:59
*** kaisers_ has quit IRC17:59
mriedemi haven't dug in in detail17:59
*** kaisers_ has joined #openstack-nova17:59
*** moshele has quit IRC18:00
*** ralonsoh has quit IRC18:01
openstackgerritDan Smith proposed openstack/nova master: Fix missing instance.delete notification  https://review.openstack.org/43722218:03
*** kaisers_ has quit IRC18:03
dansmithI also sneakily capitalized the first letter of the commit message because .. I hate that.18:03
*** cdent has joined #openstack-nova18:04
*** links has quit IRC18:04
edleaferfolco: A could of nits on https://review.openstack.org/416669, but one serious issue18:05
*** adrian_otto has quit IRC18:05
*** jpena is now known as jpena|off18:06
rfolcoedleafe, ack18:07
mriedemdansmith: but not period at the end?18:10
mriedem*no18:10
dansmithno I hate theperiod18:10
mriedemI like pizza. I also like cookies.18:10
dansmithheh18:10
*** ltomasbo is now known as ltomasbo|away18:11
* dansmith reminds himself to never say something to mriedem he doesn't want recalled years later18:11
*** efoley__ has quit IRC18:12
*** abalutoiu_ has joined #openstack-nova18:12
*** jamesdenton has joined #openstack-nova18:13
openstackgerritRafael Folco proposed openstack/nova master: DELETE all inventory for a resource provider  https://review.openstack.org/41666918:14
*** unicell has joined #openstack-nova18:15
*** zioproto has quit IRC18:21
*** Apoorva has joined #openstack-nova18:25
*** yamamoto has joined #openstack-nova18:28
*** amoralej is now known as amoralej|off18:31
*** yamamoto has quit IRC18:33
*** browne has joined #openstack-nova18:36
mdbooth... when you realise you created dicts called host_cell_mapping_mapping and cell_mapping_host_mapping18:36
*** armax has quit IRC18:38
cburgessOMG18:40
cburgessYou should be taken out back and flogged for that.18:40
mdboothcburgess: Done ;)18:40
efriedmdbooth Clearly the former is mapping the mappings of hosts to cells, and the latter is mapping the ....  sh!t, I got nothing.18:40
mdboothhosts to cell mappings, and cell mappings to hosts18:41
mdboothIt made sense when I wrote it18:41
cburgessWhere you are a ballmer peak when you wrote that?18:41
*** yamahata has quit IRC18:42
*** karimb has joined #openstack-nova18:42
openstackgerritLee Yarwood proposed openstack/nova master: compute: Move detach logic from manager into driver BDM  https://review.openstack.org/43952018:43
openstackgerritLee Yarwood proposed openstack/nova master: objects: Add attachment_id to BlockDeviceMapping  https://review.openstack.org/43766518:43
openstackgerritLee Yarwood proposed openstack/nova master: compute: Only destroy BDMs after successful detach call  https://review.openstack.org/44069318:43
openstackgerritLee Yarwood proposed openstack/nova master: db: Add attachment_id to block_device_mapping  https://review.openstack.org/43759718:43
openstackgerritLee Yarwood proposed openstack/nova master: Make BlockDeviceMapping object support uuid  https://review.openstack.org/24260318:43
openstackgerritLee Yarwood proposed openstack/nova master: Add uuid column to BlockDeviceMapping  https://review.openstack.org/24260218:43
openstackgerritLee Yarwood proposed openstack/nova master: compute: Use detach_volume during default rebuild  https://review.openstack.org/44393218:43
*** dharinic is now known as dharinic|lunch18:43
*** ducttape_ has joined #openstack-nova18:50
*** fragatina has joined #openstack-nova18:50
*** fragatina has quit IRC18:50
*** fragatina has joined #openstack-nova18:51
*** sambetts is now known as sambetts|afk18:52
*** adrian_otto has joined #openstack-nova18:55
*** ducttape_ has quit IRC18:55
*** slaweq_ has joined #openstack-nova18:56
*** dtp has quit IRC18:57
openstackgerritDan Smith proposed openstack/nova master: Make discover_hosts only query for unmapped ComputeNode records  https://review.openstack.org/42790318:57
openstackgerritDan Smith proposed openstack/nova master: Add ability to query for ComputeNodes by their mapped value  https://review.openstack.org/42790218:57
openstackgerritDan Smith proposed openstack/nova master: Add ComputeNode.mapped field  https://review.openstack.org/42790118:57
oomichimriedem: sdague: is nova-network deprecated already? I guess so, but I cannot find clear doc on nova repo19:01
*** giri has joined #openstack-nova19:02
mriedemoomichi: since newton19:02
mriedemoomichi: the config options, release notes, starting the service, etc etc19:03
*** kfarr has joined #openstack-nova19:03
mriedemthe official manuals probably don't say anything about it19:03
oomichimriedem: thanks, I got it. reno is good for me19:03
mriedemasettle: so https://docs.openstack.org/admin-guide/compute-networking-nova.html19:05
openstackgerritJay Pipes proposed openstack/nova master: virt: implement get_inventory() for Ironic  https://review.openstack.org/44154419:05
jaypipescdent: fixed up per your review ^^ dansmith, you were already +2 on it :)19:05
mriedemasettle: nova-network has been deprecated since newton, and in ocata we made it such that you can only run nova-network in a cells v1 configuration (which is not recommended for new deployments either)19:05
