Wednesday, 2016-08-03

*** code-R has quit IRC00:00
*** diogogmt has quit IRC00:00
jaypipesdansmith: hrmmm..00:00
jaypipesdansmith: yeah, gertty sync problem. pushed now.00:01
dansmiththanks00:01
jaypipesdansmith: guh, still pushing.00:01
*** TravT has joined #openstack-nova00:01
*** moshele has joined #openstack-nova00:02
*** thumpba has joined #openstack-nova00:03
mriedemsfinucan: or is that because we need https://review.openstack.org/#/c/337108/ stacked on top which is passing intel nfv ci?00:03
*** shaohe_feng has quit IRC00:04
jaypipesdansmith: now done syncing..00:04
jaypipesdansmith: please see my concern on https://review.openstack.org/#/c/314782/28 btw.00:04
*** shaohe_feng has joined #openstack-nova00:04
*** code-R_ has quit IRC00:05
*** code-R has joined #openstack-nova00:05
dansmithjaypipes: replying00:06
jaypipesdansmith: baited breath.00:07
*** sameo has quit IRC00:07
dansmithjaypipes: replied00:10
*** TravT has quit IRC00:10
*** mtanino has quit IRC00:11
*** thumpba has quit IRC00:12
openstackgerritJay Pipes proposed openstack/nova: rt: refactor test_dupe_filter unit test  https://review.openstack.org/35031900:12
jaypipesdansmith: ahhhhhh00:13
jaypipesdansmith: I totally friggin missed that...00:13
dansmithI forgive you :)00:13
*** thumpba has joined #openstack-nova00:14
*** thumpba has quit IRC00:14
*** shaohe_feng has quit IRC00:14
*** shaohe_feng has joined #openstack-nova00:14
jaypipesdansmith: approved and wallaby'd. and gertty-sync'd.00:15
mriedemYingxin: do you still need this or can it be abandoned? https://review.openstack.org/#/c/306301/00:15
dansmithjaypipes: thankyasir00:15
*** TravT has joined #openstack-nova00:15
jaypipesdansmith: for you? any time.00:16
* dansmith sets his alarm for 2am EDT Sunday to ask jaypipes for a review00:17
jaypipesdansmith: review bait: https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:master+topic:rt-unit-tests00:17
mriedemoh wow we're using cascading deletes now?00:17
jaypipesmriedem: erm, no :)00:17
dansmithjaypipes: I've been waiting for that to be unblocked, so yeah, cool00:17
mriedemthen what is https://review.openstack.org/#/c/314781/26/nova/db/sqlalchemy/api_models.py ?00:17
*** devkulkarni1 has quit IRC00:17
jaypipesmriedem: see my comment in there :)00:18
jaypipesthat apparently gertty duplicated...00:18
mriedemyeah it's not used00:18
dansmithyeah, sofrware-level cascade00:18
dansmither, software00:18
jaypipesdansmith: not even :)00:18
jaypipesdansmith: since you use query().filter().delete() it won't even be in use.00:19
dansmithyou mean doffm :)00:19
jaypipesdansmith: yeah, the doffster.00:19
dansmithI thought this would in fact generate the deletes for us00:19
dansmithbut I guess not since I had to do them manually.. I just remembered that00:19
*** TravT has quit IRC00:19
mriedemwe should just remove that in a follow up00:19
mriedemto avoid confusion00:19
dansmithsure00:20
*** jwcroppe_ has joined #openstack-nova00:20
*** jwcroppe has quit IRC00:22
*** ssurana has quit IRC00:23
*** shaohe_feng has quit IRC00:24
*** TravT has joined #openstack-nova00:25
*** shaohe_feng has joined #openstack-nova00:25
*** browne has quit IRC00:28
*** jwcroppe has joined #openstack-nova00:28
*** jwcroppe_ has quit IRC00:31
*** thorst has joined #openstack-nova00:31
*** rfolco_ has joined #openstack-nova00:32
*** rfolco_ has quit IRC00:32
*** baoli has joined #openstack-nova00:33
*** kaisers_ has joined #openstack-nova00:34
*** shaohe_feng has quit IRC00:35
*** shaohe_feng has joined #openstack-nova00:35
*** TravT has quit IRC00:38
*** kaisers_ has quit IRC00:38
*** thorst_ has joined #openstack-nova00:39
openstackgerritJay Pipes proposed openstack/nova: rt: refactor test_dupe_filter unit test  https://review.openstack.org/35031900:39
openstackgerritJay Pipes proposed openstack/nova: rt: Refactor test_dupe_filter unit test  https://review.openstack.org/35031900:39
*** hoangcx has joined #openstack-nova00:40
*** thorst has quit IRC00:41
*** thorst_ is now known as thorst00:41
*** TravT has joined #openstack-nova00:43
*** hoangcx2 has quit IRC00:43
*** baoli has quit IRC00:45
*** shaohe_feng has quit IRC00:45
*** shaohe_feng has joined #openstack-nova00:45
*** gongysh has joined #openstack-nova00:49
*** gongysh_ has joined #openstack-nova00:51
*** code-R has quit IRC00:54
*** shaohe_feng has quit IRC00:55
*** shaohe_feng has joined #openstack-nova00:55
*** TravT has quit IRC00:56
*** sdake has quit IRC00:57
*** sdake has joined #openstack-nova00:59
*** TravT has joined #openstack-nova01:01
*** phuongnh has joined #openstack-nova01:01
*** code-R has joined #openstack-nova01:02
*** shaohe_feng has quit IRC01:05
*** shaohe_feng has joined #openstack-nova01:06
*** diogogmt has joined #openstack-nova01:09
mriedemoomichi: does this make sense?01:10
mriedemコンサートで01:10
oomichimriedem: わかんない01:11
oomichimriedem: I cannot understand :)01:11
mriedemheh01:12
mriedemthat's what google told me the translation for 'in concert' was01:12
*** Sukhdev has quit IRC01:12
oomichimriedem: yeah, I can see. but cannot see the meaning01:12
mriedemjointly01:12
oomichimriedem: that is stylish English for me01:12
*** Apoorva has quit IRC01:12
mriedemi copied that line from sdague's spec so that's why01:12
mriedemhe's fancy pants01:12
oomichimriedem: I know, sdague's English is always stylish01:13
*** TravT has quit IRC01:14
*** shaohe_feng has quit IRC01:16
*** shaohe_feng has joined #openstack-nova01:16
Yingxinmriedem: I can abandon that01:16
*** mriedem has quit IRC01:19
eliqiaooomichi when you say stylish , does that mean 流行する ?01:21
*** akshai has joined #openstack-nova01:23
*** Apoorva has joined #openstack-nova01:23
*** dimtruck is now known as zz_dimtruck01:23
oomichieliqiao: I am using stylish as good-looking or Foppery ;)01:24
oomichieliqiao: yea, almost same as the above JP lang01:24
*** browne has joined #openstack-nova01:25
eliqiaooomichi: haha, get it. :) , OpenStack is so international.01:25
*** shaohe_feng has quit IRC01:26
*** shaohe_feng has joined #openstack-nova01:27
*** delattec has quit IRC01:28
Qiminghi, can someone tell me if nova emits a notification when a nova-compute service is down (could be any reason)?01:31
*** tangchen has quit IRC01:33
*** akshai has quit IRC01:35
*** shaohe_feng has quit IRC01:36
openstackgerritFang He proposed openstack/nova: use the HostPortGroupSpec.vswitchName instead of HostPortGroup.vswitch.split  https://review.openstack.org/35033501:37
*** shaohe_feng has joined #openstack-nova01:37
*** stewie925_ has quit IRC01:38
*** tangchen has joined #openstack-nova01:39
*** Apoorva_ has joined #openstack-nova01:40
*** TravT has joined #openstack-nova01:43
*** sdake has quit IRC01:44
*** Apoorva has quit IRC01:44
*** EinstCrazy has joined #openstack-nova01:44
*** browne has quit IRC01:44
*** Apoorva_ has quit IRC01:44
*** shaohe_feng has quit IRC01:46
*** shaohe_feng has joined #openstack-nova01:47
*** thorst has quit IRC01:47
*** TravT has quit IRC01:47
*** thorst has joined #openstack-nova01:48
*** oomichi has quit IRC01:48
*** baoli has joined #openstack-nova01:49
*** burt has quit IRC01:51
*** TravT has joined #openstack-nova01:52
*** thorst has quit IRC01:56
*** shaohe_feng has quit IRC01:57
*** chenfan has quit IRC01:57
*** unicell1 has quit IRC01:57
*** shaohe_feng has joined #openstack-nova01:57
*** zz_dimtruck is now known as dimtruck01:58
*** TravT has quit IRC02:01
*** yb has joined #openstack-nova02:04
*** yb has left #openstack-nova02:05
*** TravT has joined #openstack-nova02:05
*** shaohe_feng has quit IRC02:07
*** shaohe_feng has joined #openstack-nova02:08
*** chenfan has joined #openstack-nova02:09
*** kenperkins has quit IRC02:11
openstackgerritTina Kevin proposed openstack/nova-specs: Add allocation ratio to hypervisor-show  https://review.openstack.org/35034402:17
*** zerda2 has joined #openstack-nova02:17
*** shaohe_feng has quit IRC02:17
*** shaohe_feng has joined #openstack-nova02:18
*** TravT has quit IRC02:18
*** yb has joined #openstack-nova02:18
*** coolsvap has joined #openstack-nova02:19
*** rrecio_ has quit IRC02:19
*** TravT has joined #openstack-nova02:19
openstackgerritTina Kevin proposed openstack/nova-specs: Add allocation ratio to hypervisor-show  https://review.openstack.org/35034802:21
*** moshele has quit IRC02:22
*** kaisers_ has joined #openstack-nova02:23
*** TravT has quit IRC02:24
*** kaisers_ has quit IRC02:27
*** shaohe_feng has quit IRC02:27
*** shaohe_feng has joined #openstack-nova02:28
*** TravT has joined #openstack-nova02:28
*** yamahata has quit IRC02:28
*** code-R has quit IRC02:36
*** shaohe_feng has quit IRC02:38
*** shaohe_feng has joined #openstack-nova02:39
*** markvoelker has joined #openstack-nova02:39
*** dave-mccowan has joined #openstack-nova02:41
*** shaohe_feng has quit IRC02:48
*** shaohe_feng has joined #openstack-nova02:49
*** thorst has joined #openstack-nova02:54
*** TravT has quit IRC02:54
*** TravT has joined #openstack-nova02:54
*** TravT has quit IRC02:55
*** takashin has joined #openstack-nova02:56
*** pwnall has quit IRC02:57
*** gjayavelu has quit IRC02:58
*** shaohe_feng has quit IRC02:58
*** hemanthm|afk is now known as hemanthm02:59
*** shaohe_feng has joined #openstack-nova03:01
*** thorst has quit IRC03:02
*** code-R has joined #openstack-nova03:03
*** oanson has joined #openstack-nova03:03
*** rama_y has quit IRC03:04
*** dave-mccowan has quit IRC03:08
*** shaohe_feng has quit IRC03:08
*** shaohe_feng has joined #openstack-nova03:09
*** oanson has quit IRC03:11
*** code-R has quit IRC03:14
*** gongysh has quit IRC03:14
*** code-R has joined #openstack-nova03:14
*** shaohe_feng has quit IRC03:19
*** shaohe_feng has joined #openstack-nova03:20
*** dimtruck is now known as zz_dimtruck03:26
*** zz_dimtruck is now known as dimtruck03:26
*** Nakato has quit IRC03:27
*** shaohe_feng has quit IRC03:29
*** devkulkarni1 has joined #openstack-nova03:29
*** shaohe_feng has joined #openstack-nova03:30
*** Nakato has joined #openstack-nova03:32
*** dimtruck is now known as zz_dimtruck03:35
*** shaohe_feng has quit IRC03:39
*** shaohe_feng has joined #openstack-nova03:40
*** code-R has quit IRC03:41
*** code-R has joined #openstack-nova03:41
*** annp has joined #openstack-nova03:42
*** nkrinner_afk is now known as nkrinner03:47
*** gjayavelu has joined #openstack-nova03:48
*** ayoung has quit IRC03:48
*** shaohe_feng has quit IRC03:49
*** shaohe_feng has joined #openstack-nova03:50
*** sdake has joined #openstack-nova03:50
*** kevinz has joined #openstack-nova03:51
*** devkulkarni1 has quit IRC03:53
*** devkulkarni1 has joined #openstack-nova03:53
*** sdake_ has joined #openstack-nova03:54
*** markvoelker has quit IRC03:55
*** sdake has quit IRC03:58
*** thorst has joined #openstack-nova03:59
*** shaohe_feng has quit IRC04:00
*** shaohe_feng has joined #openstack-nova04:01
*** sdake_ has quit IRC04:05
*** thorst has quit IRC04:06
*** shaohe_feng has quit IRC04:10
*** links has joined #openstack-nova04:10
*** shaohe_feng has joined #openstack-nova04:11
*** Sukhdev has joined #openstack-nova04:13
*** ratailor has joined #openstack-nova04:15
*** jdillaman has quit IRC04:18
*** amotoki has joined #openstack-nova04:20
*** shaohe_feng has quit IRC04:20
*** shaohe_feng has joined #openstack-nova04:21
*** kaisers_ has joined #openstack-nova04:23
*** kaisers_ has quit IRC04:23
*** unicell has joined #openstack-nova04:24
*** devkulkarni1 has quit IRC04:27
*** unicell has quit IRC04:28
*** unicell has joined #openstack-nova04:28
*** shaohe_feng has quit IRC04:30
*** shaohe_feng has joined #openstack-nova04:31
*** markvoelker has joined #openstack-nova04:33
*** flwang has quit IRC04:35
*** markvoelker has quit IRC04:39
*** shaohe_feng has quit IRC04:41
*** shaohe_feng has joined #openstack-nova04:41
*** dikonoor has joined #openstack-nova04:45
*** TravT has joined #openstack-nova04:49
*** shaohe_feng has quit IRC04:51
*** shaohe_feng has joined #openstack-nova04:52
*** TravT has quit IRC04:58
*** shaohe_feng has quit IRC05:01
*** rfolco has joined #openstack-nova05:02
*** shaohe_feng has joined #openstack-nova05:02
*** TravT has joined #openstack-nova05:02
*** jaosorior has joined #openstack-nova05:03
openstackgerritAnush Krishnamurthy proposed openstack/nova: Improve help text of image_file_url  https://review.openstack.org/33249205:03
*** haleyb has quit IRC05:03
*** thorst has joined #openstack-nova05:04
*** yamahata has joined #openstack-nova05:06
*** rfolco has quit IRC05:06
*** ihrachys has joined #openstack-nova05:08
openstackgerritAnush Krishnamurthy proposed openstack/nova: Update tags for Image file url from filesystems config option  https://review.openstack.org/35037905:09
*** thorst has quit IRC05:11
*** gcb has quit IRC05:11
*** shaohe_feng has quit IRC05:11
*** gcb has joined #openstack-nova05:12
*** TravT has quit IRC05:12
*** pbandark has joined #openstack-nova05:12
*** shaohe_feng has joined #openstack-nova05:14
*** irenab has joined #openstack-nova05:16
*** ihrachys has quit IRC05:17
*** yamahata has quit IRC05:19
*** shaohe_feng has quit IRC05:22
*** shaohe_feng has joined #openstack-nova05:24
*** oanson has joined #openstack-nova05:26
*** armax has quit IRC05:28
*** markvoelker has joined #openstack-nova05:28
*** diga has joined #openstack-nova05:28
*** liverpooler has joined #openstack-nova05:28
*** rcernin has joined #openstack-nova05:28
*** unicell has quit IRC05:29
*** unicell has joined #openstack-nova05:30
*** liverpooler has quit IRC05:30
*** liverpooler has joined #openstack-nova05:31
*** rtheis has quit IRC05:32
*** shaohe_feng has quit IRC05:32
*** rtheis has joined #openstack-nova05:32
*** markvoelker has quit IRC05:33
*** shaohe_feng has joined #openstack-nova05:33
*** ecelik has joined #openstack-nova05:41
*** rama_y has joined #openstack-nova05:42
*** shaohe_feng has quit IRC05:42
*** shaohe_feng has joined #openstack-nova05:43
*** chandankumar has joined #openstack-nova05:43
*** jdillaman has joined #openstack-nova05:44
*** fragatina has joined #openstack-nova05:45
*** ihrachys has joined #openstack-nova05:46
*** hshiina has joined #openstack-nova05:49
*** fragatina has quit IRC05:51
*** fragatina has joined #openstack-nova05:52
*** shaohe_feng has quit IRC05:52
*** shaohe_feng has joined #openstack-nova05:53
*** Sukhdev has quit IRC05:55
openstackgerritmelanie witt proposed openstack/nova: Make InstanceGroup object favor the API database  https://review.openstack.org/33099505:56
openstackgerritmelanie witt proposed openstack/nova: Add online migration to move instance groups to API database  https://review.openstack.org/33309805:56
*** haleyb has joined #openstack-nova05:56
*** ihrachys has quit IRC05:58
openstackgerritChangBo Guo(gcb) proposed openstack/os-vif: Trivial: clean up oslo-incubator related stuff  https://review.openstack.org/35038605:59
*** rama_y has quit IRC06:00
*** ekuris_ has joined #openstack-nova06:02
*** shaohe_feng has quit IRC06:03
*** shaohe_feng has joined #openstack-nova06:03
*** mjura has joined #openstack-nova06:04
*** sdake has joined #openstack-nova06:07
*** sdake has quit IRC06:07
*** sdake has joined #openstack-nova06:07
*** tiendc has joined #openstack-nova06:07
*** thorst has joined #openstack-nova06:09
*** code-R has quit IRC06:11
*** code-R has joined #openstack-nova06:12
*** mjura has quit IRC06:12
*** mjura has joined #openstack-nova06:13
*** shaohe_feng has quit IRC06:13
*** karthiks|afk is now known as karthiks06:13
*** shaohe_feng has joined #openstack-nova06:13
*** thorst has quit IRC06:16
*** sudipto has joined #openstack-nova06:17
*** TravT has joined #openstack-nova06:17
*** rfolco has joined #openstack-nova06:18
*** lpetrut has joined #openstack-nova06:19
*** markvoelker has joined #openstack-nova06:22
*** sdake has quit IRC06:23
*** shaohe_feng has quit IRC06:23
*** rfolco has quit IRC06:24
*** shaohe_feng has joined #openstack-nova06:24
*** diana_clarke has joined #openstack-nova06:26
*** markvoelker has quit IRC06:26
*** TravT has quit IRC06:27
*** TravT has joined #openstack-nova06:31
*** code-R has quit IRC06:32
*** code-R has joined #openstack-nova06:32
*** diana_clarke has quit IRC06:33
*** shaohe_feng has quit IRC06:33
*** shaohe_feng has joined #openstack-nova06:34
openstackgerritGhanshyam Mann proposed openstack/nova: Strict ImageRef validation to UUID only  https://review.openstack.org/33880206:35
