Wednesday, 2018-01-24

*** hshiina has joined #openstack-nova00:00
*** zhurong has joined #openstack-nova00:02
*** jackie-truong has quit IRC00:03
*** owalsh_ has joined #openstack-nova00:05
*** owalsh has quit IRC00:06
*** penick has joined #openstack-nova00:07
*** owalsh_ has quit IRC00:11
*** owalsh has joined #openstack-nova00:16
*** tetsuro has joined #openstack-nova00:19
*** zhurong has quit IRC00:20
*** oanson has quit IRC00:22
*** oanson has joined #openstack-nova00:24
*** yangyapeng has quit IRC00:25
*** yangyapeng has joined #openstack-nova00:25
*** artom has joined #openstack-nova00:26
*** takashin has quit IRC00:28
*** takashin has joined #openstack-nova00:28
*** acormier has joined #openstack-nova00:29
*** yangyapeng has quit IRC00:30
*** penick has quit IRC00:30
*** mlavalle has quit IRC00:31
*** yingjun has joined #openstack-nova00:32
*** penick has joined #openstack-nova00:32
*** acormier has quit IRC00:33
*** Nil_ has quit IRC00:34
*** acormier has joined #openstack-nova00:37
*** Dinesh_Bhor has joined #openstack-nova00:38
*** sdague has joined #openstack-nova00:39
*** acormier has quit IRC00:41
*** liuzz has joined #openstack-nova00:42
*** chyka has quit IRC00:43
openstackgerritTakashi NATSUME proposed openstack/python-novaclient master: Microversion 2.61 - List/Show all server migration types
*** Dinesh_Bhor has quit IRC00:49
*** Dinesh_Bhor has joined #openstack-nova00:50
*** mariusv__ has quit IRC00:51
*** mariusv has joined #openstack-nova00:53
*** mariusv has quit IRC00:53
*** mariusv has joined #openstack-nova00:53
openstackgerritTakashi NATSUME proposed openstack/nova master: Fix missing marker functions
openstackgerritTakashi NATSUME proposed openstack/nova master: [placement] Add functional tests for traits API
openstackgerritTakashi NATSUME proposed openstack/nova master: [placement] Separate API schemas (resource_provider)
openstackgerritTakashi NATSUME proposed openstack/nova master: Fix the order of target host checks
*** zhurong has joined #openstack-nova00:58
*** penick has quit IRC00:58
*** nicolasbock has quit IRC00:58
openstackgerritTakashi NATSUME proposed openstack/nova master: Fix 500 error while passing 4-byte unicode data
*** amodi has quit IRC00:59
openstackgerritTakashi NATSUME proposed openstack/nova master: Adds view builders for keypairs controller
*** r-daneel has quit IRC01:01
*** mdnadeem has joined #openstack-nova01:06
*** zhaochao has joined #openstack-nova01:10
*** phuongnh has joined #openstack-nova01:11
*** tiendc has joined #openstack-nova01:12
openstackgerritOpenStack Proposal Bot proposed openstack/nova master: Updated from global requirements
*** stakeda has joined #openstack-nova01:14
*** crushil has joined #openstack-nova01:15
*** Dinesh_Bhor has quit IRC01:15
*** salv-orlando has joined #openstack-nova01:15
*** eandersson has quit IRC01:16
*** eandersson_ has joined #openstack-nova01:16
*** yangyapeng has joined #openstack-nova01:17
*** armax has quit IRC01:18
*** Dinesh_Bhor has joined #openstack-nova01:19
*** salv-orlando has quit IRC01:20
*** Tom-Tom has joined #openstack-nova01:21
*** david-lyle has quit IRC01:22
openstackgerritOpenStack Proposal Bot proposed openstack/os-traits master: Updated from global requirements
openstackgerritOpenStack Proposal Bot proposed openstack/os-vif master: Updated from global requirements
*** Tom-Tom_ has joined #openstack-nova01:24
*** Tom-Tom has quit IRC01:25
*** yangyapeng has quit IRC01:26
*** yangyapeng has joined #openstack-nova01:26
*** inara has quit IRC01:27
*** armax has joined #openstack-nova01:29
*** inara has joined #openstack-nova01:30
openstackgerritOpenStack Proposal Bot proposed openstack/python-novaclient master: Updated from global requirements