mriedemasettle: i guess we need a big fat warning in that page now right?19:05
cdentthank you jaypipes, I felt meh about dissing dansmith's +2, but seemed better to get it now instead of later19:06
jaypipesno worries, those were good comments, cdent.19:06
cdenti'm in review mode at the moment, so should get back round to the new version shortly19:06
* mdbooth has been trying to fix a unit test for about an hour, only to realise it's a bug in the code19:09
cdentthat means you win mdbooth19:09
mdboothIt's a mechanical change. I just added an additional assert and I couldn't see why the hell it didn't still work.19:10
mdboothdansmith: Is it possible to migrate an instance between cells? Excepting, of course that the answer's no because there's only 1 cell.19:12
dansmithmdbooth: eventually, but not currently19:12
*** david-lyle has quit IRC19:14
*** giri has quit IRC19:15
mdboothexternal_instance_event() associates a migrating instance with both its hosts. It then assumes that all instances it associated with a particular host have the same cell_mapping. However, if an instance were migrating to a new cell, it would be listed against the destination host, but have a different cell mapping to all other instances in that list.19:16
*** armax has joined #openstack-nova19:16
dansmithmdbooth: https://review.openstack.org/#/c/438025/719:18
*** annegentle has joined #openstack-nova19:18
mdboothdansmith: Sure. The assumption in this code is kinda obtuse though (I stared at it for an hour without noticing it). It'll break when you eventually allow that, and it won't be at all obvious that it needs to be fixed.19:19
mdboothdansmith: I wonder if it's worth addressing while it's in mind.19:20
dansmithmdbooth: the context in the instance doesn't have anything to do with where the rpc call will go19:20
dansmithand on the other side of the call, the context is reset to what was used for the rpc call19:21
*** jackie-truong has joined #openstack-nova19:21
*** Prophidys has quit IRC19:21
*** Prophidys has joined #openstack-nova19:22
mdboothdansmith: In this case the cell_mapping comes from a weird mappings dict passed in19:22
*** salv-orl_ has joined #openstack-nova19:22
mdboothThe bug is that if the first instance in the host list is actually from the wrong cell, it'll use the wrong cell context19:23
*** satyar has quit IRC19:23
cdentjaypipes: are we decided that step_size of 1 is okay on the ironic inventory because the modulo math comes out right in that (special) case?19:23
jaypipescdent: I think so19:23
cdentcool, just checking19:23
openstackgerritmelanie witt proposed openstack/nova master: Count resources to check quota for cells  https://review.openstack.org/41652119:24
mdboothdansmith: NM. I have to shoot, but I'm going to leave this open and look at it again first thing in the morning.19:25
dansmithmdbooth: okay I see what you mean, but I don't think that's right.. the instancemapping will point to the hostmapping where its record currently resides19:25
dansmithmdbooth: then we add in the other destination,19:25
*** salv-orlando has quit IRC19:25
dansmithbut the instance gets added to a host list based on instance.host, which matches hostmapping, and thus the cell19:25
mdboothdansmith: That's the bug, btw19:25
mdboothinstance is added to host list based on _get_relevant_hosts19:26
mdboothSo a migrating instance will be in *2* host lists, one of which it's not actually on19:26
mdboothIf the first entry in the list just happened to be a migrating instance which wasn't actually on this host19:26
*** sdague has quit IRC19:26
mdboothWe'd use its cell_mapping anyway19:26
dansmithah, this is actually what you mean:19:27
mdboothThat's ok as long as we're not migrating between cells19:27
dansmithhttps://github.com/openstack/nova/blob/master/nova/compute/api.py#L408219:27
mdboothBut it's seriously obtuse19:27
dansmithyeah, I see now19:27
dansmithI was looking at the api side19:27
mdboothdansmith: Yeah, that.19:27
* mdbooth should have linked the ode19:27
dansmithwe should just pass the mapping in from the api side instead of inferring it late there19:28
dansmithmy NOTE was probably really "be lazy and don't add a new parameter"19:28
*** sdague has joined #openstack-nova19:28
mdboothdansmith: As I say, I'll leave it open and poke at it in the morning19:29
dansmithalright19:29
openstackgerritDan Smith proposed openstack/nova master: Make discover_hosts only query for unmapped ComputeNode records  https://review.openstack.org/42790319:30
*** yamamoto has joined #openstack-nova19:30
*** sneti_ has joined #openstack-nova19:33
mriedemgah, mocking out a context manager and asserting calls on the mock is not fun19:33
*** yamamoto has quit IRC19:35
mriedemdocs bug if someone wants it https://bugs.launchpad.net/openstack-manuals/+bug/167161519:36
openstackLaunchpad bug 1671615 in openstack-manuals "Networking with nova-network in Administrator Guide" [Undecided,New]19:36