openstackgerritGhanshyam Mann proposed openstack/nova: Strict ImageRef validation to UUID only  https://review.openstack.org/33880206:38
*** ddeja|away is now known as ddeja06:39
*** TravT has quit IRC06:40
*** belmoreira has joined #openstack-nova06:42
*** shaohe_feng has quit IRC06:44
*** shaohe_feng has joined #openstack-nova06:44
*** tesseract- has joined #openstack-nova06:45
openstackgerritGhanshyam Mann proposed openstack/nova: Strict ImageRef validation to UUID only  https://review.openstack.org/33880206:47
*** moshele has joined #openstack-nova06:47
*** shaohe_feng has quit IRC06:54
*** sridharg has joined #openstack-nova06:54
*** shaohe_feng has joined #openstack-nova06:55
*** thorst has joined #openstack-nova06:55
*** annp has quit IRC06:59
*** tangchen has quit IRC07:00
*** tangchen_ has quit IRC07:00
*** ddeja has left #openstack-nova07:01
*** thorst has quit IRC07:03
*** shaohe_feng has quit IRC07:04
*** shaohe_feng has joined #openstack-nova07:05
*** ihrachys has joined #openstack-nova07:14
*** shaohe_feng has quit IRC07:14
*** jaosorior is now known as jaosorior_brb07:15
*** shaohe_feng has joined #openstack-nova07:15
*** markvoelker has joined #openstack-nova07:16
*** itoader has joined #openstack-nova07:17
openstackgerritPaul Carlton proposed openstack/nova-specs: Use libvirt storage pools  https://review.openstack.org/31050507:18
*** code-R has quit IRC07:20
*** pcaruana has joined #openstack-nova07:21
*** markvoelker has quit IRC07:21
*** cfriesen_ has quit IRC07:21
openstackgerritPaul Carlton proposed openstack/nova-specs: Migrate libvirt volumes  https://review.openstack.org/31053807:22
*** links has quit IRC07:23
*** dikonoor has quit IRC07:23
*** jlanoux has joined #openstack-nova07:24
*** ratailor has quit IRC07:24
*** jlanoux has quit IRC07:24
*** shaohe_feng has quit IRC07:25
*** ratailor has joined #openstack-nova07:25
*** paul-carlton2 has joined #openstack-nova07:25
tdurakovhi folks07:25
*** shaohe_feng has joined #openstack-nova07:26
*** jlanoux has joined #openstack-nova07:26
tdurakovwhile checking logs for ci job, found that there are lots of warnings from privsep07:26
tdurakovis it something know07:26
tdurakovhttp://logs.openstack.org/66/329466/4/check/gate-tempest-dsvm-multinode-live-migration/586c6be/logs/subnode-2/screen-n-cpu.txt.gz?level=WARNING07:26
*** code-R has joined #openstack-nova07:28
*** ekarlso has quit IRC07:29
*** tiendc has quit IRC07:30
*** TravT has joined #openstack-nova07:30
*** shaohe_feng has quit IRC07:35
*** shaohe_feng has joined #openstack-nova07:36
*** lpetrut has quit IRC07:36
*** links has joined #openstack-nova07:36
*** ekarlso has joined #openstack-nova07:37
*** claudiub has joined #openstack-nova07:37
*** dikonoor has joined #openstack-nova07:37
*** TravT has quit IRC07:39
*** Oku_OS-away is now known as Oku_OS07:41
*** TravT has joined #openstack-nova07:44
*** shaohe_feng has quit IRC07:45
*** shaohe_feng has joined #openstack-nova07:46
*** lpetrut has joined #openstack-nova07:46
*** danpb has joined #openstack-nova07:51
*** itoader has quit IRC07:53
*** TravT has quit IRC07:53
*** markus_z has joined #openstack-nova07:54
*** shaohe_feng has quit IRC07:55
openstackgerritTakashi NATSUME proposed openstack/nova: Fix an error in archiving 'migrations' table  https://review.openstack.org/32368407:56
*** shaohe_feng has joined #openstack-nova07:56
*** daemontool has joined #openstack-nova07:57
*** TravT has joined #openstack-nova07:58
*** jistr is now known as jistr|training07:59
*** zzzeek has quit IRC08:00
*** amotoki has quit IRC08:00
*** thorst has joined #openstack-nova08:01
*** zzzeek has joined #openstack-nova08:02
*** takashin has left #openstack-nova08:03
*** paul-carlton2 has quit IRC08:03
*** amotoki has joined #openstack-nova08:04
*** lpetrut has quit IRC08:04
*** amotoki has quit IRC08:05
*** shaohe_feng has quit IRC08:06
*** shaohe_feng has joined #openstack-nova08:07
*** coolsvap is now known as coolsvap_08:08
*** thorst has quit IRC08:08
openstackgerritDavanum Srinivas (dims) proposed openstack/nova: [WIP] Testing latest u-c  https://review.openstack.org/30334508:10
*** markvoelker has joined #openstack-nova08:10
*** jaosorior_brb has quit IRC08:12
*** jaosorior_brb has joined #openstack-nova08:13
*** gjayavelu has quit IRC08:14
*** gjayavelu has joined #openstack-nova08:14
*** TravT has quit IRC08:14
*** markvoelker has quit IRC08:14
*** shaohe_feng has quit IRC08:16
*** amotoki has joined #openstack-nova08:17
*** shaohe_feng has joined #openstack-nova08:17
*** amotoki_ has joined #openstack-nova08:18
*** gjayavelu has quit IRC08:19
*** TravT has joined #openstack-nova08:19
*** amotoki has quit IRC08:21
*** code-R has quit IRC08:22
*** yassine has joined #openstack-nova08:23
*** jaosorior_brb is now known as jaosorior08:23
*** yassine is now known as yassou08:23
*** shaohe_feng has quit IRC08:26
*** baoli has quit IRC08:27
*** paul-carlton2 has joined #openstack-nova08:28
*** shaohe_feng has joined #openstack-nova08:29
*** amotoki_ has quit IRC08:29
*** paul-carlton has joined #openstack-nova08:30
johnthetubaguytdurakov: wow, thats a lot. I thought there was a patch to fix that in priv sep, but I could be wrong, maybe we need another patch too08:30
danpbjohnthetubaguy: a patch to fix what ?08:31
johnthetubaguytdurakov: here it is: https://review.openstack.org/#/c/350415/08:32
johnthetubaguydanpb: all the privsep logs being warning, rather than the correct log level, or debug, or something08:32
johnthetubaguytdurakov: it follows on from this one: https://review.openstack.org/#/c/339275/08:33
*** shaohe_feng has quit IRC08:36
*** TravT has quit IRC08:37
openstackgerritBéla Vancsics proposed openstack/nova: Reduce code duplication and complexity  https://review.openstack.org/34218708:37
*** shaohe_feng has joined #openstack-nova08:37
*** gb21 has joined #openstack-nova08:40
*** TravT has joined #openstack-nova08:42
*** RuiChen has joined #openstack-nova08:43
gibijohnthetubaguy: hi! I've checked the possibility of setUpClass usage for the notification testing.08:43
gibijohnthetubaguy: the biggest problem with it that nova base TestCase class build on setUp08:44
*** amotoki has joined #openstack-nova08:45
gibijohnthetubaguy: if we want to have all the environment set up at setUpClass time then we have to move lot of set up from setUp to setUpClass in TestCase as well08:45
gibijohnthetubaguy: but that will affect all the test in nova08:45
gibijohnthetubaguy: I left some details in the review08:46
johnthetubaguygibi: hmm, thats odd, what did you need access to?08:46
*** shaohe_feng has quit IRC08:47
*** shaohe_feng has joined #openstack-nova08:47
johnthetubaguygibi: do you have a link to the review handy?08:48
*** gb21 has quit IRC08:48
gibijohnthetubaguy: I need all the set up to be able to boot an instance08:48
gibijohnthetubaguy: https://review.openstack.org/#/c/342944/08:48
*** priteau has joined #openstack-nova08:48
*** gb21 has joined #openstack-nova08:48
gibijohnthetubaguy: e.g starting up nova services08:49
openstackgerritAndrey Volkov proposed openstack/python-novaclient: Add support for microversion 2.33  https://review.openstack.org/34954708:49
johnthetubaguygibi: oops, true08:49
gibijohnthetubaguy: also fixtures does not play well with setUpClass08:49
*** jistr|training is now known as jistr08:49
johnthetubaguyyeah, its a little bit of an anti-pattern I guess, so that makes sense08:50
johnthetubaguygibi: thanks for looking, it probably means what you have is the best option08:50
gibijohnthetubaguy: If we want to spare some (e.g a lot of :)) instance boots in the test then we can do what I propose quite easly but doing that via setUpClass needs a lot more effort08:51
johnthetubaguygibi: +108:52
gibijohnthetubaguy: thanks for the review08:52
*** amotoki has quit IRC08:53
johnthetubaguygibi: sorry for the wild goose chase there, it was educational at least08:53
gibijohnthetubaguy: no problem, I also learned a good lesson during the investigation08:53
*** eliqiao has quit IRC08:54
gibijohnthetubaguy: there are some other improvements proposed in the notification area, if you have some review bandwidth08:54
*** ihrachys has quit IRC08:54
gibijohnthetubaguy: https://review.openstack.org/#/c/341422/ reservation_id handling cleanup in the sample test08:54
gibijohnthetubaguy: https://review.openstack.org/#/c/344170/ doc generation improvement08:54
*** code-R has joined #openstack-nova08:54
*** eliqiao has joined #openstack-nova08:55
kaisersmdbooth: Hi! Regarding your questions i'm trying to answer in order:08:55
openstackgerritAndrey Kurilin proposed openstack/python-novaclient: Fix python35 job failures  https://review.openstack.org/34415108:55
*** TravT has quit IRC08:56
openstackgerritAndrey Kurilin proposed openstack/python-novaclient: Fix python35 job failures  https://review.openstack.org/34415108:57
mdboothkaisers: Thanks. What TZ are you in, btw?08:57
*** shaohe_feng has quit IRC08:57
*** shaohe_feng has joined #openstack-nova08:58
tdurakovjohnthetubaguy: I see thanks, so are folks going to merge these loging improvements into privsep?08:58
mdboothkaisers: Regarding the revert, I've proposed this: https://review.openstack.org/#/c/349541/08:58
*** code-R has quit IRC08:59
johnthetubaguytdurakov: hopefully, looks like angus uploaded the fix he wanted doing, so that should mean we are closer now08:59
mdboothkaisers: Quobyte CI doesn't seem to have voted on it yet, though, which is annoying :/08:59
*** hshiina has quit IRC09:00
*** TravT has joined #openstack-nova09:00
tdurakovjohnthetubaguy: hope so:)09:00
openstackgerritMatthew Booth proposed openstack/nova: libvirt: Use fake_imagebackend in functional  https://review.openstack.org/34555309:01
openstackgerritMatthew Booth proposed openstack/nova: libvirt: Remove some unnecessary mocking in test_driver  https://review.openstack.org/33397909:01
openstackgerritMatthew Booth proposed openstack/nova: libvirt: Create console.log consistently  https://review.openstack.org/34954109:01
openstackgerritMatthew Booth proposed openstack/nova: libvirt: Improve mocking of imagebackend disks  https://review.openstack.org/33324209:01
openstackgerritMatthew Booth proposed openstack/nova: libvirt: Pass object to _create_images_and_backing in test  https://review.openstack.org/33324309:01
openstackgerritMatthew Booth proposed openstack/nova: libvirt: Reset can_fallocate in test setUp()  https://review.openstack.org/34416809:01
*** amotoki has joined #openstack-nova09:02
*** rmart04 has joined #openstack-nova09:03
kaisersmdbooth: qemu is running with nova:cinder in our systems, the CI uses ubuntu:ubuntu in a devstack env.09:04
*** markvoelker has joined #openstack-nova09:04
*** gb21 has quit IRC09:05
*** thorst has joined #openstack-nova09:06
mdboothkaisers: Why do nova and cinder require the same write permissions? Is cinder touching the nova instances directory at any point?09:06
openstackgerritOpenStack Proposal Bot proposed openstack/os-vif: Updated from global requirements  https://review.openstack.org/33991909:06
mdboothkaisers: Incidentally, I just pushed https://review.openstack.org/#/c/349541/ to the front of my queue because it was causing me problems locally. If you could run it against your CI that would be great.09:07
kaisersmdbooth: regarding same instance storage: Quobyte is a shared file system, volumes are provided in Nova by mounting the Quobyte Cinder volume on a given compute host. Does that answer your question?09:07
*** shaohe_feng has quit IRC09:07
mdboothkaisers: Where are they mounted?09:08
*** shaohe_feng has joined #openstack-nova09:08
*** roeyc has joined #openstack-nova09:08
kaisersmdbooth: Why console.log is created with root:root ownership despite qemu running as non-root i did not understand, too. I'm happy for ideas on this.09:08
mdboothkaisers: It's created by libvirt, which is running as root.09:09
*** markvoelker has quit IRC09:09
mdboothlibvirt expects to change its ownership when it launches.09:09
mdboothAs danpb mentioned, I suspect this model is broken in other cases, btw.09:09
*** TravT has quit IRC09:10
mdboothI'd want to go audit the nova driver to see everywhere to do stuff as root.09:10
*** gb21 has joined #openstack-nova09:10
mdboothThe Quobyte CI setup just sounds fragile to me. If I were you I'd invest some time in making the change danpb suggests to tell libvirt to leave the quobyte volumes alone.09:11
kaisersdanpb: mriedem: i saw your pointer to the Quobyte CI/driver topic, thanks for that. I opened a ticket to provide a place where to discuss / improve this: https://bugs.launchpad.net/nova/+bug/160929809:11
openstackLaunchpad bug 1609298 in OpenStack Compute (nova) "libvirt should not require dynamic_ownership off for secure Cinder/Quobyte settings" [Undecided,New]09:11
danpbkaisers: why does the CI system turn off dynamic_ownership is the biggest question  - that is certainly the wrong thing to do09:11
danpbkashyap: if quobyte requires that its voliumes are not chown'd, then we  should use a <seclabel> to stop it, never dynamic_ownership09:12
*** annp has joined #openstack-nova09:12
*** annp_ has joined #openstack-nova09:13
danpbmdbooth: i'm looking at your change here https://review.openstack.org/#/c/333243/17/nova/tests/unit/virt/libvirt/test_driver.py09:13
pkholkinmriedem: Hi Matt, could you please look this patch with tests https://review.openstack.org/34746209:13
danpbmdbooth: and struggling to understand how the hell tests currently succeeed !??!09:13
*** thorst has quit IRC09:13
*** annp has left #openstack-nova09:13
*** annp_ has quit IRC09:14
mdboothdanpb: Could well be because it's not really testing anything09:14
danpbmdbooth: afaict, _fetch_isntance_kerneLramdisk will call  utils.get_instance_path which calls    instance.name09:14
danpbwhich will break when given a dict09:14
*** TravT has joined #openstack-nova09:14
mdboothutils != utils09:14
mdboothutils == fake_utils09:14
mdboothiirc09:14
danpboh wait, no,  its mocked out the _fetch_instancE_kernel_ramdisk method entirely09:14
kaisersmdbooth: mriedem: danpb: I just worked through the irc history, just catching up with the lower end now, sorry. I'm in TZ UTC+2.09:14
mdboothdanpb: Or that. We over-mock all over the place :/09:15
kaisersmdbooth: yep, i saw you're new change, will check this09:15
danpband god i hate that  fake_libvirt_Utils even exists09:15
*** EinstCrazy has quit IRC09:15
danpbwe should kill that someday and just have tests mockk out the individual utils methods that are relevant to them09:15
mdboothkaisers: I'd work backwards if I were you :) If you go backwards you'll find where people answered their own previous questions.09:15
openstackgerritAndrey Kurilin proposed openstack/python-novaclient: Fix python35 job failures  https://review.openstack.org/34415109:16
*** EinstCrazy has joined #openstack-nova09:16
mdboothdanpb: Some of these tests are really dumb, basically pro-forma.09:16
mdboothdanpb: Incidentally, chances are I had to make that change because I removed some mocking and hit the actual error.09:17
*** shaohe_feng has quit IRC09:17
*** shaohe_feng has joined #openstack-nova09:18
openstackgerritAndrey Kurilin proposed openstack/python-novaclient: Fix python35 job failures  https://review.openstack.org/34415109:18
kaisersmdbooth: yep, in hindsight probably better. :)09:18
mdboothkaisers: Thanks for poking the CI. Any idea what its response time is like right now?09:20
kaisersmdbooth: ~1h09:20
mdboothkaisers: Awesome, thanks.09:20