*** takashin has quit IRC01:34
*** oanson has quit IRC01:38
*** sdague has quit IRC01:40
*** oanson has joined #openstack-nova01:41
*** hieulq_ has quit IRC01:42
*** jackie-truong has joined #openstack-nova01:54
jackie-truongIf anyone has a free moment to review this patch, it would be much appreciated:
*** hieulq_ has joined #openstack-nova01:59
*** Guest87011 has quit IRC02:00
*** yassine has joined #openstack-nova02:00
*** yassine is now known as Guest5065102:00
SpazmoticThanks jianghuaw,  I will spin up that bug report tomorrow night when I am at my dev machine and can pull some XenAPI logs for it :)02:00
jianghuawSpazmotic, good:-)02:01
SpazmoticAlso should go ahead and rebase and fix that inline while i'm at it tomorrow.02:01
*** dikonoor has joined #openstack-nova02:03
*** esberglu has joined #openstack-nova02:03
*** esberglu has quit IRC02:06
*** chyka has joined #openstack-nova02:07
*** chyka has quit IRC02:08
*** gcb has joined #openstack-nova02:09
*** zhaochao has quit IRC02:11
*** armax_ has joined #openstack-nova02:12
*** yamamoto has joined #openstack-nova02:13
openstackgerritOpenStack Proposal Bot proposed openstack/os-traits master: Updated from global requirements
openstackgerritOpenStack Proposal Bot proposed openstack/os-vif master: Updated from global requirements
*** armax has quit IRC02:15
*** armax_ is now known as armax02:15
*** Dinesh_Bhor has quit IRC02:15
*** Dinesh__Bhor has joined #openstack-nova02:15
*** hongbin has joined #openstack-nova02:16
*** salv-orlando has joined #openstack-nova02:17
*** harlowja has quit IRC02:20
*** salv-orlando has quit IRC02:21
*** dikonoo has joined #openstack-nova02:22
*** dikonoor has quit IRC02:22
alex_xumriedem: dansmith good news is we have allocation_req_version in Selection obj, then I can convert the alloc_req dict to consistent format in the begining of claim method probably02:24
alex_xumriedem: dansmith here is
*** zhaochao has joined #openstack-nova02:25
openstackgerritsean mooney proposed openstack/nova master: Change 'InstancePCIRequest' spec field
*** gyee has quit IRC02:31
sean-k-mooneystephenfin: now uses objects. ill be in around lunch tomorow if you have any questions.02:32
*** wxy has joined #openstack-nova02:33
*** gongysh has joined #openstack-nova02:35
*** lyan has quit IRC02:47
*** jackie-truong has quit IRC02:52
*** zhurong has quit IRC02:57
*** vladikr has quit IRC03:02
*** vladikr has joined #openstack-nova03:02
*** damien_r has quit IRC03:08
*** damien_r has joined #openstack-nova03:08
*** crushil has quit IRC03:09
openstackgerritDao Cong Tien proposed openstack/nova master: virt/ironic: Implement rescue and unrescue
*** tuanla____ has joined #openstack-nova03:14
*** smatzek has joined #openstack-nova03:17
*** salv-orlando has joined #openstack-nova03:17
*** salv-orlando has quit IRC03:22
*** sree has joined #openstack-nova03:22
*** sree_ has joined #openstack-nova03:25
*** sree_ is now known as Guest6242003:25
*** sree has quit IRC03:29
*** lyan has joined #openstack-nova03:31
*** lyan has quit IRC03:31
*** david-lyle has joined #openstack-nova03:31
*** fragatina has quit IRC03:32
*** david-lyle has quit IRC03:32
*** david-lyle has joined #openstack-nova03:33
*** david-lyle has quit IRC03:39
*** felipemonteiro_ has joined #openstack-nova03:41
*** smatzek has quit IRC03:41
*** felipemonteiro__ has joined #openstack-nova03:42
*** hshiina has quit IRC03:42
*** tbachman has quit IRC03:44
*** felipemonteiro_ has quit IRC03:45
*** bhujay has joined #openstack-nova03:46
*** abhishekk has joined #openstack-nova03:47
*** lyan has joined #openstack-nova03:47
*** annp has joined #openstack-nova03:48
*** brault has quit IRC03:49
*** BenderRodriguez has left #openstack-nova03:50
*** psachin` has joined #openstack-nova03:52
*** psachin has joined #openstack-nova03:52
*** hshiina has joined #openstack-nova03:54
*** bhujay has quit IRC04:00
*** tovin07 has quit IRC04:00
*** tiendc has quit IRC04:00
*** tuanla____ has quit IRC04:00
*** tiendc has joined #openstack-nova04:01
*** tovin07 has joined #openstack-nova04:01
*** tuanla____ has joined #openstack-nova04:01
openstackgerritDeepak Mourya proposed openstack/nova master: Handle TZ change in iso8601 >=0.1.12