*** dharinic|lunch is now known as dharinic19:36
*** salv-orl_ has quit IRC19:37
dansmithmriedem: so this is my patch series to make the discovery faster: https://review.openstack.org/#/c/427903/19:37
dansmithmaybe should have a specless blueprint?19:37
mriedemyeah probably19:37
*** gfhellma has quit IRC19:38
mriedemmdbooth: dansmith: got any tricks for asserting calls on a context manager method?19:38
*** salv-orlando has joined #openstack-nova19:38
mriedemlike, i want to use mock.assert_has_calls,19:38
mriedemon just the args to the method,19:38
mriedembut the context manager enter/exit stuff messes that up19:38
dansmithmriedem: mock it with a side effect that is a context manager19:38
dansmiththere's some other way to mock just __enter__ but it's always obscure so I don't do it19:39
dansmithmriedem: I have been meaning to switch the way that fixture works a bit so we can more easily mock the non-context-manager set_target_cell in the tests19:39
efried__enter__ can be mock.patched like anything else, neh?19:39
mriedemyes, the problem is i don't need to mock __enter__ and __exit__19:40
mriedemthe problem is,19:40
mriedemassert_has_calls things i have 3 more calls than i actually care about19:40
dansmithyeah19:40
dansmithit's ugly19:40
dansmithI made it work one time,19:40
dansmithand then undid it19:40
mriedemmaybe i can just use mock.method_calls19:41
*** crusio has quit IRC19:41
mriedemwhich doesn't include magic methods19:41
*** crusio has joined #openstack-nova19:41
asettleSorry mriedem was off for the evening (I'm UK based), but basically, yes :) big fat warning.19:42
asettle"This is deprecated as of Ocata" type thing19:42
asettleThanks for noticing that!19:42
efriedmriedem https://github.com/powervm/pypowervm/blob/develop/pypowervm/tests/tasks/test_storage.py#L93-L9919:43
asettlecdent and jaypipes - how did you both get on?19:43
efriedmriedem Or perhaps https://github.com/powervm/pypowervm/blob/develop/pypowervm/tests/tasks/test_storage.py#L355-L363 is closer to what you want.19:43
jaypipesasettle: was there an updated revision pushed? I didn't see one?19:44
asettlejaypipes: nope. sorry, that's not what I meant. Just checking in to see if you or cdent had other thoughts/concerns since we last spoke19:44
mriedemefried: that's a bit ugly imo19:44
asettleI see your comments on patch 819:44
asettleThank you :)19:44
mriedemthe .__enter__.return_value19:44
cdentasettle: i put more comments on the review, but basically I was able to confirmation that the placement parts are okay19:44
*** karimb has quit IRC19:45
efriedmriedem Would it be less ugly if __enter__ wasn't a double-underscore method?19:45
openstackgerritDan Smith proposed openstack/nova master: Make discover_hosts only query for unmapped ComputeNode records  https://review.openstack.org/42790319:45
openstackgerritDan Smith proposed openstack/nova master: Add ability to query for ComputeNodes by their mapped value  https://review.openstack.org/42790219:45
openstackgerritDan Smith proposed openstack/nova master: Add ComputeNode.mapped field  https://review.openstack.org/42790119:45
asettleTerrific! Ugh thank you so much cdent :D19:45
jaypipesasettle: nope, besides those suggestions I put on the initial review, I'm good! :)19:45
*** READ10 has quit IRC19:47
asettleUgh you guys are excellent :D thank you both and to mriedem for looking at this all week :) Crossing my fingers that some of these edits should set us on the right path for both RDO and Ubuntu19:47
asettleI'll check back in with Brian in the morning.19:47
*** karimb has joined #openstack-nova19:48
openstackgerritchristopher.uhler proposed openstack/nova master: Libvirt volume driver for Veritas HyperScale  https://review.openstack.org/44395119:50
*** ducttape_ has joined #openstack-nova19:51
*** adrian_otto1 has joined #openstack-nova19:52
*** david-lyle has joined #openstack-nova19:53
*** ducttape_ has quit IRC19:56
*** adrian_otto has quit IRC19:56
mriedemguh20:00
mriedemannoying when you have the fix for a thing all day, but spend 2 hours trying to get tests to work20:00
mriedem*unit* tests20:00
cdentmriedem: after all this time you haven't been able to accept that as the new normal?20:01
mriedemsometimes i don't mind,20:01
mriedemi only mind when i have other stuff to be doing20:02
*** kaisers_ has joined #openstack-nova20:02
*** mtanino has quit IRC20:02
efriedmriedem Then you have to rebase the thing...20:04
mriedemi do actually have to rebase this20:06
mriedembut want it working first20:06
mriedemi want the *mocks* working first20:06
* dansmith hands mriedem pizza and cookies20:07
* melwitt plays "you're the best" montage for mriedem 20:08
*** salv-orlando has quit IRC20:08
mriedemgee thanks everyone20:08
mriedemwhat i really need is coffee20:09
dansmithmriedem: basically, we all acknowledge your mundane plight, now get back to work20:09
mriedemand metal20:09
*** kfarr has quit IRC20:09
mriedemha20:09
dansmith:)20:09
*** yamahata has joined #openstack-nova20:09
mriedemi need comstud here to kick me in my ass20:10
*** gfhellma has joined #openstack-nova20:11
* dansmith puts on his red foreman costume to kick eric "riedeman" foreman in his ass20:11