*** xinliang has quit IRC09:21
kaisersdanpb: reason for dynamic_ownership=0 is described in https://bugs.launchpad.net/nova/+bug/1609298 , pls ping me if i can provide further info09:22
openstackLaunchpad bug 1609298 in OpenStack Compute (nova) "libvirt should not require dynamic_ownership off for secure Cinder/Quobyte settings" [Undecided,New]09:22
*** null0 has joined #openstack-nova09:23
*** TravT has quit IRC09:23
*** tangchen has joined #openstack-nova09:24
kaisersmdbooth: Regarding where the volumes are mounted: The Nova driver mounts a Quobyte volume containing the Cinder volume files at a configurable mount point in the local file system, when running instances of a Quobyte based Cinder volume.09:25
*** gb21 has quit IRC09:25
kaisersmdbooth: please note that there's a difference between a Quobyte volume, which is a volume in Quobyte, and a Cinder volume which is in essence an image file on a Quobyte volume in our drivers.09:26
*** lpetrut has joined #openstack-nova09:26
*** shaohe_feng has quit IRC09:28
mdboothkaisers: Got it, thanks.09:28
*** shaohe_feng has joined #openstack-nova09:28
mdboothkaisers: I think that answers my main concern. From the original bug report I had (fortunately erroneously) understood that cinder was writing to files owned by Nova.09:29
openstackgerritGleb Stepanov proposed openstack/nova: Skip malformed cookies  https://review.openstack.org/34609009:29
mdboothIIUC it's not doing that. It just needs to be able to write to cinder volumes with common ownership.09:29
kaisersmdbooth: ah, ok. No, cinder (driver) does not touch those.09:30
kaisersmdbooth: Makes me think where i can clarify the env better in the tickets...09:31
*** null0 has quit IRC09:33
kaisersmdbooth: correct (regarding your last question), Nova needs to be able to write to the cinder volume files.09:36
*** shaohe_feng has quit IRC09:38
*** shaohe_feng has joined #openstack-nova09:39
openstackgerritGleb Stepanov proposed openstack/nova: [WIP] Init allocated ports set  https://review.openstack.org/34747909:40
openstackgerritPawel Koniszewski proposed openstack/nova: [libvirt] Remove live_migration_flag & block_migration_flag  https://review.openstack.org/33486009:41
openstackgerritPawel Koniszewski proposed openstack/nova: Change default value of live_migration_tunnelled to False  https://review.openstack.org/35048009:41
lpetrutHi, any chance anyone could take a look over this patch, updating the HyperV live migrate data object? https://review.openstack.org/#/c/347435/3 We need this in order to fix our issues regarding shared storage. This is part of a series, fixing our issues on this matter.09:44
*** xinliang has joined #openstack-nova09:44
*** Qiming has quit IRC09:46
*** roeyc has quit IRC09:47
*** shaohe_feng has quit IRC09:48
*** code-R has joined #openstack-nova09:49
*** shaohe_feng has joined #openstack-nova09:49
*** Qiming has joined #openstack-nova09:50
*** edand has joined #openstack-nova09:51
*** TravT has joined #openstack-nova09:51
*** code-R has quit IRC09:53
tdurakovdanpb: hey, are you around?09:55
*** TravT has quit IRC09:56
*** yamahata has joined #openstack-nova09:56
tdurakovdanpb, pkoniszewski https://review.openstack.org/#/c/350480/1 - do you think it's worth to log warning for operator that there is no tunneled flag any more, or it's enough to have release notes?09:56
pkoniszewskitdurakov: i don't think that warning is required there, reno might be09:56
danpbyeah certainly no warning is appropriate09:57
openstackgerritxiexs proposed openstack/nova-specs: Make API errors conform to API-WG schema  https://review.openstack.org/30767509:57
tdurakovpkoniszewski, danpb kk, than only reno is needed09:57
*** Jaison has joined #openstack-nova09:58
*** markvoelker has joined #openstack-nova09:58
*** shaohe_feng has quit IRC09:58
*** links has quit IRC09:59
*** shaohe_feng has joined #openstack-nova09:59
openstackgerritPawel Koniszewski proposed openstack/nova: Change default value of live_migration_tunnelled to False  https://review.openstack.org/35048010:00
*** fwiesel has joined #openstack-nova10:00
openstackgerritPawel Koniszewski proposed openstack/nova: [libvirt] Remove live_migration_flag & block_migration_flag  https://review.openstack.org/33486010:01
*** chenfei has quit IRC10:02
*** chenfei has joined #openstack-nova10:02
*** fwiesel is now known as fabianw10:03
*** markvoelker has quit IRC10:04
*** TravT has joined #openstack-nova10:05
fabianwHi, is anyone knowledgeable about the "nova.virt.vmwareapi.vm_util.get_vmdk_info" here? I'm curious, how it is supposed to work. As I understand it, it is trying to get the root-disk based on the VMDK name10:07
fabianwWhich a) fails for me right now, and b) I am not sure how it is supposed to work, since the vcenter renames the files practically at will10:08
*** shaohe_feng has quit IRC10:09
*** fragatina has quit IRC10:09
*** shaohe_feng has joined #openstack-nova10:09
openstackgerritDavanum Srinivas (dims) proposed openstack/nova: [WIP] Testing latest u-c  https://review.openstack.org/30334510:10
openstackgerritAlvaro Lopez Garcia proposed openstack/nova: Ensure that periodic reclaim cleans DB deleted instances  https://review.openstack.org/32325010:10
*** yamahata has quit IRC10:11
*** thorst has joined #openstack-nova10:11
*** ManishD has joined #openstack-nova10:12
*** Wenzhi has quit IRC10:12
*** Qiming has quit IRC10:12
*** chenfei has quit IRC10:12
*** bhagyashris has joined #openstack-nova10:12
*** Wenzhi has joined #openstack-nova10:13
*** kevinz has quit IRC10:13
*** TravT has quit IRC10:14
ManishDHI, I am running into one case:-- 1. Attach FC and ICSI volume to single instance (same instance) -- Successful  2. Detach ICSI volume -- Successful 3. Reattach ICSI volume -- Error :- 24 - iSCSI login failed due to authorization failure10:15
*** serverascode has quit IRC10:15
ManishDplease can someone help10:15
*** bhagyashri has quit IRC10:16
*** Dinesh_Bhor has quit IRC10:16
*** serverascode has joined #openstack-nova10:16
*** Dinesh_Bhor has joined #openstack-nova10:17
*** Qiming has joined #openstack-nova10:17
*** thorst has quit IRC10:18
*** gb21 has joined #openstack-nova10:18
*** chenfei has joined #openstack-nova10:18
*** TravT has joined #openstack-nova10:18
paul-carlton2danpb, looking at your comments on storage pools spec and thinking I may have missed something with live migration. I think the problem is it is not nova but libvirt/qemu that creates the instance on the target by copying the domain.xml from the source, nova just creates items like underlay files for qcow2 disks etc10:19
*** shaohe_feng has quit IRC10:19
paul-carlton2danpb, I was thinking live migration could be used to create a new instance on the target with new disk layout10:20
*** shaohe_feng has joined #openstack-nova10:20
tdurakovdanpb could you please take a look, when have time https://review.openstack.org/#/c/348284/10:20
*** rfolco has joined #openstack-nova10:20
paul-carlton2danpb, I can investigate this further to see if there is anyway to get the target instance created with new disk locations10:21
tdurakovdanpb: oh, already, will update it soon)10:21
danpbpaul-carlton2: if you have NFS for the instances directory there's no point at which you create new disk on the target10:21
danpbpaul-carlton2:  as the disks already exist10:21
danpbif you don't have shared storage, then nova could update the libvirt xml to point to the new location10:22
paul-carlton2danpb, agreed, but in the non shared case10:22
danpbwe have to assume some people will have NFS though10:22
paul-carlton2danpb, that is what I was hoping10:22
danpbtdurakov: err i already commented on it and its not been addressed10:23
paul-carlton2danpb, if the disks are shared via NFS live migration will detect this and not copy the disks which means live migration won't ffer an opportunity to move the disk?10:24
danpbpaul-carlton2: exactly10:24
paul-carlton2danbp but we could still fix up the instance xml on the target and copy the disk to the new location before letting the target instance be created?10:25
*** rfolco has quit IRC10:25
danpbpaul-carlton2: the whole point of NFS is that you're not copying disk images10:26
*** hoangcx has quit IRC10:27
danpband iirc nova will refuse to let you request block migration if using nfs, as that would ordinarily cause corruption as the target overwrote the existing image10:28
paul-carlton2danpb, agreed, but as a special case if the live migration detects it is migrating and old style instance it could use 'mv' to rename the file, the source would continue to access it via the existing open and the target would open it in the new location because we changed its xml to say find this disk here?10:29
*** shaohe_feng has quit IRC10:29
*** amotoki has quit IRC10:30
*** shaohe_feng has joined #openstack-nova10:30
*** rodrigods has quit IRC10:30
*** rodrigods has joined #openstack-nova10:30
*** TravT has quit IRC10:32
danpbpaul-carlton2: you'd need logic to move it back to the orginal location on  migration failure too10:34
paul-carlton2danpb, agreed, I'll update spec and will try this out to verify there are no hidden issues10:35
danpbpaul-carlton2: also what happens if you have a vm running on new nova, and live migration wants to move it to an old nova with old layout10:35
danpbafaik, we have to cope with live migration in both directions10:36
danpblikewise if guest state is saved on a new node, and resumed on an old node, we're doomed10:36
paul-carlton2danpb, new to old would not be allowed, I came across this in my implementation of live migration of rescued instances10:36
danpbbecause the old node won't have the disk in the right location for the saved xml state10:37
danpbpaul-carlton2: i don't think we can simply rule out new to old in that way10:37
danpbits a pretty significant restriction on deployments during process of updating their compute nods10:38
paul-carlton2danpb, I got feed back that all compute nodes need to be at same level before live migration would be allowed if I recall correctly, I'll check10:38
*** chenfan has quit IRC10:38
danpbanyway this is all stuff needed in the spec10:38
paul-carlton2danpb, I agreed I had code to detect the version of the target node but was told to remove it10:38
*** amotoki has joined #openstack-nova10:39
*** EinstCrazy has quit IRC10:39
*** EinstCrazy has joined #openstack-nova10:39
danpbpaul-carlton2: wasn't that related to RPC API versions10:39
*** shaohe_feng has quit IRC10:39
paul-carlton2danpb, agreed,  need to update it but first we need to agree on if new to old support is required, I think not10:40
johnthetubaguymikal: did you see the questions for you here: https://review.openstack.org/#/c/33495610:40
*** shaohe_feng has joined #openstack-nova10:40
*** amotoki has quit IRC10:41
johnthetubaguypaul-carlton2: folks do use live-migrate to perform the upgrade of their nodes, so we certainly need old -> new, it seems a shame to break the reverse if we don't have to I guess10:41
paul-carlton2danpb, yep, I was getting RPC API version of target host and making sure it was at the level needed to support new feature with out worrying that other hosts in the clouds might not yet be updated but I was advised to simply check for minimum version of all comoute nodes and prevent live migration of rescused instances till that was at the right level, I wasn't happy but accepted the guidance10:42
danpbnew -> old is pretty critical in the early stages of upgrade process where the number new nods is tiny compared to number of old nodes10:42
*** code-R has joined #openstack-nova10:43
paul-carlton2johnthetubaguy, old to new is alway supported10:43
danpbpaul-carlton2: i think that's a different scenario becasue  migration os rescued instances is a new feature10:43
danpbpaul-carlton2 ie you were not breaking an existing scenario - you were simply not adding a new scenario10:43
johnthetubaguypaul-carlton2: there was a test in the grenade multi-node job that mriedman created as a temporary thing that did a move old to new and new to old, we might be able to add that properly now, but tdurakov might have more details on that10:43
johnthetubaguyah...10:43
johnthetubaguyyeah, we did say, migration of rescued things is a new feature10:44
*** EinstCrazy has quit IRC10:44
johnthetubaguyso for simplicity, you only allow the new feature once all nodes are upgraded10:44
danpbwith the movement of disk image locations, you would be breaking existing valid feature10:44
johnthetubaguyI admit that doesn't help with the upgrade till a cycle later, which does suck a bit10:44
danpbso that would basically imply you would have to add code for the new disk layout in ocata, but not actually use that code until Pxxxx10:45
johnthetubaguywell you use the code once the upgrade is complete, i.e. its tested in the gate right away?10:46
paul-carlton2johnthetubaguy, danpb I think it is reasonable to disallow new to old, surely all compute nodes should be updated within a short period?10:46
johnthetubaguypaul-carlton2: short might be a week I guess?10:46
danpbpaul-carlton2: depends on the scale of your deployment - i could easily imagine it taking weeks or more10:46
paul-carlton2I played with this for live migration of rescued instances and found that I could detect the RPC API level of the specific target node10:47
*** code-R has quit IRC10:47
paul-carlton2and prevent migration to that node.10:47
*** eliqiao has quit IRC10:47
johnthetubaguydanpb I don't get the comment about having to wait till Pxxx, am I missing something?10:47
danpbjohnthetubaguy: if you start using the new disk layout on a new VM on a Ocata node, that'd prevent it being  live migrated to an Newton node10:48
danpbjohnthetubaguy: so if you add support for the new layout in Ocata, but don't actually start launching VMS with the new layout until P10:48
paul-carlton2If the node is up it should be upgraded by deployment tooling within a few hours, if it is down when it restarts nova compute has to be upgrade before it can rejoin10:49
danpbjohnthetubaguy: then you preserve the ability to live migration from Ocata to Newton10:49
*** Jaison has quit IRC10:49
danpbjohnthetubaguy: and also enable migration from Pxxxx to Ocata  (both support new layout)10:49
*** shaohe_feng has quit IRC10:50
*** shaohe_feng has joined #openstack-nova10:50
paul-carlton2I'm reasonably happy with an approach that implements this across two releases but still contend that as mentioned above new to old scenario can be blocked without operational issues10:51
*** eliqiao has joined #openstack-nova10:51
johnthetubaguydanpb: ah, gotcha. so I guess post live-migrate you would get the new disk layout?10:51
*** gongysh_ has quit IRC10:52
johnthetubaguywe could just only start using the new layout post upgrade, and support old a new layouts for a little bit?10:52
*** chenfan has joined #openstack-nova10:52
johnthetubaguyso create new VMs with old layouts until the upgrade completes10:52
*** links has joined #openstack-nova10:53
danpbjohnthetubaguy: yeah, i guess you could do that - we'd need to have some config option to say whether to use the new layout10:53
johnthetubaguywe can detect that, using min_service version10:53
danpboh, i guess we could do that10:54
danpbthis would also solve the save/restore problem10:54
paul-carlton2danpb, johnthetubaguy, my point is we can prevent new to old migrations, most of the cloud should be upgraded within hours, what scenario am I missing?10:54
danpbpaul-carlton2: no, we absolutely cannot assumed every compute node is upgraded with hours10:55
*** sdague has joined #openstack-nova10:55
johnthetubaguyyeah, within hours isn't a reasonable assumption, sadly10:55
danpbwe have to assume it can be spread over weeeks10:55
*** zerda2 has quit IRC10:55
johnthetubaguylive-migrates put a large load on the system, so you need to stagger them10:55
paul-carlton2danpb, johnthetubaguy , ok10:56
danpbit can also kill VM performance while going on10:56