*** yingjun has quit IRC04:06
*** gongysh has quit IRC04:07
*** Tom-Tom_ has quit IRC04:10
*** jaianshu has joined #openstack-nova04:11
*** liverpooler has quit IRC04:11
*** lyan has quit IRC04:12
*** mdnadeem has quit IRC04:18
*** salv-orlando has joined #openstack-nova04:18
*** janki has joined #openstack-nova04:22
*** salv-orlando has quit IRC04:22
*** mdnadeem has joined #openstack-nova04:25
openstackgerritsean mooney proposed openstack/nova master: Change 'InstancePCIRequest' spec field
openstackgerritsean mooney proposed openstack/nova master: Add Neutron port capabilities to devspec in request
openstackgerritsean mooney proposed openstack/nova master: Format NIC features using os-traits definitions
*** dave-mccowan has quit IRC04:27
*** gongysh has joined #openstack-nova04:29
*** felipemonteiro__ has quit IRC04:31
*** tuanla____ has quit IRC04:33
*** tovin07 has quit IRC04:33
*** annp has quit IRC04:33
*** annp has joined #openstack-nova04:34
*** tuanla____ has joined #openstack-nova04:34
*** tovin07 has joined #openstack-nova04:34
*** Tom-Tom has joined #openstack-nova04:36
*** pramodrj07 has quit IRC04:39
*** Dinesh__Bhor has quit IRC04:40
*** Tom-Tom has quit IRC04:40
*** Dinesh__Bhor has joined #openstack-nova04:40
*** masber has joined #openstack-nova04:42
openstackgerritAlex Xu proposed openstack/nova master: placement: using the dict format for the allocation in claim_resources
openstackgerritAlex Xu proposed openstack/nova master: placement: enable required traits from the flavor extra specs
*** vladikr has quit IRC04:52
*** sridharg has joined #openstack-nova04:52
*** hongbin has quit IRC04:54
*** ratailor has joined #openstack-nova04:56
*** dikonoo has quit IRC05:12
*** blkart has joined #openstack-nova05:14
*** salv-orlando has joined #openstack-nova05:19
openstackgerritMerged openstack/nova master: Set server status to ERROR if rebuild failed
*** harlowja has joined #openstack-nova05:19
*** salv-orlando has quit IRC05:23
*** dikonoor has joined #openstack-nova05:29
*** links has joined #openstack-nova05:32
openstackgerritAlex Xu proposed openstack/nova master: placement: using the dict format for the allocation in claim_resources
openstackgerritAlex Xu proposed openstack/nova master: placement: enable required traits from the flavor extra specs
*** lajoskatona has joined #openstack-nova05:39
*** Sandy619 has joined #openstack-nova05:41
*** salv-orlando has joined #openstack-nova05:45
*** Sandy619 has quit IRC05:46
*** Tom-Tom has joined #openstack-nova05:48
*** chyka has joined #openstack-nova05:50
*** dikonoor has quit IRC05:51
*** dikonoor has joined #openstack-nova05:52
*** Guest62420 has quit IRC05:53
*** sree has joined #openstack-nova05:54
*** chyka has quit IRC05:55
*** bhujay has joined #openstack-nova05:58
*** clayton has quit IRC05:58
*** sree has quit IRC05:59
*** Dinesh__Bhor has quit IRC05:59
*** Eran_Kuris has joined #openstack-nova06:03
*** rcernin_ has joined #openstack-nova06:03
*** rcernin has quit IRC06:03
*** clayton has joined #openstack-nova06:04
*** dikonoor has quit IRC06:06
*** dikonoor has joined #openstack-nova06:06
*** armax has quit IRC06:07
*** Eran_Kuris has quit IRC06:08
*** sree has joined #openstack-nova06:08
*** Dinesh__Bhor has joined #openstack-nova06:08
*** Eran_Kuris has joined #openstack-nova06:10
*** sree has quit IRC06:12
*** fragatina has joined #openstack-nova06:13
*** fragatina has quit IRC06:14
*** acormier has joined #openstack-nova06:14
*** fragatina has joined #openstack-nova06:14
*** tbachman has joined #openstack-nova06:15
*** dikonoor has quit IRC06:15
*** tbachman has quit IRC06:15
ameedaMorning :)06:18