efriedI could put you in an armbar, but that would be counterproductive to your typing abilities.20:11
melwittwhat's special about comstud kicking you? what about us??20:12
* dansmith points at the boots and makes scardy eyes20:12
dansmither, "scaredy"20:12
*** kfarr has joined #openstack-nova20:15
*** Anticimex has quit IRC20:21
mriedemmelwitt: because you'd probably apologize afterward20:22
mriedemcomstud on the other hand...20:22
melwittheh20:22
openstackgerritSujitha proposed openstack/nova master: Add description to policies in admin_password.py  https://review.openstack.org/44395720:26
*** slaweq_ has quit IRC20:27
mriedemf yes20:28
*** dtp has joined #openstack-nova20:29
*** slaweq_ has joined #openstack-nova20:29
*** dtp has quit IRC20:30
*** dave-mccowan has quit IRC20:30
*** dtp has joined #openstack-nova20:31
*** Anticimex has joined #openstack-nova20:31
*** yamamoto has joined #openstack-nova20:31
*** jackie-truong has quit IRC20:33
*** yamamoto has quit IRC20:35
*** raunak has quit IRC20:36
openstackgerritSujitha proposed openstack/nova master: Add description to policies in admin_password.py  https://review.openstack.org/44395720:36
*** tasker has joined #openstack-nova20:36
*** fragatina has quit IRC20:36
taskeris there a way to "build" a new instance in an un-powered state?20:37
*** Anticimex has quit IRC20:40
*** Anticimex has joined #openstack-nova20:40
*** sdague has quit IRC20:41
taskerb'ah. probably the wrong channel, but I'm trying to do the above programatticaly.20:42
dansmithyou're revising that quota set and un-WIPing them all I assume right?20:42
*** raunak has joined #openstack-nova20:43
*** xyang1 has joined #openstack-nova20:43
*** sree has joined #openstack-nova20:44
mriedemdansmith: yeah20:44
mriedemplumber called, more distractoins20:44
*** awaugama has quit IRC20:46
efriedtasker That doesn't sound... easy.  I can think of some ways to hack it.  But I don't think the APIs provide anything.20:46
taskerefried: i was afraid of that.20:46
*** dave-mcc_ has joined #openstack-nova20:47
efriedtasker Example: create it with no boot disk, and it (maybe) boots into grub/OF.  The nova side will see this as "active", so the spawn will succeed, but the VM didn't _really_ boot.20:47
*** sree has quit IRC20:48
taskeri have an instance with two volumes, one boot, one mounted at /usr. not ideal, but not my instance. i need to be able to, at will, make a backup of the volumes, restore them, and spin up a new copy of the instance. but if the instance boots from 'vda' alone, it'll have problems with no 'vdb' and I can't attach a volume until it's active.20:48
taskerefried: that's not too bad.20:49
mriedemdansmith: like the water in my basement, here comes the flood20:49
openstackgerritMatt Riedemann proposed openstack/nova master: Decrement quota when deleting an instance in cell0  https://review.openstack.org/44339520:49
openstackgerritMatt Riedemann proposed openstack/nova master: Temporarily untarget context when deleting from cell0  https://review.openstack.org/44340320:49
efried...can't attach volume until active, why?20:49
taskerConflict: Cannot 'attach_volume' while it is in vm_state building20:49
mriedemtasker: i think cburgess is trying to solve a similar issue20:50
mriedemat least with that conflict20:50
mriedemyou should join forces20:50
taskeroh? do you have a commit review i can check out?20:50
*** ducttape_ has joined #openstack-nova20:52
taskerI'll checkout gerrit and get send a message their way.20:52
mriedemsec20:52
mriedemi thought this https://review.openstack.org/#/c/41754120:54
mriedemtasker: see L33 here for PTG discoussion https://etherpad.openstack.org/p/nova-ptg-pike-cinder20:54
mriedem*discussion20:54
*** priteau has quit IRC20:54
*** JoseMello has quit IRC20:55
taskerthanks! I'll check it out.20:55
mriedemdansmith: are you checking those changes out? if they look ok at first glance, i'll start working on rebasing on top of gibi's changes20:55
dansmithI will yeah20:55
*** ducttape_ has quit IRC20:56
*** dimtruck is now known as zz_dimtruck20:56
openstackgerritmelanie witt proposed openstack/nova master: Count resources to check quota for cells  https://review.openstack.org/41652120:57
dansmithgross20:58
dansmithwell, I don't like the context mocking20:58
dansmithbut I shan't hold up on that20:58
*** gjayavelu has joined #openstack-nova21:00
cburgessmriedem tasker What did I do?21:00
*** rfolco has quit IRC21:00
cburgessmriedem Oh that thing yeah... so thats now part of johnthetubaguy spec on moving to the new cinder attach api21:00
*** smatzek_ has quit IRC21:01
*** gfhellma has quit IRC21:03
mriedemdansmith: for the assertions?21:03
dansmithmriedem: yeah21:03
mriedemor just the 20 context calls?21:03
dansmiththe context calls21:03
mriedemoh21:03
mriedemyeah, well,21:03
mriedemsee ^ for the last 3 hours :)21:03
*** fragatina has joined #openstack-nova21:03
dansmithheh yeah I know21:03
*** eharney has quit IRC21:05
mriedemdansmith: comments in https://review.openstack.org/#/c/437222/21:05