danpbso some places may need to get end user buyin to schedule migration at a specific convenient time10:56
paul-carlton2danpb, surely we can have saved instances that remain saved for months, thus preventing us turning on new format?10:56
danpbpaul-carlton2: you would still need code to re-write the XML before restoring them10:57
*** dikonoor has quit IRC10:57
danpbit just avoids possibility of new VMs with new layout being restored on old nodes, if we wait for min_service cap10:57
paul-carlton2danpb agreed, that is not a problem10:57
paul-carlton2danpb, you mean a saved instance might be cold migrated from new to old?10:58
paul-carlton2danpb does cold migration actually copy the qemu save file?10:59
*** shaohe_feng has quit IRC11:00
danpbpaul-carlton2: i don't think it copies the qemu save file, but the save file could be on nfs11:00
paul-carlton2danpb, If we need to support live and cold migration from new to old I need to think about this some more.11:01
*** shaohe_feng has joined #openstack-nova11:01
gcbjohnthetubaguy,  I just saw you approved https://review.openstack.org/325761 .  In fact, there is a  series of checking config options in tests https://review.openstack.org/#/q/topic:bug/1517839+status:open+project:openstack/nova11:02
paul-carlton2danpb, I think it is doable, possibly with a two step upgrade, do we support upgrade from N to N+2, ie Mitka to Ocata?11:02
gcbthey are in a stack, and  enforce type check in the last commit , So would be nic if you can review others when you'r free. :-)11:04
*** gszasz has joined #openstack-nova11:04
danpbpaul-carlton2: we only ever allow N to N+111:04
*** kolesovdv has joined #openstack-nova11:04
paul-carlton2danpb, good then we could implement code in Ocata to support all this but not use the new format unless the operator sets a flag to say use it in which case they are saying new to old is not an issue because they have all nodes on Ocata?11:06
johnthetubaguygcb: ah, cool, I didn't got digging very far, just trying to kill off changes waiting for another +2, will put those on my list of later11:07
paul-carlton2danpb, in P we'd remove the flag and the new format would become the default, because Ocata code would know how to support new format to old11:08
danpbpaul-carlton2:  as johnthetubaguy said, we don't need a flag11:08
gcbjohnthetubaguy,  thanks11:08
danpbpaul-carlton2: we can look to see if the min  api version is older than Ocata, and if it is, then assume we have to continue using old layout11:08
paul-carlton2danpb, you mean check min version on every instance create, rebuild etc and use new format if it is Ocata?11:09
danpbyes11:10
*** shaohe_feng has quit IRC11:10
*** sameo has joined #openstack-nova11:10
danpbnb, this would imply 2 live migrations of each VM before they use the new layout11:10
danpbafter first live migration from Newton to Ocata  they'll still be using old layout11:11
*** shaohe_feng has joined #openstack-nova11:11
danpbafter second live migration from Ocata to Ocata, they'll be using new layout (assuming all Newton nodes have been killed at that point)11:11
johnthetubaguyyeah, that sounds right11:11
*** roeyc has joined #openstack-nova11:12
johnthetubaguyI don't think we can assume people use live-migrate to do upgrade, so there will be older instances around using the old layout, basically for ever11:12
paul-carlton2danpb, johnthetubaguy surely if you live migrate an instance after min version reaches Ocata then it will upgrader to new format?11:12
*** roeyc has quit IRC11:13
johnthetubaguyyeah, thats when you to Ocata to Ocata though, or Ocata to P11:13
johnthetubaguyoh, hang on, yeah11:13
paul-carlton2danpb, johnthetubaguy, the old format needs to be supported till every instances has been rebuit, hard rebooted or live migrated11:14
johnthetubaguyit would be Newton to Newton that starting changing things, if we merge it in Newton11:14
johnthetubaguypaul-carlton2: right, which is basically until the end of time11:14
paul-carlton2johnthetubaguy, we won't get this into Newton now11:14
*** diga has quit IRC11:14
paul-carlton2johnthetubaguy, my proposal was we enforce an end date, maybe Q but maybe that won't fly?11:16
*** thorst has joined #openstack-nova11:16
danpbNewton is out of the question - we're not getting any of this into that11:17
*** gb21 has quit IRC11:17
paul-carlton2danpb, johnthetubaguy I was thinking we'd provide a utility to identify instances still using the legacy disk layout so that operators can live migrate their instances before moving to Q or whatever the cutoff is11:18
danpbpaul-carlton2: if we can make live migration do the right thing with conversions, then we can be fairly aggresive with removing code for old layout - Q would be reasonable11:18
paul-carlton2danpb, agreed11:18
danpbpaul-carlton2: if on the other hand we require a cold reboot, then we'd have to keep the old code around for many many many releases11:18
paul-carlton2danpb, agreed, we can't enforce instance restart, I know an MOD Tandem system that they did not reboot for 14 years!11:20
*** abalutoiu has joined #openstack-nova11:20
*** shaohe_feng has quit IRC11:20
*** lucasagomes is now known as lucas-hungry11:21
*** rfolco has joined #openstack-nova11:22
paul-carlton2danpb, johnthetubaguy will update spec and do some testing to make we can implement it so live migration does old to new conversion for all scenarios11:22
johnthetubaguyyeah, I would just make it start doing the conversions only once min_version reaches the point we add the new layout support, I guess?11:23
*** gb21 has joined #openstack-nova11:23
*** shaohe_feng has joined #openstack-nova11:24
*** roeyc has joined #openstack-nova11:25
*** thorst has quit IRC11:25
*** rfolco has quit IRC11:26
*** bvanhav has joined #openstack-nova11:27
alogapaul-carlton2, johnthetubaguy, danpb: a couple of commends regarding the previous discussion, from an ops point of view :)11:28
*** ManishD has quit IRC11:28
alogapaul-carlton2, johnthetubaguy, danpb: you cannot assume that the upgrade from N to N+1 is done in hours, I would say that weeks is a much more reasonable timeframe11:28
johnthetubaguyaloga: agreed, thats certainly what I am seeing as an operator11:29
johnthetubaguyaloga: well, if you include live-migrate11:29
alogapaul-carlton2, johnthetubaguy, danpb: and the same goes for the migration/reboot/whatever11:29
johnthetubaguywithout live-migrate, I think we are doing around 1k nodes per hour ish, last time I checked11:30
alogaI mean, sometines you cannot simply restart a VM, as the workload can be lost (for instance MPI computing requiring coordination between nodes)11:30
johnthetubaguyaloga: so we can never restart a VM11:31
openstackgerritPavel Kholkin proposed openstack/nova: removed db_exc.DBDuplicateEntry in bw_usage_update  https://review.openstack.org/25080711:31
*** shaohe_feng has quit IRC11:31
alogaor long-running VMs11:31
*** edand has quit IRC11:31
*** shaohe_feng has joined #openstack-nova11:32
alogajohnthetubaguy: well, we're not a commercial site, so we can reboot things without losing money11:32
johnthetubaguyyeah, it varies11:32
alogajohnthetubaguy: but sometimes, critial computations cannot be stopped in several days11:32
johnthetubaguyas a service provider, its basically never, unless the sky is falling down11:32
alogajohnthetubaguy: indeed11:32
alogaI do not know if stablishing a timeframe for the migration could be reasonable for other folks11:34
alogafor us it would be OK, assuming that it is large enough :)11:34
* johnthetubaguy goes for lunch11:36
alogajohnthetubaguy: enjoy11:36
*** code-R has joined #openstack-nova11:37
*** fabianw has quit IRC11:37
*** sudipto has quit IRC11:40
*** thorst has joined #openstack-nova11:41
*** shaohe_feng has quit IRC11:41
*** shaohe_feng has joined #openstack-nova11:42
*** code-R has quit IRC11:42
tdurakovmdbooth: hey, are you around?11:43
mdboothtdurakov: Yup.11:44
andreykurilin johnthetubaguy: hi! Can you put your +2 for https://review.openstack.org/#/c/344151/ ? :) it is about py35 support in novaclient11:44
*** smatzek has joined #openstack-nova11:45
*** hshiina has joined #openstack-nova11:45
jrolljaypipes: dansmith: given the networking feature is bumping the minimum anyway, are we good on https://review.openstack.org/#/c/345082/ then?11:45
tdurakovmdbooth: got question on this(want to be sure I understand this change properly) https://review.openstack.org/#/c/344168/6 why you change order of calls only for EncryptedLvmTestCase, while there are several classes that also inherits from _ImageTestCase ?11:45
*** coolsvap_ is now known as coolsvap11:45
mdboothtdurakov: TBH, this will have been simply a case of me attempting to boil slightly less ocean.11:46
mdboothI'll have spotted the test failure, and fixed it locally.11:46
mdboothI don't think I spent a week deliberating on this one.11:46
tdurakovmdbooth: ok, so the bug you trying to fix affects only EncryptedLvmTestCase, right?11:47
danpbtdurakov: the other test clases already have the call in the correct order11:47
danpbEncryptedLvmTestCase was the only one that was in a different order11:48
tdurakovdanpb: ok11:48
openstackgerritChangBo Guo(gcb) proposed openstack/nova: Set enforce_type=True in method flags  https://review.openstack.org/32574311:48
*** shaohe_feng has quit IRC11:51
*** shaohe_feng has joined #openstack-nova11:52
*** markvoelker has joined #openstack-nova11:52
openstackgerritAndrey Volkov proposed openstack/python-novaclient: Add support for microversion 2.33  https://review.openstack.org/34954711:53
*** shaohe_feng has quit IRC12:01
*** shaohe_feng has joined #openstack-nova12:02
*** phuongnh has quit IRC12:03
*** TravT has joined #openstack-nova12:05
mdboothjohnthetubaguy: https://review.openstack.org/#/c/334956/ So my poor gerrit-fu may mean that you missed my response to jaypipes, because I put it in the wrong place :)12:06
mdboothIt's a top-level comment12:06
*** ducttape_ has joined #openstack-nova12:06
mdboothI have no idea why I did that, btw.12:07
* mdbooth cuts/pastes to the right place.12:07
*** shaohe_feng has quit IRC12:12
*** shaohe_feng has joined #openstack-nova12:13
*** xyang1 has joined #openstack-nova12:16
*** gb21 has quit IRC12:19
*** edmondsw has joined #openstack-nova12:19
kaisersdanpb: Regarding https://bugs.launchpad.net/nova/+bug/1609298 : what type & mode would that seclabel require? I'm not familiar with seclabels but i'll create a change for inserting it as proposed.12:20
openstackLaunchpad bug 1609298 in OpenStack Compute (nova) "libvirt should not require dynamic_ownership off for secure Cinder/Quobyte settings" [Undecided,New]12:20
*** shaohe_feng has quit IRC12:22
*** takedakn has joined #openstack-nova12:22
*** shaohe_feng has joined #openstack-nova12:22
*** TravT has quit IRC12:23
openstackgerritHironori Shiina proposed openstack/nova: ironic: Cleanup instance information when spawn fails  https://review.openstack.org/34125312:23
*** jamesdenton has joined #openstack-nova12:25
*** fwiesel has joined #openstack-nova12:25
danpbkaisers: iirc, you'll need  <seclabel type="dac" relabel="no">  inside the <source> element for the <disk>12:25
*** fwiesel has quit IRC12:25
*** rfolco has joined #openstack-nova12:25
*** baoli has joined #openstack-nova12:26
*** fabianw has joined #openstack-nova12:27
*** baoli_ has joined #openstack-nova12:27
*** fabianw has quit IRC12:27
*** TravT has joined #openstack-nova12:28
*** baoli has quit IRC12:30
kaisersdanpb: thnx!12:31
*** hoangcx has joined #openstack-nova12:31
*** shaohe_feng has quit IRC12:32
*** shaohe_feng has joined #openstack-nova12:33
*** daemontool_ has joined #openstack-nova12:35
openstackgerritAlex Szarka proposed openstack/nova: Recover instances when physical data is missing  https://review.openstack.org/32234112:36
*** sdake has joined #openstack-nova12:37
*** daemontool has quit IRC12:37
johnthetubaguymdbooth: ah, I think I get you, I am really not that familiar with that bit of the imagecache code12:38
openstackgerritAlex Szarka proposed openstack/nova: Recover instances when physical data is missing  https://review.openstack.org/32234112:39
*** c64cosmin has joined #openstack-nova12:39
*** ljxiash has joined #openstack-nova12:41
openstackgerritAlex Szarka proposed openstack/nova: Recover instances when physical data is missing  https://review.openstack.org/32234112:41
*** shaohe_feng has quit IRC12:42
*** lucas-hungry is now known as lucasagomes12:43
*** shaohe_feng has joined #openstack-nova12:43
openstackgerritClaudiu Belu proposed openstack/nova-specs: Reproposes the Hyper-V Cluster spec to Ocata  https://review.openstack.org/35055912:46
*** ratailor has quit IRC12:47
alex_xunova api subteam meeting in 10 mins at #openstack-meeting-412:50
alex_xu\o/ yeah, i sent reminder on time12:50
*** annegentle has joined #openstack-nova12:52
*** jamesdenton has quit IRC12:52
*** shaohe_feng has quit IRC12:53
*** shaohe_feng has joined #openstack-nova12:54
sdagueo/12:59
*** gmann_ has joined #openstack-nova12:59
*** annegentle has quit IRC13:00
mdboothjohnthetubaguy: I don't think anybody is :) That's why I try to submit mini patch series every time I think I understand a little of it.13:00
mdboothkswiatek: I just got here: http://paste.openstack.org/show/547756/13:02
*** rlrossit has joined #openstack-nova13:02
mdboothI think something like that would allow you to do what melwitt was asking for13:02
mdboothThat basically mocks the entirety of the external libraries13:02
*** itoader has joined #openstack-nova13:03
*** burgerk has joined #openstack-nova13:03
* kashyap vaguely recalls that there was a spec submitted to allow for applying IO disk limits via QEMU IO throttling infrastructure ('block_set_io_throttle'). Wonder if it's just my imagination...13:03
mdboothAnd has the client return a list of values we'd be interested in testing the filtering of13:03
*** shaohe_feng has quit IRC13:03
*** shaohe_feng has joined #openstack-nova13:04
mdboothkswiatek: It also doesn't mock any of RBDDriver, which melwitt was keen on13:04
openstackgerritJim Rollenhagen proposed openstack/nova-specs: Re-re-propose ironic-multiple-compute-hosts  https://review.openstack.org/35028313:06
jrolldansmith: ^ updated, replied to comments13:06
*** mriedem has joined #openstack-nova13:06
mdboothkswiatek: What problem did you hit when you tried to mock rados?13:08
*** diogogmt has quit IRC13:09
kswiatekhi, I need help with testing patch: https://review.openstack.org/#/c/314928/13:09
*** tblakeslee has joined #openstack-nova13:10
mdboothkswiatek: ^^^ You mentioned you hit a problem when you tried to mock rados13:10
*** dave-mccowan has joined #openstack-nova13:10
mdboothDid you see the above pastebin, btw?13:10
*** diogogmt has joined #openstack-nova13:10
doffmdansmith, alaski: Thanks for working on those aggregate changes. I feel pretty guilty for not being able to see it through.13:10
kswiateksorry didn't saw i had scrolled window up :)13:10
*** sudipto has joined #openstack-nova13:11
*** ducttape_ has quit IRC13:11
*** mdrabe has joined #openstack-nova13:12
kswiatekwhen I mocked rado (like was mocked in other test) i get error in line: ret = self.librbd.rbd_list(ioctx.io, byref(c_names), byref(size)) in rbd.py13:12
kswiatekctypes.ArgumentError: argument 1: <type 'exceptions.TypeError'>: Don't know how to convert parameter 113:13
kswiatekmdbooth i tried your solution13:13
kswiatektry*13:13
*** kolesovdv has quit IRC13:13
*** shaohe_feng has quit IRC13:13
*** TravT has quit IRC13:14
*** shaohe_feng has joined #openstack-nova13:14
*** agireud has quit IRC13:14
mriedemeasy fix to get neutronclient 5.0 going in u-c https://review.openstack.org/#/c/349297/13:15
mdboothkswiatek: What line of the test is that stack trace in?13:15
mdboothI guess it would be in the cleanup call...13:15
mdboothIs it during _destroy_volume?13:16
*** ljxiash has quit IRC13:16
mdboothThe above little test runs to completion for me13:16
mdboothObviously it doesn't actually test anything :)13:16
mdboothBut I stuck a probe in to make sure the volume list is being returned int he right place in cleanup_volumes13:17
*** agireud has joined #openstack-nova13:17
mdboothkswiatek: Pastebin a stack trace for me13:18
*** ljxiash has joined #openstack-nova13:18
*** burgerk has quit IRC13:18
kswiatekmy error was while executing this line in rbd_utils.py: volumes = rbd.RBD().list(client.ioctx)13:18
*** ihrachys has joined #openstack-nova13:18
*** TravT has joined #openstack-nova13:18
mdboothkswiatek: Ok, that suggests it's not mocked13:18
mdboothThat line definitely wfm13:18
openstackgerritJoshua Hesketh proposed openstack/nova: Filter for ironic uuid  https://review.openstack.org/34341313:19
*** markvoelker has quit IRC13:19
mdboothMy test fragment mocks the whole rbd library, so that shouldn't be able to raise an exception, because it shouldn't be running any code13:19
kswiatekhttp://paste.openstack.org/show/547764/13:20
mdboothkswiatek: Can you pastebin the test function which resulted in that?13:20
*** cleong has joined #openstack-nova13:20
mdboothIs it literally what I posted?13:20
kswiatekI just run code with your code and waiting for result13:21
kswiatekno13:21
kswiatekit's my previous code :)13:21
mdboothOk, pastebin what you ran13:21
mdboothAh...13:21
mdboothSo, yeah. You need to mock those libraries, otherwise it's going to fail.13:21
mdboothWe don't want it actually attempting to do real ceph stuff.13:21
mdboothYou asked about what ioctx should be13:23
*** ljxiash has quit IRC13:23
mdboothThe real answer is it's not at all important to the test, so we should sweep it under the carpet :)13:23
openstackgerritJoshua Hesketh proposed openstack/nova: Filter for ironic uuid  https://review.openstack.org/34341313:23
*** shaohe_feng has quit IRC13:23
*** shaohe_feng has joined #openstack-nova13:24
kswiatekyour code seems working13:26
*** roeyc has quit IRC13:27
*** TravT has quit IRC13:27
kswiateki mean no error, so what's left is to check if it clean what it should (or don't clean)13:27
*** jwcroppe has quit IRC13:28
mdboothkswiatek: That snippet is just a demo of some mocking.13:28
*** fragatina has joined #openstack-nova13:28
mdboothYou need to write the actual test :)13:28
*** takedakn has quit IRC13:28
*** ihrachys has quit IRC13:29
*** nkrinner is now known as nkrinner_afk13:29
mdboothSo you'll want the volume list to be something sensible, then assert that _destroy_volume was called as expected based on the filter we passed it.13:29
*** ljxiash has joined #openstack-nova13:31
*** belmoreira has quit IRC13:31
openstackgerritJim Rollenhagen proposed openstack/nova: Ironic: enable multitenant networking  https://review.openstack.org/29789513:31
*** ljxiash has quit IRC13:32
jrolljaypipes: mriedem: ^ updated depends-on for the ironicclient bump13:32
*** tonytan4ever has joined #openstack-nova13:32
*** ljxiash has joined #openstack-nova13:32
*** ttx has quit IRC13:32
*** ttx has joined #openstack-nova13:33
*** roeyc has joined #openstack-nova13:33
*** diogogmt has quit IRC13:33
*** ttx has quit IRC13:33
*** shaohe_feng has quit IRC13:34
*** ttx has joined #openstack-nova13:34
*** shaohe_feng has joined #openstack-nova13:34
*** ttx has quit IRC13:35
*** ttx has joined #openstack-nova13:36
mriedemjroll: thanks13:37
*** hoangcx has quit IRC13:37
*** hwoarang has quit IRC13:37
jrollnp13:39
*** TravT has joined #openstack-nova13:41
*** jwcroppe has joined #openstack-nova13:42
*** shaohe_feng has quit IRC13:44
openstackgerritAnush Krishnamurthy proposed openstack/nova: Improve help text of image_file_url  https://review.openstack.org/33249213:44
*** shaohe_feng has joined #openstack-nova13:45
dansmithjroll: if you remove that one thing about testing with services down, I'll +2 that spec13:46
*** markvoelker has joined #openstack-nova13:46
jrolldansmith: sure13:47
jrolldansmith: "I wonder, could we mitigate that by starting a new compute service in disabled mode" <- sounds right13:47
dansmithjroll: cool13:47
openstackgerritJim Rollenhagen proposed openstack/nova-specs: Re-re-propose ironic-multiple-compute-hosts  https://review.openstack.org/35028313:48
jrollthere you are sir, thanks13:48
*** sudipto has quit IRC13:50
*** zz_dimtruck is now known as dimtruck13:50
*** fragatina has quit IRC13:51
openstackgerritDaniel Berrange proposed openstack/nova: network: introduce helper APIs for dealing with os-vif objects  https://review.openstack.org/26967213:52
openstackgerritDaniel Berrange proposed openstack/nova: libvirt: convert over to use os-vif for Linux Bridge & OVS  https://review.openstack.org/35059513:52
*** hwoarang has joined #openstack-nova13:52
mriedemdanpb: good call on splitting that change up :)13:52
*** devkulkarni1 has joined #openstack-nova13:53
*** eharney has joined #openstack-nova13:54
*** ecelik has quit IRC13:54
*** shaohe_feng has quit IRC13:54
*** akshai has joined #openstack-nova13:54
*** TravT has quit IRC13:55
*** shaohe_feng has joined #openstack-nova13:55
*** achinm has joined #openstack-nova13:55
*** burgerk has joined #openstack-nova13:57
dansmithmriedem: when were you going to do a mitaka release?13:59
*** TravT has joined #openstack-nova14:00
openstackgerritPaul Carlton proposed openstack/nova-specs: Use libvirt storage pools  https://review.openstack.org/31050514:00
*** fragatina has joined #openstack-nova14:00
*** krot_sickleave is now known as krotscheck14:00
*** c64cosmin has quit IRC14:00
*** anush_ has joined #openstack-nova14:00
*** akshai has quit IRC14:00
gmann_sdague: alex_xu mriedem johnthetubaguy do we need to backport imageRef bug  to mitaka/liberty ?14:01
*** tongli has joined #openstack-nova14:01
gmann_if we are fixing this as bug, may be in mitaka also make sense14:01
mriedemdansmith: soon grasshopper14:01
bhagyashrismriedem: Hi, could you please review the patch https://review.openstack.org/#/c/320787/14:02
*** akshai has joined #openstack-nova14:02
mriedemdansmith: have some stuff queued up first this morning, but wanted to see what was flushed through last night too14:02
lyarwoodmriedem: can you take another look at https://review.openstack.org/#/c/349897/ for stable/mitaka now that I've dropped the rootwrap changes?14:02
dansmithmriedem: I think we should get that crypto fix from lyarwood in there.. it's simple but completely breaks encrypted volumes for people with real setups14:02
sdaguegmann_: no, because it's tweaking the image API. Which we can accept as part of moving to glance v2 for the backend, but I don't think we should force earlier14:02
mriedemlyarwood: dansmith: i see how it is, the double team14:02
dansmithmriedem: you know you like it14:03
dansmithmriedem: I've spent the last two days watching this unfold and lyarwood swoops in with a fix, so.. :)14:03
mriedemi can't say the image i'm thinking in my head14:03
*** eharney_ has joined #openstack-nova14:04
mriedemwell, the fix was in master14:04
mriedemlyarwood swooped in with hitting a backport button in gerrit14:04
gmann_sdague: ah nice point. glance v2 not there.14:04
lyarwoodmriedem: well that and I had to type stable/mitaka14:04
*** eharney has quit IRC14:04
*** shaohe_feng has quit IRC14:04
*** tongli has quit IRC14:04
*** tongli has joined #openstack-nova14:05
mriedemlyarwood: dansmith: so i brought the rootwrap thing up on the original change in master and was somehow sold on that not being an issue for upgrades, but i'd have to go back and look at the reasoning14:05
dansmithmriedem: I know, I meant swooped in with a pointer to the fix14:05
*** shaohe_feng has joined #openstack-nova14:05
dansmithmriedem: it's easy to drop that from the backport, and I think that's fine.. the backport makes it more specific, which is probably not a problem for upgrades anyway,14:05
*** jwcroppe has quit IRC14:05
dansmithbut there's really no reason to include it, IMHO14:05
*** ducttape_ has joined #openstack-nova14:06
dansmithit's a minor tightening on a system that already lets us run dd unrestricted, so..14:06
mriedemyeah it goes from wildcard to specific'ish14:06
dansmithwe could easily have to backport something that requires it to go back to wildcard and then be screwed14:06
dansmithso I'd just leave it out, personally14:06
mriedemi'm fine with that approach14:06
mriedemlyarwood: just leave a note in the commit message about the deviation from the original fix and why14:07
mriedemand then i think i'm happy14:07
mriedemwell, as happy as i can be14:07
*** cfriesen_ has joined #openstack-nova14:07
vladikrany chance someone could please look this spec, https://review.openstack.org/#/c/350211  - has been resubmitted yesterday, it was previously approved for Newton14:08
lyarwoodmriedem: thanks, will do.14:08
mriedemgmann_: no on backporting the imageref thing to liberty14:08
vladikrhope it wont miss the freeze tomorrow14:08
mriedemgmann_: it's a latent bug, changes api behavior, and isn't critical/security related14:08
mriedemvladikr: we aren't really opening up reviews en masse for ocata14:09
openstackgerritGhanshyam Mann proposed openstack/nova: Strict ImageRef validation to UUID only  https://review.openstack.org/33880214:10
gmann_mriedem: yea, true.14:12
*** ayoung has joined #openstack-nova14:12
jaypipesjroll: will re-review the resource_class patch shortly.14:13
jrollthanks jaypipes14:13
*** TravT has quit IRC14:13
jaypipesjroll: let's definitely get that done today :)14:13
jrolldansmith: random question, are o.vo fields singletons shared between instances of an object?14:13
jrolldansmith: context, we're hacking default to be calculated at object init >.>14:14
dansmithjroll: no14:14
kswiatekmdbooth: you mean volume with instance.uuid which will pass condition and execute destroy_volume? and assert that destroy_volume will be called?14:14
dansmithjroll: oh, you mean between instances of a single class definition?14:14
dansmithjroll: fields = {} is on the class not the instance14:15
*** shaohe_feng has quit IRC14:15
dansmithjroll: so there is one instance of the field per class14:15
mdboothkswiatek: Yeah. It should destroy the correct volumes, and no others.14:15
jrolldansmith: yeah, between two instances of Node() - if I modify node_instance.fields['foo']._default, should that change it for a second instance?14:15
*** shaohe_feng has joined #openstack-nova14:15
dansmithjroll: yeah, because it's on the class14:15
jrolldansmith: aha, so I think the answer is yes14:15
jrolldamn14:15
dansmithjroll: that's kindof the point, right? if you have two instances of the same object that don't share the same set of fields, then it's hard to argue that they're compatible :)14:16
jrolldansmith: right, yeah14:17
*** oanson has quit IRC14:17
mriedemgmann_: you should also update the api-ref for imageRef14:17
mriedemsince it says " The image reference, as a UUID or full URL, for the image to use for your server instance."14:17
*** mlavalle has joined #openstack-nova14:18
*** amotoki has joined #openstack-nova14:18
*** TravT has joined #openstack-nova14:18
mriedemalex_xu: did you guys talk about the bug for restricting sort_keys when listing instances?14:19
openstackgerritLee Yarwood proposed openstack/nova: block_device: Make refresh_conn_infos py3 compatible  https://review.openstack.org/34211114:19
gmann_mriedem: ah, yea that should be update.14:20
*** siva_krish has joined #openstack-nova14:21
alex_xumriedem: sorry, we didn't get chance14:21
*** cdelatte has joined #openstack-nova14:21
alex_xumriedem: we want to remove the sort_keys?14:23
mriedemalex_xu: https://bugs.launchpad.net/nova/+bug/160903914:23
openstackLaunchpad bug 1609039 in OpenStack Compute (nova) "Should not be able to sort instances based on joined tables" [High,Triaged]14:23
mriedemwe want to restrict sorting on joined tables14:23
*** links has quit IRC14:23
mriedemwhen listing instances14:23
*** claudiub has quit IRC14:24
mriedemi'm not sure what the best way to implement that is, whitelist via jsonschema in the api or just check in the db api for sort_keys that are on joined columns14:24
sdaguemriedem: jsonschema doesn't apply to parameters14:24
*** siva_krish has quit IRC14:24
mriedemthat settles that then14:24
alex_xumriedem: I can take at look at that bug14:25
sdaguewe don't have a validation mechanism for parameters yet... which we should14:25
alex_xuyes14:25
*** shaohe_feng has quit IRC14:25
*** jaosorior has quit IRC14:25
mriedemi wasn't crazy about doing jsonschema validation on this anyway14:25
mriedemb/c if a new column is added to the instance table that's not a joined column you'd have to update the schema and microversion it14:25
mriedemmaybe that's actually desirable, idk14:26
*** shaohe_feng has joined #openstack-nova14:26
tdurakovmriedem, alex_xu hi folks, read this thread http://lists.openstack.org/pipermail/openstack-dev/2016-August/100606.html - I'd prefer to not bump microversion for this honestly, but what I'm concerned more, is whetherit's ok to allow such resizes? do we have enough code on compute.manager side to perform this?14:26
*** siva_krish has joined #openstack-nova14:26
sdaguemriedem: well, sorting by any collumn that doesn't have an index on it, is basically a DOS vector14:26
sdagueso I do think the allowed sort columns should be explicit14:26
*** unicell1 has joined #openstack-nova14:27
mriedemalright maybe comment that in the bug so we don't lose it14:27
mriedemi have to run to a meeting14:27
*** ekuris_ has quit IRC14:27
kswiatekmdbooth: something like this: http://paste.openstack.org/show/547785/14:28
*** gszasz has quit IRC14:28
*** yamahata has joined #openstack-nova14:28
*** unicell has quit IRC14:29
mdboothkswiatek: Yep, but with more representative volume names.14:29
kswiatek:) ok14:29
mdboothI can't remember exactly how rbd names them14:29
mdboothBut I think it's like <uuid>_<name>14:29
mdboothAnyway, we should ensure that data is representative14:30
alex_xutdurakov: emm...basically that is the patch https://review.openstack.org/#/c/334747/ want to resolve14:30
kswiatekok14:30
tdurakovalex_xu: yes, I saw it14:30
alex_xutdurakov: at least, when I review the code, it looks like ok allow the resize14:30
tdurakovalex_xu: tbh, I'd solve this problem with another approach14:31
*** gszasz has joined #openstack-nova14:31
*** ihrachys has joined #openstack-nova14:31
tdurakovif smth fails during resize, do rolback, not just change instance state14:31
*** TravT has quit IRC14:32
tdurakovalso why not to wrap finish resize the same way, so instance will moved to active state, as it's done on source node14:32
alex_xutdurakov: yea, i see your point now, when I review I remember I think about that case, but I can't remember the detail now...14:33
*** siva_krish has quit IRC14:33
tdurakovalex_xu: from my sight this should be internal thing, from operator/user it's no matter where it's failed, so let's just not expose that at all14:33
tdurakovmriedem: what do you think^14:34
*** shaohe_feng has quit IRC14:35
*** shaohe_feng has joined #openstack-nova14:36
alex_xutdurakov: yea, i understand your point and agree, but i remember I say no to that way when i review, but i can't remember the detail now. I will try to revisit that code again.14:36
tdurakovalex_xu: yes, please:)14:37
*** code-R has joined #openstack-nova14:37
alex_xutdurakov: you can write your comment on the patch, i will try to revisit it tomorrow.14:37
*** jlanoux has quit IRC14:37
tdurakovalex_xu: sure, will do this. will be waiting for your feedback on that approach14:38
alex_xutdurakov: yea, thanks14:38
*** burgerk has quit IRC14:38
*** slucidi has quit IRC14:38
*** burt has joined #openstack-nova14:38
*** achinm has quit IRC14:38
*** code-R_ has joined #openstack-nova14:38
alex_xusee you tomorrow nova...14:39
*** efried has joined #openstack-nova14:40
openstackgerritGhanshyam Mann proposed openstack/nova: Strict ImageRef validation to UUID only  https://review.openstack.org/33880214:40