mriedemthe notification one21:05
cburgessmriedem So https://review.openstack.org/#/c/417541 should be abandoned. I'll have yuming do that shortly. Our work is now covered by johnthetubaguy work in https://review.openstack.org/#/c/373203/21:06
*** gfhellma has joined #openstack-nova21:07
mriedemcburgess: i can abandon it if you want21:07
cburgessGo ahead21:07
cburgessmriedem Abandon with great abandon.21:07
dansmithmriedem: so in the delete_while_booting, we can't soft delete, so it's always going to be a hard instance-is-gone21:07
*** tbachman has quit IRC21:08
dansmithbecause we're just deleting the mapping21:08
dansmithwell, wait21:08
dansmithyeah, right,21:08
dansmithbecause if we deleted that, then we're going to end up with no instance record when the conductor realizes that we won21:09
*** salv-orlando has joined #openstack-nova21:09
dansmithL171221:09
*** zioproto has joined #openstack-nova21:09
*** Jack_I has quit IRC21:10
dansmithwell, wait,21:10
mriedemit seems like we just don't support soft_delete for the case that the instance isn't in a cell yet21:11
dansmithyeah21:11
dansmithright21:11
*** gyee has joined #openstack-nova21:11
mriedemsoft_delete is also totally dependent on the compute deleting the instance21:11
mriedemand if it never landed in a cell, there is no compute to clean it up21:11
mriedemseems like a good comment to add in the code as a reminder of that21:12
mriedemwe could do that in a follow up, like my ask for a comment about why we avoid the lazy-load in the context manager for sys meta21:12
mriedemcburgess: done, and the bp in lp too21:12
*** ujjain has quit IRC21:13
cburgesscool21:13
mriedemthis patch still doesn't have a test to make sure that the soft_delete case is handled when the instance is in cell021:13
dansmithmriedem: yeah21:13
dansmithmriedem: I can do that, or you can throw it in while you're stacking your others on top here21:13
mriedemlet me think about what i just said,21:14
dansmithyeah, that...21:14
dansmithwould be a problem21:14
mriedemif it's in cell0, again, it's we can't soft delete it21:14
dansmithbut there's also no resources to clean up21:14
mriedemhttps://review.openstack.org/#/c/437222/7/nova/compute/api.py@212621:15
dansmiththat cast is just going to go into the ether21:16
dansmithif you delete it again, it hard deletes right?21:16
openstackgerritSarafraj Singh proposed openstack/nova-specs master: WIP: Replace live migration steps and delays with memory iteration logic  https://review.openstack.org/44333921:18
dansmithmriedem: from the api perspective, what we have _might_ make the most sense21:20
dansmithyou soft delete, it stays around half deleted,21:20
dansmithbut never gets culled because there is no compute for it21:20
dansmiththen you hard delete it to make it go away21:20
dansmithon the other hand,21:20
mriedemyou mean force-delete?21:21
dansmithjust making it kill the instance immediately if it's in cell0 also makes sense, since there's no reason to keep a dead instance around21:21
dansmithsoft-delete being for "oops, I deleted a thing that I wanted to keep"21:21
mriedemi don't think force-delete would fix anytihng here21:21
mriedembut i'm not following the half-deleted case21:22
openstackgerritSarafraj Singh proposed openstack/nova-specs master: Spec to add per instance timeout for LM  https://review.openstack.org/44280921:22
mriedemi put some comments in there if you want to read those and reply21:22
dansmithmriedem: https://github.com/openstack/nova/blob/master/nova/api/openstack/compute/servers.py#L782-L78521:22
mriedembasically i think the only issue now is the notification is lying if it says we do a soft-delete but we actually hard-delete in cell021:22
mriedemwhat triggers that InstanceInvalidState though?21:23
dansmithmriedem: lies how? because it doesn't call it a soft delete?21:23
openstackgerritSarafraj Singh proposed openstack/nova-specs master: Add live-migration-force-after-timeout spec  https://review.openstack.org/43846721:23
*** liverpooler has quit IRC21:23
mriedemmust_have_launched=True21:23
*** tasker has quit IRC21:23
dansmithmriedem: must_have_launched21:23
dansmithyeah21:23
*** aysyd has quit IRC21:24
mriedemwhat sets launched_at? compute?21:24
mriedemyup _update_instance_after_spawn21:24
dansmithyeah21:24
mriedemsee https://review.openstack.org/#/c/437222/7/nova/compute/api.py@179421:25
mriedemfor the lie21:25
*** jaosorior has joined #openstack-nova21:25
dansmithoh, calls it a soft delete when it's not21:26
mriedemyeah21:26
mriedemi mean that's very minor21:26
mriedembut i think it's technically wrong21:26
dansmithif we do that,21:26
dansmithhmm21:26
*** ujjain has joined #openstack-nova21:27
*** ujjain has quit IRC21:27
*** ujjain has joined #openstack-nova21:27
dansmithyou can get there if you have a non-cell0 cell too right?21:27
mriedembut no host?21:27
dansmithif you race with compute owning the instance21:27
dansmiththe instance_mapping has a cell,21:27
dansmithbecause scheduler decided,21:27
dansmithbut compute hasn't instance.host=$me yet21:27
mriedemwhat sets instance.host in the compute?21:29
mriedemfor spawn i mean, i'm not seeing that21:29
dansmithyeah, hang on21:29
dansmithI always have to dig for this I dunno why21:30
*** zz_dimtruck is now known as dimtruck21:30
*** tbachman has joined #openstack-nova21:30
melwittit's in instance_claim in the RT21:31
dansmithoh, that's probably why21:31
mriedemi figured that gd low down RT was involved21:31
dansmithhttps://github.com/openstack/nova/blob/master/nova/compute/resource_tracker.py#L336-L34421:31
melwitthaha, yeah. it's hidden there, for fun!21:32
*** yamamoto has joined #openstack-nova21:32
mriedemRT is like a leprachaun21:32
mriedemfull of tricks21:32
*** jaosorior has quit IRC21:33
mriedemdansmith: ok so back to https://review.openstack.org/#/c/437222/7/nova/compute/api.py@1794 if it's not in cell0 and it doesn't have a host yet, but let's say it's in cell1,21:33
mriedemwe still delete the instance21:34
mriedemso i think that just means, when the reclaim task (or whatever it's called) in the compute runs, it just won't see that instance21:34
mriedemand that's fine, it's already gone21:34
mriedemso it was an immediate soft delete :)21:34
mriedemright?21:34
*** Jeffrey4l_ has quit IRC21:35
dansmithunless we got here because instance wasn't created in the database21:35
dansmithbut I _think_ that means it's gone-gone too21:35
mriedemwhich database?21:35
mriedemcell db?21:35
*** Jeffrey4l_ has joined #openstack-nova21:35
dansmithyeah the only place instances get created21:36
dansmithI think we can also fail to delete there if host changes, right?21:36
dansmithwe're in a try..except there21:36
*** dave-mcc_ has quit IRC21:36
dansmithand I think that's for the "if host changed" case,21:36
dansmithwhere we can either fall through or bail21:36
dansmithI think we might be over-obsessing here21:37
mriedemyeah if the host changes we're here https://github.com/openstack/nova/blob/master/nova/compute/api.py#L181621:37
mriedemand then it falls through to the normal delete path21:37
dansmithI'm not sure whether notifications consumers will do something useful with the soft/hard indication21:37
mriedemi agree this is probably not the end of the world21:37
*** yamamoto has quit IRC21:38
dansmiththe missing notifications, and the quota bug behind this are definitely more important21:38
dansmithgibi might have a better handle on how much this does or does not matter21:38
melwittI was thinking my main concern would be whether the start/end notifications match, and also that they match the request that was issued from the user21:39
melwittfor the first, I saw the context manager makes sure the start/end match21:39
dansmiththe context manager makes sure they match I think21:39
melwittfor the last, I guess that's where the delete_type comes in. I was just trying to think, how someone might use this and if they're listening to notifications, they may also be looking at what requests were issued and correlating notifications with those21:40
melwittto get like a bundle of "what happened" which would be 1) user issued delete_type X 2) nova emitted delete_type X start 3) nova emitted delete_type X end21:41
*** krtaylor has quit IRC21:42
*** tbachman has quit IRC21:42
dansmithI'm not sure what the point of correlating the requests and the notifications really is, but we don't really know what (all) people use these for21:42
melwittif they're doing log collecting and correlating. anyway, that's the thing I could think of it mattering21:42
dansmithmriedem: I guess my point is, this is demonstrably better than what we have now, I think, especially considering the quota thing behind it21:43
*** tbachman has joined #openstack-nova21:43
dansmithin some very teensy windows of a less-mainstream feature, you might get a delete type you didn't expect21:43
melwittat my last company, use used the old notification driver that emits notifications to the logs. and we had those in splunk and correlated things like that, to look at when people opened bugs etc21:43
melwittI think what I'm saying is what the patch is doing, so it makes sense to me to keep everything saying "soft delete" if the user issued "soft delete"21:45
*** dave-mccowan has joined #openstack-nova21:45
dansmithwell, it's not in the first case he brought up21:45
mriedemthe user doesn't technically request "soft" delete21:46
mriedemit's all driven by config21:46
*** kaisers_ has quit IRC21:46
mriedemit's totally not interoperable21:46
melwitterg, right. notifications probably have the request-id too right? so that's what anyone would use to correlate anyway21:46
dansmithyeah21:47
melwittokay, that addresses what I was thinking21:47
*** hamzy has quit IRC21:48
mriedemso i think i might still pull this down, add some comments, and add a test for the soft-delete case where there isn't a host but it's in a cell (any cell) and the build request was already gone21:48
mriedembecause i'm pretty sure i'm right here https://review.openstack.org/#/c/437222/7/nova/tests/unit/utils.py21:49
mriedemthen i'll rebase my stuff on top of it21:49