*** mtanino has joined #openstack-nova14:40
*** diogogmt has joined #openstack-nova14:40
kswiatekmdbooth; last thing, quite silly, random names hardcode or generate?14:40
*** amotoki has quit IRC14:40
*** TravT has joined #openstack-nova14:41
mdboothkswiatek: I'd use the uuids module for instance uuid. For the name component, use a couple of real names. e.g. 'disk' and 'disk.swap'14:41
kswiatekok thanks14:41
*** code-R has quit IRC14:41
mdboothkswiatek: But check what the actual format is, because I haven't.14:42
*** burgerk has joined #openstack-nova14:43
*** chenfan_ has joined #openstack-nova14:44
*** chandankumar has quit IRC14:44
*** gmann_ has quit IRC14:44
*** Swami has joined #openstack-nova14:45
*** TravT has quit IRC14:45
*** amotoki has joined #openstack-nova14:45
*** shaohe_feng has quit IRC14:45
*** toabctl has quit IRC14:46
*** TravT has joined #openstack-nova14:46
*** shaohe_feng has joined #openstack-nova14:46
*** kenperkins has joined #openstack-nova14:46
*** toabctl has joined #openstack-nova14:47
*** annegentle has joined #openstack-nova14:47
*** hieulq_ has joined #openstack-nova14:47
*** knikolla has joined #openstack-nova14:47
*** code-R_ has quit IRC14:48
*** code-R has joined #openstack-nova14:48
*** fragatina has quit IRC14:49
*** TravT_ has joined #openstack-nova14:49
*** fragatina has joined #openstack-nova14:49
*** TravT has quit IRC14:50
*** TravT_ is now known as TravT14:51
*** eharney_ is now known as eharney14:51
*** mdrabe has quit IRC14:53
*** tjones has joined #openstack-nova14:54
*** shaohe_feng has quit IRC14:56
*** shaohe_feng has joined #openstack-nova14:56
*** siva_krish has joined #openstack-nova14:58
*** mdrabe has joined #openstack-nova15:01
*** gszasz has quit IRC15:02
*** fragatina has quit IRC15:02
dansmithjaypipes: FYI, I'm fixing that rt unit test series for the pep8 fail and the useless code gibi pointed out15:03
*** hieulq__ has joined #openstack-nova15:05
*** nkrinner_afk has quit IRC15:05
openstackgerritDan Smith proposed openstack/nova: rt: Remove duplicate unit test for missing mig ctx  https://review.openstack.org/35029315:05
openstackgerritDan Smith proposed openstack/nova: rt: Refactor resize claim abort unit test  https://review.openstack.org/35029215:05
openstackgerritDan Smith proposed openstack/nova: rt: Refactor test_dupe_filter unit test  https://review.openstack.org/35031915:05
openstackgerritDan Smith proposed openstack/nova: rt: Refactor resize_claim unit test  https://review.openstack.org/34974615:05
*** jlanoux has joined #openstack-nova15:06
*** shaohe_feng has quit IRC15:06
*** amotoki has quit IRC15:06
*** shaohe_feng has joined #openstack-nova15:06
*** mjura has quit IRC15:06
*** hieulq_ has quit IRC15:07
jaypipesdansmith: cool, thx. btw, that useless code was just copy/pasted from the original :)15:07
dansmithjaypipes: okay I didn't realize, but since this was a cleanup, seems okay to fix right?15:08
jaypipesdansmith: oh, absolutely, and I appreciate it!15:08
*** akshai has quit IRC15:09
kswiatekmdbooth: one more thing: commit changes one by one, or change everything form comments and then commit?15:09
*** rama_y has joined #openstack-nova15:09
mdboothkswiatek: Not sure I understand. There should only be a single resulting change.15:10
*** chenfan_ has quit IRC15:10
kswiatekok15:11
*** rrecio has joined #openstack-nova15:11
*** siva_krish has quit IRC15:12
*** gszasz has joined #openstack-nova15:14
mriedemtdurakov: i tend to agree with you and alaski (who commented in the change) that if we can rollback we should just do that automatically15:14
*** dimtruck is now known as zz_dimtruck15:15
*** zz_dimtruck is now known as dimtruck15:16
*** chenfan_ has joined #openstack-nova15:16
dansmithjaypipes: oops, forgot to remove your datetime import on that last one, so re-pushing in a sec15:16
*** shaohe_feng has quit IRC15:16
*** nkrinner_afk has joined #openstack-nova15:17
dansmithoh, nm, it's used for a timedelta, so.. it's good15:17
*** shaohe_feng has joined #openstack-nova15:17
mriedemmikal: since you were on the change on master maybe you want to hit the backport https://review.openstack.org/#/c/349986/15:17
mriedemdansmith: https://review.openstack.org/#/c/314207/15:17
mriedemand https://review.openstack.org/#/c/304746/15:17
mriedemor johnthetubaguy ^15:18
*** akshai has joined #openstack-nova15:18
*** priteau has quit IRC15:19
*** paul-carlton2 has quit IRC15:20
*** READ10 has joined #openstack-nova15:20
mriedemmdbooth: on https://review.openstack.org/#/c/329930/ i'm pretty sure the bdm.save() actually happens in the virt driver15:20
mriedemwhen prepping block devices15:20
mriedemmdbooth: it doesn't hurt to add the save(), but i'm assuming that was the thought process15:21
mdboothmriedem: IIRC I couldn't immediately see where it was called, and it wasn't documented that it was called elsewhere15:21
mriedemin driver_block_device.attach_block_devices15:21
mdboothI'm engaged in a war on non-local context15:22
mriedemvia the @update_db decorator or whatever15:22
mriedemmdbooth: yeah i understand15:22
mdboothWill update_db do it?15:22
mdboothIt'll flush the sqla object, but without the save() does the change get pushed there?15:22
*** amotoki has joined #openstack-nova15:22
*** markus_z has quit IRC15:22
mriedemyes i believe it happens indirectly15:23
*** sheel has joined #openstack-nova15:23
mriedemthe DriverBlockDevice is created from the bdm object after that field is set,15:23
mriedemthen we call attach which has the update_db decorator15:23
*** chandankumar has joined #openstack-nova15:24
*** dave-mccowan has quit IRC15:24
mriedemit's totally tight coupling and you wouldn't know if you didn't memorize that entire code flow15:24
*** hoangcx has joined #openstack-nova15:24
dansmithyeah, it's the terribles15:24
*** dimtruck is now known as zz_dimtruck15:25
*** shaohe_feng has quit IRC15:26
*** shaohe_feng has joined #openstack-nova15:27
*** ihrachys has quit IRC15:29
*** ihrachys has joined #openstack-nova15:30
*** roeyc has quit IRC15:31
moshelejaypipes: reminder for pci reviews https://review.openstack.org/#/c/328983/ , https://review.openstack.org/#/c/347444/ https://review.openstack.org/#/c/347558/, thanks in advance :)15:31
johnthetubaguymdbooth: mriedem: dansmith: I keep thinking about attaching the BDM to the instance, and start using instance.save() to save it, we seem to fetch BDM so many times during the boot process to double check things, and save it only some times, its so dam confusing. Is that a crazy thought?15:31
jaypipesmoshele: you got it. will do shortly.15:32
mdboothjohnthetubaguy: Makes sense to me. For the persistent instance metadata stuff I'll want to maintain the ability to poke them individually via DriverBlockDevice, though.15:33
mriedemjohnthetubaguy: idk, a bdm doesn't need an instance_uuid set right away15:33
mriedemand i don't know what changes with multiattach15:34
mriedemor i forget rather15:34
johnthetubaguywell multiattach just has a bdm for each attachment, I assumed, but I don't remember either15:34
johnthetubaguyyeah, it feels like a the request_spec could store the details for a bdm without a uuid, to some extent, so we can then always set the uuid on a bdm15:35
*** sridharg has quit IRC15:36
johnthetubaguymaybe thats the bit I never understood actually, a bdm without the instance_uuid set15:36
andreykurilinmriedem: https://review.openstack.org/344151 +6, -5 change and py35 is green in novaclient :)15:36
mdboothjohnthetubaguy: Isn't it just a quirk of how they're created in api?15:36
* mdbooth digs the code out again.15:37
*** shaohe_feng has quit IRC15:37
johnthetubaguymdbooth: that was my hope15:37
*** shaohe_feng has joined #openstack-nova15:37
*** paul-carlton has quit IRC15:38
openstackgerritAndrey Volkov proposed openstack/python-novaclient: Add support for microversion 2.33  https://review.openstack.org/34954715:39
*** itlinux has joined #openstack-nova15:39
andreykurilinmriedem: thanks15:39
*** zz_dimtruck is now known as dimtruck15:40
openstackgerritMatt Riedemann proposed openstack/nova-specs: Deprecate os-interface proxy API  https://review.openstack.org/35027715:40
dansmithjohnthetubaguy: I've always thought it was weird that BDM has instance inside *it* instead of the other way around, yes15:41
mdboothjohnthetubaguy: BlockDeviceMapping.instance_uuid isn't nullable. However, when the object is originally created (before it's persisted) it's not set to handle the multiple create case15:42
dansmithjohnthetubaguy: I'd also like to unfsck a bunch of the BDM code if we're going to do that though15:42
mdboothSee _bdm_validate_set_size_and_instance()15:42
johnthetubaguydansmith: +115:42
mdboothdansmith: +115:42
johnthetubaguymdbooth: yeah, thats OK, as long as you can't save without and instance_uuid15:42
johnthetubaguymdbooth: yeah, I was just looking in here: https://github.com/openstack/nova/blob/841fc2ecde608a012660828011cf1f5dea322026/nova/compute/api.py#L143415:43
*** devkulkarni2 has joined #openstack-nova15:44
edleafeif a patch adds a new test method, and the rest of the tests in that file use mox, is it OK for the new test to also use mox?15:44
dansmithedleafe: mriedem will smack you down for that15:45
*** devkulkarni1 has quit IRC15:45
*** pece has joined #openstack-nova15:45
mdboothedleafe: I'd say it depends if the new test follows an existing (anti-)pattern.15:45
johnthetubaguyso we are currently trying to remove all mox tests15:45
mdboothIf there are 4 almost identical mox-using tests and you want to add a 5th variation15:45
edleafeWell, I'm looking at https://review.openstack.org/#/c/338802/9/nova/tests/unit/api/openstack/compute/test_block_device_mapping.py15:45
mdboothYou should either follow the pattern, or also fix all the others15:46
mdboothconsistency ftw15:46
dansmithedleafe: I _just_ got smacked for that in the last couple weeks, so .. :)15:46
edleafemdbooth: yeah, but fixing the others is really not related to the current change15:46
mriedemuse mock15:46
mriedemif it's a new test, use mock15:46
dansmithedleafe: charges are pending, but he's out on bail15:46
mriedemdon't need to fix all the others15:46
edleafedansmith: I think mriedem just likes smacking you15:46
mriedemuse mock with a new test15:46
mdboothBut basically, you should do what mriedem says :)15:46
mriedemmox races under py315:47
mriedemand there are no more docs for mox15:47
mriedemit's basically EOL15:47
mriedemuse mock15:47
mriedemgdi15:47
*** shaohe_feng has quit IRC15:47
mriedemkeep in mind we have a whole fing blueprint for removing mox15:47
*** armax has joined #openstack-nova15:47
mriedemso let's not http://s2.quickmeme.com/img/0c/0c0fcf30234ad83af446aaa5f1484afa60a6dd9dac943371fb4c026c7802fb04.jpg15:48
edleafemriedem: thanks. Wasn't sure if I should smack gmann for that or not15:48
*** shaohe_feng has joined #openstack-nova15:48
mriedemsmack away15:48
mdboothmriedem: So I'd argue that ^^^ is an example of where we want consistency15:48
* edleafe gets out the long paddle...15:48
mdboothSo either use mox, or fix them all15:48
mdboothAnd if the former isn't an option...15:49
edleafemdbooth: or fix the others in a follow-up patch15:49
dansmithor pre-patch15:49
mriedemthat ^15:49
mriedemwhatever, don't do it all in the same thing15:49
mriedembecause backports15:49
mriedempeople15:49
mriedemplus review sanity15:49
edleafethat ^^15:49
mriedemthere is like a series of 5 changes in stable/mitaka for converting mox to mock15:50
mriedemso the backport thing is an issue15:50
*** ihrachys has quit IRC15:51
*** mdrabe_ has joined #openstack-nova15:51
mriedem4th bullet down http://docs.openstack.org/infra/manual/developers.html#peer-review15:51
mriedems/bullet/item/15:51
melwittmy thinking is that someone has to start a new consistency and making a new test use mock helps to do that. that is, if you add a new test with mock, that will let the next person copy the pattern for the next new test, and so on15:52
mriedemyes yes yes15:52
mriedemthat's why we have TestNeutronv2WithMock15:52
mriedembecause i forced it15:52
edleafemelwitt: and provide a pattern for updating others15:52
*** rmart04 has quit IRC15:52
mriedemmelwitt: poke https://review.openstack.org/#/c/349297/15:53
mriedemspeaking of ^15:53
dansmithmriedem: why are you so hyper this morning?15:53
mriedemanger15:53
*** moshele has quit IRC15:53
mriedem+coffee15:53
dansmithheh15:53
* melwitt looks15:54
*** Sukhdev has joined #openstack-nova15:54
*** mdrabe has quit IRC15:55
*** locust has quit IRC15:57
*** JoseMello has joined #openstack-nova15:57
*** shaohe_feng has quit IRC15:57
*** shaohe_feng has joined #openstack-nova15:58
*** akshai has quit IRC15:59
dansmithmriedem: FWIW, jaypipes and I are good with the ironic multi-compute spec change, after a few rounds with jroll: https://review.openstack.org/#/c/35028316:00
*** annegentle has quit IRC16:01
*** mgoddard_ has joined #openstack-nova16:01
mriedemok, will queue it up16:02
dansmiththanks16:02
*** annegentle has joined #openstack-nova16:02
*** chenfan has quit IRC16:03
jrollty16:03
*** mgoddard has quit IRC16:05
*** pbandark has quit IRC16:06
*** jistr is now known as jistr|biab16:06
openstackgerritPushkar Umaranikar proposed openstack/nova: Api-ref: Improve os-migrateLive input parameters  https://review.openstack.org/34565216:07
openstackgerritAnusha Unnam proposed openstack/nova: Config options consistency for consoleauth.py  https://review.openstack.org/34805016:07
*** shaohe_feng has quit IRC16:07
*** chenfan_ has quit IRC16:07
*** shaohe_feng has joined #openstack-nova16:08
*** efried has quit IRC16:11
*** nic has joined #openstack-nova16:12
*** paul-carlton2 has joined #openstack-nova16:12
openstackgerritOpenStack Proposal Bot proposed openstack/nova: Updated from global requirements  https://review.openstack.org/35067316:13
openstackgerritOpenStack Proposal Bot proposed openstack/os-vif: Updated from global requirements  https://review.openstack.org/33991916:14
mriedembut first https://review.openstack.org/#/c/324068/16:15
*** stewie925_ has joined #openstack-nova16:17
*** siva_krishnan has joined #openstack-nova16:17
siva_krishnanjohnthetubaguy: this is regarding your review comments on https://review.openstack.org/#/c/348946/. I just need a small clarification on it.16:18
*** shaohe_feng has quit IRC16:18
jaypipeslbeliveau: so, 326174 is currentl-Workflow'd by you, but it is the dependent patch upon which other patches around SR-IOV live migration from you and moshele are based. Can you fix that?16:18
*** shaohe_feng has joined #openstack-nova16:19
*** chenfan has joined #openstack-nova16:20
*** chandankumar has quit IRC16:20
*** jdurgin1 has joined #openstack-nova16:21
*** hieulq__ has quit IRC16:21
lbeliveaujaypipes: let me have a look16:22
*** jistr|biab is now known as jistr16:22
jrollsdague: when y'all add a new API endpoint, do you 404 it for previous microversions, or?16:22
mriedemhmm, where is our resize policy https://github.com/openstack/nova/tree/master/nova/policies16:23
*** hshiina has quit IRC16:23
openstackgerritdevdatta-kulkarni proposed openstack/nova: TrivialFix: Fixed a typo in nova/test.py  https://review.openstack.org/35068016:24
*** dansmith has quit IRC16:27
*** Apoorva has joined #openstack-nova16:27
*** dansmith has joined #openstack-nova16:28
*** dansmith is now known as Guest2098816:28
*** shaohe_feng has quit IRC16:28
*** shaohe_feng has joined #openstack-nova16:29
*** RuiChen has quit IRC16:29
*** Guest20988 is now known as dansmith16:29
openstackgerritLudovic Beliveau proposed openstack/nova: Allocate PCI devices on migration  https://review.openstack.org/32898316:29
*** ssurana has joined #openstack-nova16:29
*** akshai has joined #openstack-nova16:30
openstackgerritLudovic Beliveau proposed openstack/nova: Fixes for race conditions in revert migration with SR-IOV  https://review.openstack.org/33976516:32
*** akshai_ has joined #openstack-nova16:32
mriedemsdague: comments are in https://review.openstack.org/#/c/324068/16:33