*** markmcclain has joined #openstack-nova21:49
mriedemand then it's pants off dance off21:49
*** aojea has quit IRC21:49
*** jose-phi_ has joined #openstack-nova21:51
*** gyee has quit IRC21:51
*** ducttape_ has joined #openstack-nova21:52
*** jose-phillips has quit IRC21:52
dansmithso, wait,21:53
dansmithwe notify delete.started..delete.ended in the api, or soft_delete.started..ended,21:53
dansmithand then again in compute in both cases?21:53
dansmithso if you normal delete, you get two delete.started notifications/21:53
dansmithand if soft delete then you get the soft one first and the regular one later?21:53
*** ducttape_ has quit IRC21:57
*** sneti_ has quit IRC21:57
melwittthe notifys should be limited to the local delete cases, right?21:59
melwittthe new ones, I mean21:59
*** crushil has quit IRC21:59
mriedemmelwitt: i think so https://review.openstack.org/#/c/437222/7/nova/compute/api.py@186621:59
melwittso they wouldn't be going on to compute to get a second notify21:59
mriedemright22:00
mriedemand where gibi has made this change, it's a local delete22:00
dansmithnot if we fail the last race check, but yeah22:00
dansmithso do we ever emit a soft-delete notification normally?22:01
dansmithmeaning before all of this changed..22:02
dansmithonly in the local delete, which is totally not a soft delete, is that the only time we emit soft-deleted.start?22:02
*** edmondsw has quit IRC22:03
mriedemit looks like it...22:03
dansmithhow stupid is that?22:04
dansmiththe only time we emit soft-deleted is when .. we didn't22:04
mriedemdown in _delete_instance in the compute manager, even if it was a soft_delete, we issue instance.delete.start22:04
melwittI wonder how soft-deleted notify ever came about then. because like you said, there isn't really a soft-delete from the user perspective22:06
mriedemdansmith: that's a good point22:06
dansmithmelwitt: right, I can't think of any reason it makes sense to emit soft-deleted22:06
dansmithexcept maybe in the case where we're deferring it, which we _don't_ notify about22:06
*** rfolco has joined #openstack-nova22:08
mriedemwe'll also issue soft_delete.start/end in the case that the instance is shelved_offloaded, because that's another local delete case22:08
*** rfolco has quit IRC22:08
dansmithyeah22:08
mriedemso yeah, seems totally pointless to even distinguish22:08
mriedemsince you can't restore it anyway22:08
dansmithwe totally cover all the cases where we're not actually soft deleting22:08
mriedemit's mfing gone22:08
mriedemso going to back to https://review.openstack.org/#/c/437222/7/nova/compute/api.py@1792 i don't think we should even pass delete_type there, it's not a soft_delete22:09
mriedemi know we're talking in circles, but you could argue that we should never send soft_delete.start/end22:09
dansmithI think I would be in favor of just never emit soft_delete22:09
mriedemyeah22:09
mriedemfor the existing parts that do this, i'd leave that for a separate change22:10
dansmithso for this patch, I say remove that one case where we do,22:10
mriedemyeah22:10
dansmithand then yeah clean up the others22:10
mriedemcool22:10
mriedemhey, high five22:10
dansmithfuck you you -1d my +2, no high five for you22:10
dansmithhmm, that sounded darker and more non-sensical ranty than I meant it to :P22:11
* dansmith digs up the rainbows and lollipops link22:11
mriedemi stand by my -122:12
*** cleong has quit IRC22:13
*** jose-phi_ has quit IRC22:14
*** jose-phillips has joined #openstack-nova22:19
mriedemcrap22:22
mriedemall 3 speaking sessions at the summit were accepted22:22
mriedemarmax: from boot to woot22:23
*** gyee has joined #openstack-nova22:24
*** adrian_otto1 has quit IRC22:28
*** ducttape_ has joined #openstack-nova22:28
armaxmriedem: I just saw the emaik22:30
armaxemail22:30
*** adrian_otto has joined #openstack-nova22:30
armaxwe should probably sit down and talk about it at some point22:30
*** slaweq_ has quit IRC22:30
*** lucasxu has quit IRC22:31
*** edmondsw has joined #openstack-nova22:33
mriedemyeah22:34
tonybsean-k-mooney: It's all good.  As you can see I'm trying to work with you but for the most part I'm learning on my feet.  It'd be cool if we had a clear direction on what level of support we have for Windows platforms.  Are they for development? or just runtime?  I'm treatign win32 vs cygwin much like Ubuntu vs Fedora.  Anyway we don't overlap much but we can keep going back and forth on the review22:37
*** edmondsw has quit IRC22:38
*** lpetrut has quit IRC22:39
*** thorst has quit IRC22:39
*** karimb has quit IRC22:39
*** catintheroof has joined #openstack-nova22:39
*** gfhellma has quit IRC22:40
*** gfhellma_ has joined #openstack-nova22:40
openstackgerritMatt Riedemann proposed openstack/nova master: Fix missing instance.delete notification  https://review.openstack.org/43722222:40
*** jamielennox is now known as jamielennox|away22:42
*** annegentle has quit IRC22:43
*** sree has joined #openstack-nova22:44
*** jamielennox|away is now known as jamielennox22:45
*** gouthamr has quit IRC22:48
*** gyee has quit IRC22:48
*** sree has quit IRC22:48
*** adrian_otto has quit IRC22:49
*** jamesdenton has quit IRC22:51
*** priteau has joined #openstack-nova22:55
mriedemgreat, now the functional test isn't passing after i rebased on top of the notification fix22:57
*** simondodsley has quit IRC22:57
*** felipemonteiro_ has quit IRC22:58
*** ducttape_ has quit IRC22:59
*** gyee has joined #openstack-nova23:00
*** iceyao has joined #openstack-nova23:00
*** priteau has quit IRC23:00
*** thorst has joined #openstack-nova23:03
*** yassine has joined #openstack-nova23:04
*** gyee has quit IRC23:04
*** armax has quit IRC23:04
*** ducttape_ has joined #openstack-nova23:04
*** iceyao has quit IRC23:05
*** gyee has joined #openstack-nova23:07
*** thorst has quit IRC23:07
*** gouthamr has joined #openstack-nova23:11
openstackgerritSamantha Blanco proposed openstack/nova master: [WIP] Make flavor-rxtx policy more granular  https://review.openstack.org/44403623:12
*** diga has joined #openstack-nova23:19
*** baoli has quit IRC23:20
*** baoli has joined #openstack-nova23:21
*** zioproto has quit IRC23:21
*** esberglu has quit IRC23:21
*** baoli has quit IRC23:26
*** dimtruck is now known as zz_dimtruck23:28
mriedemdansmith: ha23:31
mriedemso,23:31
mriedemthe notification patch doesn't have a test that hits notify_about_instance_delete23:31
mriedembut mine on top of it do23:31
mriedemUnboundLocalError: local variable 'system_metadata' referenced before assignment23:31
mriedemfixing23:32
*** esberglu has joined #openstack-nova23:33
*** gyee has quit IRC23:33
*** xyang1 has quit IRC23:33
*** yamamoto_ has joined #openstack-nova23:35
efriedmriedem That whole initializing-a-variable-within-a-try-and-then-using-it-in-the-except-or-finally thing...23:35
mriedemefried: i even thought about it when i was looking at it before23:35
mriedemefried: er i mean, it was a test to see if you would have caught and you didn't!23:35
efriedmriedem Rat farts!  I _did_ review that change set!23:36
mriedemrat farts indeed23:37
*** esberglu has quit IRC23:37
*** marst has quit IRC23:37
*** marst has joined #openstack-nova23:37
*** yamamoto_ has quit IRC23:40
*** cdent has quit IRC23:41
*** dave-mccowan has quit IRC23:43
openstackgerritDan Smith proposed openstack/nova master: Get instance availability_zone without hitting the api db  https://review.openstack.org/43975423:46
openstackgerritDan Smith proposed openstack/nova master: Teach HostAPI about cells  https://review.openstack.org/44216223:46
openstackgerritDan Smith proposed openstack/nova master: Make os-availability-zones know about cells  https://review.openstack.org/44216323:46
openstackgerritDan Smith proposed openstack/nova master: Change MQ targeting to honor only what is in the context  https://review.openstack.org/43802323:46
openstackgerritDan Smith proposed openstack/nova master: Make scheduler target cells to get compute node instance info  https://review.openstack.org/43989123:46
openstackgerritDan Smith proposed openstack/nova master: Sort CellMappingList.get_all() for safety  https://review.openstack.org/44317423:46
openstackgerritDan Smith proposed openstack/nova master: Add get_by_instance_uuids() to InstanceMappingList  https://review.openstack.org/44329223:46
openstackgerritDan Smith proposed openstack/nova master: Make server_groups determine deleted-ness from InstanceMappingList  https://review.openstack.org/44329323:46
openstackgerritDan Smith proposed openstack/nova master: Remove Mitaka-era service version check  https://review.openstack.org/44286123:46
openstackgerritDan Smith proposed openstack/nova master: Add cell field to Destination object  https://review.openstack.org/43802423:46
openstackgerritDan Smith proposed openstack/nova master: Make conductor ask scheduler to limit migrates to same cell  https://review.openstack.org/43802523:46
openstackgerritDan Smith proposed openstack/nova master: Add workaround to disable group policy check upcall  https://review.openstack.org/44273623:46
openstackgerritDan Smith proposed openstack/nova master: Add identity helper property to CellMapping  https://review.openstack.org/44403823:46
*** tblakes has quit IRC23:46
*** kaisers_ has joined #openstack-nova23:46
*** ducttape_ has quit IRC23:48
mriedemdansmith: i hope you ran pep823:50
*** kaisers_ has quit IRC23:51
* melwitt is going to have to start saying "rat farts" more23:52
melwittpep8 takes forever to run now, like before it even runs flake8. dunno what it's doing23:53
*** zz_dimtruck is now known as dimtruck23:54
mriedemmelwitt: you know about: tox -e pep8 -- -HEAD?23:55
mriedemcinder has fast823:55
mriedemsame thing23:55
melwittno, what does that do?23:55
mriedemruns only on the changes in the commit23:55
mriedem-HEAD23:55
mriedemdanpb wrote that quite a while ago23:55
mriedemhttps://github.com/openstack/nova/blob/master/tools/flake8wrap.sh#L1223:56
melwittoh, hm23:56
mriedemi never run full pep8 on nova anymore23:56
melwittwow. thanks23:56
efriedSecond that.23:57
*** eglynn has quit IRC23:57
*** ducttape_ has joined #openstack-nova23:58

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!