*** sdague has quit IRC16:34
openstackgerritLudovic Beliveau proposed openstack/nova: Fix revert on migration with SR-IOV  https://review.openstack.org/32617416:34
openstackgerritLudovic Beliveau proposed openstack/nova: Fixes for race conditions in revert migration with SR-IOV  https://review.openstack.org/33976516:34
*** akshai has quit IRC16:35
openstackgerritAnusha Unnam proposed openstack/nova: Config options consistency of notifications.py  https://review.openstack.org/34871516:35
openstackgerritBrian Haley proposed openstack/nova: Implement setup_networks_on_host for Neutron networks  https://review.openstack.org/27507316:36
openstackgerritRoman Podoliaka proposed openstack/nova: [WIP] Optional separate database for placement API  https://review.openstack.org/34902116:36
*** abalutoiu has quit IRC16:37
jaypipeslbeliveau: k, let's let the check pipelines run for that patch now. I've reviewed the code in the patch and it all looks good to me.16:37
*** corey_ has joined #openstack-nova16:37
*** rcernin has quit IRC16:38
*** iuliat has joined #openstack-nova16:38
*** cleong has quit IRC16:38
*** shaohe_feng has quit IRC16:38
*** shaohe_feng has joined #openstack-nova16:39
*** itoader has quit IRC16:39
*** Sukhdev has quit IRC16:40
*** seldenr has joined #openstack-nova16:41
openstackgerritLudovic Beliveau proposed openstack/nova: Fixes for race conditions in revert migration with SR-IOV  https://review.openstack.org/33976516:42
*** paul-carlton2 has quit IRC16:43
*** sdague has joined #openstack-nova16:43
openstackgerritRyan Selden proposed openstack/python-novaclient: Improve character encoding in client messages  https://review.openstack.org/35023016:43
*** pece has quit IRC16:44
*** danpb has quit IRC16:44
*** lucasagomes is now known as lucas-dinner16:45
*** irenab has quit IRC16:46
*** hoangcx has quit IRC16:46
lbeliveaujaypipes: should be good now16:48
jaypipeslbeliveau: +2 from me on both patches there. triggered a recheck on the first that had the UNSTABLE thing from the intel nfv ci16:48
lbeliveaujaypipes: thanks !16:48
*** shaohe_feng has quit IRC16:48
*** shaohe_feng has joined #openstack-nova16:49
*** iuliat has quit IRC16:49
*** hieulq_ has joined #openstack-nova16:51
mriedemmtreinish: this single vendor thread in the dev list - what's the point of that again?16:52
mriedemdo projects get kicked out of the arena if they are single vendor?16:52
mriedemor lack diversity (which i thought was the same thing)16:52
mriedemisn't this just a way to scope down the summit and PTG?16:52
*** krotscheck is now known as kro_focused16:53
mtreinishmriedem: yeah, that's what's being proposed. If a project has a single vendor tag for at least 3 cycles in a row we discuss removing them from governance16:55
mriedemand what does that mean?16:55
mriedemif you're out of governance?16:55
mriedemi thought you could still use the CI and all?16:55
mtreinishyou can, you basically just can't call yourself an openstack project16:55
mtreinishthat's about it16:56
mtreinishand horizontal teams don't have to support you16:56
*** unicell1 has quit IRC16:56
dansmithand you don't get a mascot, that's the biggest downside16:56
mriedemlike qa and docs16:56
*** browne has joined #openstack-nova16:56
mriedemno mascot?!16:56
mriedemwell f that16:56
mtreinishmriedem: yeah, qa, docs, osc, etc.16:57
mriedembut docs and qa want to kick that back to the big tent projects via plugins anyway right?16:57
mriedemisn't that already happening?16:57
mriedemhence the whole plugins thread16:58
dansmithalaski: cells meeting in a minute?16:58
alaskidansmith: yes16:58
mriedemwe should just bring back integrated vs non-integrated projects16:58
mtreinishmriedem: it was definitely simpler back then16:58
mriedemthe good old days16:58
mriedemwe need trump to make openstack great again16:58
mriedemhe knows the tricks16:59
mriedemand is the only one that can fix it16:59
*** shaohe_feng has quit IRC16:59
openstackgerritAndrew Laski proposed openstack/nova: Create Instance from BuildRequest if not in a cell  https://review.openstack.org/26392716:59
openstackgerritAndrew Laski proposed openstack/nova: instance.name should be blank if instance.id is not set  https://review.openstack.org/32734216:59
openstackgerritAndrew Laski proposed openstack/nova: Cells: Handle delete with BuildRequest  https://review.openstack.org/32598516:59
mriedemooo right before the meeting so alaski can tell us to review it all, sneaky16:59
*** pcaruana has quit IRC17:00
*** shaohe_feng has joined #openstack-nova17:00
* alaski pretends that was on purpose17:00
mtreinishmriedem: well lets just rewrite it all in: https://github.com/samshadwell/TrumpScript17:00
mriedemhaha17:01
mriedem"No floating point numbers, only integers. America never does anything halfway."17:01
mriedemThere are no import statements allowed. All code has to be home-grown and American made.17:01
mriedemnice17:01
*** annegentle has quit IRC17:02
*** tesseract- has quit IRC17:03
*** jwcroppe has joined #openstack-nova17:04
*** openstackgerrit_ has joined #openstack-nova17:06
tdurakovalaski: are you around?17:07
*** xinliang has quit IRC17:07
*** openstackgerrit_ has quit IRC17:08
*** xinliang has joined #openstack-nova17:08
tdurakovalaski: https://review.openstack.org/#/c/327342/13/nova/objects/instance.py - line 287, why not to re-raise exception here?17:08
mriedemtdurakov: cells v2 meeting atm17:08
tdurakovmriedem: oh, sorry17:08
*** shaohe_feng has quit IRC17:09
*** jlanoux has quit IRC17:09
*** shaohe_feng has joined #openstack-nova17:10
*** akshai_ has quit IRC17:11
*** lpetrut has quit IRC17:11
*** mgoddard_ has quit IRC17:12
*** mgoddard has joined #openstack-nova17:12
*** yamahata has quit IRC17:15
openstackgerritBrian Haley proposed openstack/nova: Implement setup_networks_on_host for Neutron networks  https://review.openstack.org/27507317:15
*** rtheis_ has joined #openstack-nova17:16
*** oomichi has joined #openstack-nova17:17
alaskitdurakov: I'm around, will look after this meeting17:17
*** rtheis has quit IRC17:18
*** moshele has joined #openstack-nova17:19
*** shaohe_feng has quit IRC17:19
*** shaohe_feng has joined #openstack-nova17:20
*** eharney has quit IRC17:20
openstackgerritMikhail Chernik proposed openstack/nova: Fix require thread policy for multi-NUMA computes  https://review.openstack.org/33465917:20
*** dave-mccowan has joined #openstack-nova17:20
*** dimtruck is now known as zz_dimtruck17:22
*** Swami has quit IRC17:22
*** dave-mcc_ has joined #openstack-nova17:22
*** Swami has joined #openstack-nova17:23
*** abalutoiu has joined #openstack-nova17:25
*** paul-carlton2 has joined #openstack-nova17:25
*** hieulq_ has quit IRC17:25
*** zhangjn has quit IRC17:26
*** zhangjn has joined #openstack-nova17:27
*** dave-mccowan has quit IRC17:27
*** hieulq_ has joined #openstack-nova17:27
*** browne1 has joined #openstack-nova17:27
*** browne has quit IRC17:27
*** rahuls has quit IRC17:29
*** shaohe_feng has quit IRC17:29
*** shaohe_feng has joined #openstack-nova17:30
*** unicell has joined #openstack-nova17:31
alaskitdurakov: I didn't want to raise an exception there because no code that checks instance.name expects an exception currently. And I really don't expect that code path to ever get run17:34
*** TravT has quit IRC17:35
*** TravT has joined #openstack-nova17:35
*** lpetrut has joined #openstack-nova17:37
*** daemontool_ has quit IRC17:38
*** shaohe_feng has quit IRC17:40
*** shaohe_feng has joined #openstack-nova17:41
jaypipesdansmith, mriedem, johnthetubaguy: trying to untangle some RT unit tests around resize. Have a question for you all. In the case of a migration/resize to another host, should the source host have the migration object in its tracked_migrations dict? I was assuming that it *would* until the resize is either confirmed (in which case a drop_move_claim() would be sent to the source host for that17:41
jaypipesinstance) or is reverted (in which case the source host would simply remove the migration from its tracked_migrations dict. But what I'm seeing is that the source host does *not* include ongoing migrations in its tracked_migrations dict for migrations/resizes to another host.17:41
dansmithI have no idea17:42
mriedemwas going to say, your guess is as good as mine17:42
openstackgerritWalter A. Boring IV (hemna) proposed openstack/nova: config option: rename libvirt iscsi_use_multipath  https://review.openstack.org/33135717:42
jaypipesmriedem, dansmith, johnthetubaguy: to be clear, this is the block of code that is confusing as hell: https://github.com/openstack/nova/blob/master/nova/compute/resource_tracker.py#L770-L77517:43
dansmiththat's the one?17:43
dansmiththe one confusing block?17:43
jrollheh17:43
*** paul-carlton2 has quit IRC17:43
jrollthe rest of the RT is straightforward, ofc17:43
jaypipesdansmith: yes...17:44
melwittlol17:45
jaypipesbelliott: you around?17:45
jaypipesbelliott: git annotate points at you as a person who might answer this: https://github.com/openstack/nova/commit/5bc0ff6354a399f712c777fa43153f05a50a1ce317:45
*** code-R has quit IRC17:45
*** moshele has quit IRC17:46
mriedemjaypipes: i'd maybe also look for sahid in place of ndipanov17:46
*** mat128 is now known as mat128|afk17:47
tdurakovalaski, I wonder could we end up with instance.uuid as a name in the worst case, i mean the uuid will be used for libvirt domains for example17:47
jaypipesmriedem: ndipanov?17:48
tdurakovIf answer is no, than it's ok)17:48
mriedemjaypipes: git blame showed ndipanov in that code block17:48
mriedemhe worked on a lot of the migration RT stuff17:48
mriedemand sahid has taken over some of that17:48
dansmithyeah I would have thought ndipanov was responsible for that17:48
jaypipesmriedem: ah, gotcha.17:49
jaypipesno, belliott was original code author.17:49
mriedemin the days of yre17:49
mriedem*yore17:49
*** shaohe_feng has quit IRC17:50
sdaguejroll: to earlier question, an API endpoint is an construct that's above microversions. The end point will have semantics around the version range it accepts and tells you is or is not allowed17:50
*** shaohe_feng has joined #openstack-nova17:50
jrollsdague: so if nova were to add /foo, then /foo on 2.2 would be a 406 or something not a 404?17:51
sdaguejroll: ... that's not an endpoint, that's a resource17:52
*** lpetrut has quit IRC17:52
jrollsdague: okay, resource then, sorry :)17:52
sdagueare we talking about major version number?17:52
dansmithjroll: you have to have your pedantic pants on to be in the nova channel17:52
sdague /foo is actually kind of ambiguous17:52
jrollsdague: no, minor versions17:53
jrollok so more concretely17:53
jaypipesmriedem, johnthetubaguy, dansmith: was it your understanding that drop_move_claim() would be called on the source compute host for an instance that had a resize reverted?17:53
sdagueright so /v1/my-awesome-new-thing17:53
jrollyeah17:53
dansmithjaypipes: I'm so touched that you think I know anything about this stuff17:53
jaypipesheh17:53
dansmithhonored, really.17:53
sdagueit's a 404 I'm pretty sure in our code17:53
*** yamahata has joined #openstack-nova17:53
jrollsdague: so GET /v1/my-awesome-new-thing, on an old version, I would think would 404, because that's the old behavior17:54
sdaguejroll: right17:54
jrollsounds like you agree17:54
sdagueyes17:54
jrollthat's helpful, thank you :D17:54
johnthetubaguyjaypipes: I would say yes, but I would totally feel like I am guessing while wearing a blindfold17:54
*** diogogmt has quit IRC17:54
sdaguethe point is the new resource only exists in new versions17:54
sdaguejroll: cool :)17:54
*** claudiub has joined #openstack-nova17:54
alaskitdurakov: that could happen, but that means that instance.id is not available which means the instance is not in the database at all. If we're in the virt driver building an instance from that object we have larger issues than an invalid name.17:55
sdaguejroll: http://developer.openstack.org/api-ref/compute/#server-tags-servers-tags - server tags is an instance where we did that in our API (not top level, but the semantics should be the same)17:56
melwittmriedem: on the neutronclient call patch, do you know why this still works fine but the show_quota doesn't? https://review.openstack.org/#/c/349297/2/nova/tests/unit/network/test_neutronv2.py@168017:56
johnthetubaguyjaypipes: hmm, the more I think about it, I have no idea actually17:56
tdurakovYeah, thats why I'm thinking about reraise17:56
alaskitdurakov: I'm open to raising an exception there, but I'm a bit concerned about side effects of it. I don't know what exception handling exists around it17:56
*** diogogmt has joined #openstack-nova17:57
mriedemmelwitt: no decorator https://github.com/openstack/python-neutronclient/blob/5.0.0/neutronclient/v2_0/client.py#L62517:57
openstackgerritRafael Folco proposed openstack/nova: Improve NUMA packing to fit more instances  https://review.openstack.org/34620517:57
mriedemmelwitt: compared to https://github.com/openstack/python-neutronclient/blob/5.0.0/neutronclient/v2_0/client.py#L59917:57
tdurakovI'd prefer not to handle such at all, as you said at the point there are bigger problems around17:58
*** harlowja has quit IRC17:58
melwittmriedem: ah, okay. I had looked in the change to neutronclient mentioned in lp and only saw the decoration of the client class itself and didn't realize there was per-method stuff too17:58
*** haplo37__ has joined #openstack-nova17:59
*** shaohe_feng has quit IRC18:00
*** harlowja has joined #openstack-nova18:00
*** shaohe_feng has joined #openstack-nova18:01
*** gszasz has quit IRC18:02
*** hieulq_ has quit IRC18:02
*** lpetrut has joined #openstack-nova18:03
*** tonytan4ever has quit IRC18:08
*** dave-mcc_ has quit IRC18:08
*** zz_dimtruck is now known as dimtruck18:09
*** shaohe_feng has quit IRC18:10
*** shaohe_feng has joined #openstack-nova18:11
*** devkulkarni2 has quit IRC18:11
*** syjulian has quit IRC18:12
*** stewie925 has quit IRC18:12
*** xenogear has quit IRC18:12
*** jinli has quit IRC18:12
*** MikeG451 has quit IRC18:13
*** stewie925 has joined #openstack-nova18:17
openstackgerritChristopher Martin proposed openstack/nova: Fix config opt indentation, opt registration and opt group and type check in hyperv.py  https://review.openstack.org/34848418:17
*** code-R has joined #openstack-nova18:17
openstackgerritSarafraj Singh proposed openstack/nova: Merged barbican and key_manager conf files into one.  https://review.openstack.org/34963218:18
*** yamahata has quit IRC18:18
*** yamahata has joined #openstack-nova18:19
*** devkulkarni1 has joined #openstack-nova18:19
*** code-R_ has joined #openstack-nova18:20
*** shaohe_feng has quit IRC18:21
*** claudiub has quit IRC18:21
*** shaohe_feng has joined #openstack-nova18:21
*** code-R has quit IRC18:23
*** sheel has quit IRC18:26
openstackgerritSivasathurappan Radhakrishnan proposed openstack/nova: Config options consistency of pci.py  https://review.openstack.org/35072418:26
*** MikeG451 has joined #openstack-nova18:29
*** jaosorior has joined #openstack-nova18:29
*** jaosorior has quit IRC18:30
*** xinliang has quit IRC18:30
*** shaohe_feng has quit IRC18:31
*** shaohe_feng has joined #openstack-nova18:31
*** xinliang has joined #openstack-nova18:32
*** jinli has joined #openstack-nova18:36
*** amotoki has quit IRC18:37
*** ayoung has quit IRC18:38
*** annegentle has joined #openstack-nova18:39
*** unicell1 has joined #openstack-nova18:40
*** unicell has quit IRC18:41
*** shaohe_feng has quit IRC18:41
*** itoader has joined #openstack-nova18:41
*** syjulian has joined #openstack-nova18:41
*** shaohe_feng has joined #openstack-nova18:42
*** xenogear has joined #openstack-nova18:44
*** ametts has joined #openstack-nova18:45
*** ljxiash has quit IRC18:47
*** iuliat has joined #openstack-nova18:48
*** jdurgin1 has quit IRC18:48
alaskimelwitt: some comments on https://review.openstack.org/#/c/33099518:50
*** itoader has quit IRC18:51
*** shaohe_feng has quit IRC18:51
*** shaohe_feng has joined #openstack-nova18:52
*** haplo37__ has quit IRC18:55
*** lpetrut has quit IRC18:55
*** jed56 has quit IRC18:55
*** lpetrut has joined #openstack-nova18:55
*** claudiub has joined #openstack-nova18:56
*** eharney has joined #openstack-nova18:58
*** diogogmt has quit IRC19:00
*** READ10 has quit IRC19:01
*** shaohe_feng has quit IRC19:02
*** claudiub has quit IRC19:02
*** gjayavelu has joined #openstack-nova19:02
*** efoley has joined #openstack-nova19:02
*** shaohe_feng has joined #openstack-nova19:02
*** itoader has joined #openstack-nova19:04
*** dave-mccowan has joined #openstack-nova19:04
*** iuliat has quit IRC19:06
*** iuliat has joined #openstack-nova19:06
*** READ10 has joined #openstack-nova19:07
*** fifieldt has quit IRC19:07
*** haplo37__ has joined #openstack-nova19:07
*** tonytan4ever has joined #openstack-nova19:08
*** itoader has quit IRC19:09
mikalMorning19:09
mikalmriedem: I'll take a look at stable reviews sometime today19:09
mriedemmikal: thanks19:10
mriedemi'm going to push for a stable/mitaka release tomorrow19:10
*** shaohe_feng has quit IRC19:12
*** shaohe_feng has joined #openstack-nova19:13
mikalmriedem: ok, I will try very hard to not suck at reviews there today then19:13
mikalmriedem: tonyb is off this week minding a sick child, which is why he hasn't been doing much19:13
*** tonytan4ever has quit IRC19:13
*** itoader has joined #openstack-nova19:14
*** Sukhdev has joined #openstack-nova19:15
mriedemmikal: totally not a problem19:15
*** iuliat has quit IRC19:16
*** iuliat has joined #openstack-nova19:17
*** jdurgin1 has joined #openstack-nova19:18
*** fifieldt has joined #openstack-nova19:18
*** itoader has quit IRC19:19
*** abalutoiu has quit IRC19:20
*** oomichi has quit IRC19:20
*** READ10 has quit IRC19:21
*** shaohe_feng has quit IRC19:22
*** efoley has quit IRC19:22
*** shaohe_feng has joined #openstack-nova19:23
*** zul_ has joined #openstack-nova19:23
*** iuliat has quit IRC19:27
alaskimriedem: sdague I have a question on release note tense in https://review.openstack.org/#/c/333098/619:28
*** itoader has joined #openstack-nova19:28
mriedemi don't think have been is correct19:28
mriedema command is provided to migrate server groups to the api database, new server groups are automatically created in the api database19:29
*** efried has joined #openstack-nova19:29
*** tblakeslee is now known as tblakes19:29
sdaguemaybe we need some standard boiler plate for online migrations, because existing ones get moved if updated as well right?19:31
mriedemno19:31
alaskinot in this case19:31
*** tesseract- has joined #openstack-nova19:31
mriedembut that's what everyone thinks by default b/c we were spoiled in kilo with the only flavor migration that dansmith did19:31
mriedems/only/online/19:32
mriedemnew things go to the api db, lookups start in the api db, and fallback to the cell db19:32
*** ametts has quit IRC19:32
mriedemexisting things are manually migrated using the online_data_migrations command19:32
*** bvanhav has quit IRC19:32
dansmithdepends19:32
openstackgerritJay Pipes proposed openstack/nova: rt: Refactor test_dupe_filter unit test  https://review.openstack.org/35031919:32
openstackgerritJay Pipes proposed openstack/nova: rt: Remove incorrect unit test of resize revert  https://review.openstack.org/35075419:32
openstackgerritJay Pipes proposed openstack/nova: rt: Remove dup tests in test_resource_tracker  https://review.openstack.org/35075519:32
openstackgerritJay Pipes proposed openstack/nova: rt: Remove useless unit test in resource tracker  https://review.openstack.org/35075619:32
openstackgerritJay Pipes proposed openstack/nova: rt: Move unit tests for update mig usage  https://review.openstack.org/35075719:32
dansmiththese api moves work that way, but not all online migrations do/will19:32
*** shaohe_feng has quit IRC19:32
*** clenimar has quit IRC19:34
mriedemwznoinsk: btw, is it all intel nfv ci jobs or just Intel NFV CI?19:34
*** diogogmt has joined #openstack-nova19:34
mriedemnvm, i guess the ci is Intel NFV CI19:34
mriedemwith 3 jobs19:34
*** iuliat has joined #openstack-nova19:34
*** kenperkins has quit IRC19:34
*** clenimar has joined #openstack-nova19:35
*** zul_ has quit IRC19:35
*** shaohe_feng has joined #openstack-nova19:35
*** zul_ has joined #openstack-nova19:35
*** tesseract- has quit IRC19:37
*** itoader has quit IRC19:37
jaypipesdansmith: ^^ is some more pushes (and a fix) of those resource tracker unit tests.19:38
*** ihrachys has joined #openstack-nova19:41
*** shaohe_feng has quit IRC19:43
*** shaohe_feng has joined #openstack-nova19:43
*** ametts has joined #openstack-nova19:45
*** ayoung has joined #openstack-nova19:45
*** tonytan4ever has joined #openstack-nova19:49
openstackgerritMatt Riedemann proposed openstack/nova: Add REST API support for get me a network  https://review.openstack.org/31639819:50
openstackgerritLee Yarwood proposed openstack/nova: compute: Use stashed connector during local deletes  https://review.openstack.org/34095119:51
*** itoader has joined #openstack-nova19:52
openstackgerritSarafraj Singh proposed openstack/nova: WIP: Added "networks" object in server-create-request  https://review.openstack.org/30936219:53
*** shaohe_feng has quit IRC19:53
openstackgerritJay Pipes proposed openstack/nova: rt: Move unit tests for update mig usage  https://review.openstack.org/35075719:53
*** shaohe_feng has joined #openstack-nova19:53
*** iuliat has quit IRC19:55
openstackgerritSarafraj Singh proposed openstack/nova: Deprecate cert option  https://review.openstack.org/34805519:55
*** TravT has quit IRC19:56
*** iuliat has joined #openstack-nova19:56
*** TravT has joined #openstack-nova19:59
*** itoader has quit IRC19:59
*** permalac has quit IRC20:02
openstackgerritMichael Glaser proposed openstack/nova: Check opt consistency for neutron.py  https://review.openstack.org/34960520:03
*** shaohe_feng has quit IRC20:03
*** permalac has joined #openstack-nova20:03
*** shaohe_feng has joined #openstack-nova20:04
*** iuliat has quit IRC20:05
*** matrohon has joined #openstack-nova20:06
openstackgerritSarafraj Singh proposed openstack/nova: config options: Improve help for conductor  https://review.openstack.org/32478420:08
*** lpetrut has quit IRC20:10
*** daemontool has joined #openstack-nova20:10
*** kenperkins has joined #openstack-nova20:11
*** coolsvap has quit IRC20:11
*** shaohe_feng has quit IRC20:13
*** shaohe_feng has joined #openstack-nova20:14
*** iuliat has joined #openstack-nova20:19
*** ihrachys has quit IRC20:20
*** itoader has joined #openstack-nova20:23
*** shaohe_feng has quit IRC20:24
*** shaohe_feng has joined #openstack-nova20:24
*** iuliat has quit IRC20:25
*** itoader has quit IRC20:33
*** shaohe_feng has quit IRC20:34
*** Jeffrey4l_ has joined #openstack-nova20:35
*** Jeffrey4l has quit IRC20:35
*** shaohe_feng has joined #openstack-nova20:35
*** devkulkarni1 has quit IRC20:39
*** baoli_ has quit IRC20:39
*** baoli has joined #openstack-nova20:40
openstackgerritAnush Krishnamurthy proposed openstack/nova: Deprecate cert option  https://review.openstack.org/34805520:40
*** shaohe_feng has quit IRC20:44
*** oomichi has joined #openstack-nova20:47
*** shaohe_feng has joined #openstack-nova20:47
*** daemontool has quit IRC20:48
*** smatzek has quit IRC20:49
*** Sukhdev has quit IRC20:52
*** ametts has quit IRC20:52
*** Sukhdev has joined #openstack-nova20:53
*** shaohe_feng has quit IRC20:54
*** shaohe_feng has joined #openstack-nova20:55
*** itoader has joined #openstack-nova20:56
*** jwcroppe has quit IRC20:58
*** itoader has quit IRC21:02
*** shaohe_feng has quit IRC21:05
*** shaohe_feng has joined #openstack-nova21:06
*** READ10 has joined #openstack-nova21:06
*** liverpooler has quit IRC21:08
*** rlrossit has quit IRC21:08
*** JoseMello has quit IRC21:10
mriedemdoffm: you were asking about this this morning21:13
mriedemhttps://github.com/openstack/nova/blob/master/nova/db/sqlalchemy/api.py#L239221:13
mriedemdefault sort order for listing instances is created_at,id21:13
mriedemwhich is why we have https://github.com/openstack/nova/blob/master/nova/db/sqlalchemy/models.py#L21521:13
mriedemwhich is from https://review.openstack.org/#/c/288765/21:14
mriedemthat's in mitaka21:14
*** fifieldt has quit IRC21:15
*** shaohe_feng has quit IRC21:15
*** shaohe_feng has joined #openstack-nova21:16
mriedemdansmith: jaypipes: jroll: you know what's confusing? reading about compute nodes and ironic nodes in the same paragraph21:24
mriedemcompute host / service / node is already a confusing concept21:24
dansmithyup21:24
mriedemso i'm going to ask probably a terribly dumb question, but is there only a single instance in an ironic compute node or?21:25
*** shaohe_feng has quit IRC21:25
*** fifieldt has joined #openstack-nova21:25
dansmithmriedem: correct21:25
dansmithone ironic node is one baremetal host, which can take one instance21:26
mriedemok, because i was reading, "It's also important to note that, due to the way nodes with instances are handled, if an instance is deleted while the compute service is down, that node will be removed from the compute_nodes table when the service comes back up"21:26
*** shaohe_feng has joined #openstack-nova21:26
mriedemand thinking, did he mean 'instances', plural?21:26
*** itoader has joined #openstack-nova21:27
*** baoli has quit IRC21:27
dansmithyeah, it's definitely a bastardization of the rest of our stuff, because we require you to fully consume a node with an instance right now21:27
dansmithresource providers will make that a little less terrible21:27
*** jwcroppe has joined #openstack-nova21:28
jaypipesdansmith: a lot less terrible. :)21:29
dansmithmeasurably less terrible21:30
jaypipes:)21:30
mriedemwe should say, 'at a fraction of the cost' more often because i always love that marketing trick21:31
dansmithheh21:31
mriedemit's .97 of the cost21:31
mriedemyou can't afford not to buy it!21:31
jaypipesheh21:31
*** matrohon has quit IRC21:32
*** moshele has joined #openstack-nova21:32
*** itoader has quit IRC21:34
*** shaohe_feng has quit IRC21:35
*** shaohe_feng has joined #openstack-nova21:36
mriedemok, my comments in https://review.openstack.org/#/c/350283/21:36
mriedemonly real question is https://review.openstack.org/#/c/350283/3/specs/newton/approved/ironic-multiple-compute-hosts.rst@5821:37
*** chenfei has quit IRC21:37
mriedemdansmith: jaypipes: ^ can you answer that for me quick?21:37
*** Qiming has quit IRC21:37
dansmithjust a sec21:37
*** Qiming has joined #openstack-nova21:39
dansmithmriedem: see if that helps21:40
*** chenfei has joined #openstack-nova21:40
*** thorst has quit IRC21:40
jaypipesmriedem: yeah, dansmith answered.21:40
dansmithmriedem: commented on your other thing too21:42
*** sdake has quit IRC21:43
mriedemalright21:43
dansmithmriedem: thanks21:44
dansmithmriedem: so you're good to remove the -2 on the code? or was there something else?21:45
mriedemsomething else http://m.target.com/p/graco-highback-turbobooster/-/A-1528805521:45
mriedemi'm shopping21:45
dansmithgross21:45
*** rtheis_ has quit IRC21:46
mriedemthat kid http://scene7.targetimg1.com/is/image/Target/50881510_Alt05?wid=450&hei=450&fmt=pjpeg21:46
*** shaohe_feng has quit IRC21:46
mriedemand this kid http://scene7.targetimg1.com/is/image/Target/50881510_Alt04?wid=450&hei=450&fmt=pjpeg21:46
mriedemit's going to be rough for them21:46
dansmithhaha21:46
jaypipeslol21:46
mriedemi dropped the -2 on the change21:47
*** shaohe_feng has joined #openstack-nova21:47
*** itoader has joined #openstack-nova21:49
dansmithrock on21:50
mriedemalaski: looks like you have some ci failures in https://review.openstack.org/#/c/325985/21:52
mriedemand therefore i get out of reviewing it tonight21:53
*** burgerk_ has joined #openstack-nova21:53
openstackgerritAlex Szarka proposed openstack/nova: Recover instances when physical data is missing  https://review.openstack.org/32234121:53
*** itoader has quit IRC21:55
*** burgerk has quit IRC21:56
*** shaohe_feng has quit IRC21:56
*** jwcroppe_ has joined #openstack-nova21:56
*** shaohe_feng has joined #openstack-nova21:57
*** jwcroppe has quit IRC21:58
*** burgerk_ has quit IRC21:58
*** corey_ has quit IRC21:58
*** markvoelker has quit IRC21:58
*** markvoelker has joined #openstack-nova21:59
*** mriedem has quit IRC22:01
*** oomichi has quit IRC22:02
*** eharney has quit IRC22:02
*** eharney has joined #openstack-nova22:03
*** thorst has joined #openstack-nova22:03
*** markvoelker has quit IRC22:03
*** shaohe_feng has quit IRC22:06
*** mdrabe_ has quit IRC22:07
*** shaohe_feng has joined #openstack-nova22:07
*** thorst has quit IRC22:07
*** burt has quit IRC22:08
*** cdelatte has quit IRC22:08
*** moshele has quit IRC22:10
*** annegentle has quit IRC22:13
*** yassou has quit IRC22:14
*** jwcroppe has joined #openstack-nova22:14
*** shaohe_feng has quit IRC22:16
*** jwcroppe_ has quit IRC22:17
*** shaohe_feng has joined #openstack-nova22:17
*** dimtruck is now known as zz_dimtruck22:20
*** shaohe_feng has quit IRC22:27
*** shaohe_feng has joined #openstack-nova22:28
*** rfolco has quit IRC22:30
*** mlavalle has quit IRC22:37
*** shaohe_feng has quit IRC22:37
*** shaohe_feng has joined #openstack-nova22:38
*** ducttape_ has quit IRC22:41
*** efried has quit IRC22:43
*** zul_ has quit IRC22:46
*** shaohe_feng has quit IRC22:47
*** zz_dimtruck is now known as dimtruck22:48
*** shaohe_feng has joined #openstack-nova22:48
*** hemna is now known as hemnafk22:52
*** xyang1 has quit IRC22:56
*** shaohe_feng has quit IRC22:57
*** sameo has quit IRC22:58
*** shaohe_feng has joined #openstack-nova22:58
*** ducttape_ has joined #openstack-nova23:02
*** sdake has joined #openstack-nova23:03
*** seldenr has quit IRC23:07
*** shaohe_feng has quit IRC23:08
*** shaohe_feng has joined #openstack-nova23:08
*** diogogmt has quit IRC23:09
*** tblakes has quit IRC23:12
*** shaohe_feng has quit IRC23:18
*** shaohe_feng has joined #openstack-nova23:19
*** diogogmt has joined #openstack-nova23:19
*** ducttape_ has quit IRC23:24
*** ducttape_ has joined #openstack-nova23:25
*** diogogmt has quit IRC23:27
*** shaohe_feng has quit IRC23:28
*** sdague has quit IRC23:29
jaypipesdansmith, Yingxin, edleafe, alex_xu, jroll: https://github.com/jaypipes/os-capabilities23:29
*** shaohe_feng has joined #openstack-nova23:29
*** tonytan4ever has quit IRC23:33
*** shaohe_feng has quit IRC23:38
*** hoonetorg has quit IRC23:39
*** shaohe_feng has joined #openstack-nova23:39
*** tblakes has joined #openstack-nova23:40
*** Swami has quit IRC23:41
alex_xujaypipes: cool!23:43
*** hoonetorg has joined #openstack-nova23:44
*** takashin has joined #openstack-nova23:47
alex_xujaypipes: do we need spec for os-capabilities, or we just comment on github?23:47
jaypipesalex_xu: just sent a ML thread on it. please feel free to comment on that with your suggestions! :) Also, of course, feel free to submit pull requests!23:48
*** jamielennox is now known as jamielennox|away23:48
*** jaypipes has quit IRC23:48
*** shaohe_feng has quit IRC23:49
*** ducttape_ has quit IRC23:49
*** shaohe_feng has joined #openstack-nova23:49
alex_xujaypipes: cool, thanks! not sure have any comment yet, I and YingXin just begin to discuss the interface of os-capabilities, but we didn't any time to do any actual thing :) but will try to catch up and see anything we can help on23:49
*** jamesdenton has joined #openstack-nova23:50
*** sdake has quit IRC23:51
*** rrecio has quit IRC23:54
*** hoangcx has joined #openstack-nova23:55
*** code-R_ has quit IRC23:55
*** sdake has joined #openstack-nova23:55
gmannedleafe: ops, yea i should use mock. just copy paste from old tests :). Thanks for catching that23:58
*** TravT has quit IRC23:59
*** shaohe_feng has quit IRC23:59
*** shaohe_feng has joined #openstack-nova23:59

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