ameedacan you please check this code to detach volumes "" and let me know if that need to fix or anything else06:18
*** acormier has quit IRC06:18
*** tbachman has joined #openstack-nova06:18
*** gcb has quit IRC06:22
*** slaweq has joined #openstack-nova06:26
*** slaweq has quit IRC06:28
*** sridharg has quit IRC06:28
*** dikonoor has joined #openstack-nova06:30
*** threestrands has joined #openstack-nova06:39
*** slaweq has joined #openstack-nova06:44
*** Dinesh__Bhor has quit IRC06:46
*** slaweq has quit IRC06:49
*** blkart has quit IRC06:49
*** threestrands has quit IRC06:49
*** Dinesh__Bhor has joined #openstack-nova06:50
*** Dinesh__Bhor has quit IRC06:52
*** Dinesh__Bhor has joined #openstack-nova06:52
*** tbachman has quit IRC06:54
*** slaweq has joined #openstack-nova07:03
*** annp has quit IRC07:10
alex_xugibi: hi, I update the patch and addressed a upgrade case which pointed by Matt
*** pcaruana has joined #openstack-nova07:10
*** rcernin has joined #openstack-nova07:12
*** rcernin_ has quit IRC07:12
*** slaweq has quit IRC07:14
*** Dinesh__Bhor has quit IRC07:17
*** janki has quit IRC07:18
*** pcaruana has quit IRC07:21
*** pcaruana has joined #openstack-nova07:22
*** markvoelker has quit IRC07:25
*** markvoelker has joined #openstack-nova07:27
*** jdurgin has quit IRC07:28
*** edand__ has joined #openstack-nova07:29
*** markvoelker has quit IRC07:32
*** chyka has joined #openstack-nova07:39
*** slaweq has joined #openstack-nova07:40
openstackgerritMaciej Jozefczyk proposed openstack/nova master: Do not normalize allocation_ratios
openstackgerritMaciej Jozefczyk proposed openstack/nova master: Do not normalize allocation_ratios
*** chyka has quit IRC07:44
*** slaweq has quit IRC07:45
*** jdurgin has joined #openstack-nova07:45
openstackgerritJianghua Wang proposed openstack/nova master: XenAPI: provide VGPU_DISPLAY_HEAD inventory in compute node
openstackgerritMaciej Jozefczyk proposed openstack/nova master: Do not normalize allocation_ratios
*** itlinux has joined #openstack-nova07:56
*** brault has joined #openstack-nova08:01
*** pcaruana has quit IRC08:01
*** rgerganov|away is now known as rgerganov08:02
*** matrohon has joined #openstack-nova08:02
*** alexchadin has joined #openstack-nova08:10
*** sgordon has quit IRC08:17
kashyapdmsimard: Hey, I'm fully occupied with something more urgent until 04-Feb, I'm afraid.  But I've added that to my TODO queue.08:20
*** sahid has joined #openstack-nova08:20
*** ragiman has joined #openstack-nova08:20
*** amoralej|off is now known as amoralej08:22
*** sgordon has joined #openstack-nova08:23
*** AlexeyAbashkin has joined #openstack-nova08:23
*** itlinux has quit IRC08:24
*** lpetrut has joined #openstack-nova08:25
*** tesseract has joined #openstack-nova08:27
*** alexchadin has quit IRC08:28
*** alexchadin has joined #openstack-nova08:29
*** lpetrut has quit IRC08:32
*** mdnadeem has quit IRC08:36
*** alexchadin has quit IRC08:36
*** mdnadeem has joined #openstack-nova08:36
*** mdnadeem is now known as mdnadeem|lunch08:36
*** alexchadin has joined #openstack-nova08:36
*** lpetrut has joined #openstack-nova08:38
*** pcaruana has joined #openstack-nova08:39
*** alexchad_ has joined #openstack-nova08:40
*** diegows has quit IRC08:41
*** dikonoo has joined #openstack-nova08:41
*** dikonoor has quit IRC08:42
*** alexchadin has quit IRC08:42
*** rcernin has quit IRC08:44
*** jpena|off is now known as jpena08:45
*** diegows has joined #openstack-nova08:46
*** Guest50651 has quit IRC08:49
*** Guest50651 has joined #openstack-nova08:52
gibialex_xu: looking...08:52
alex_xugibi: thanks08:52
*** matrohon has quit IRC08:53
*** yamamoto has quit IRC08:53
*** tesseract has quit IRC08:54
ameedaplease please please, can anyone approve my gerrit here ?
*** jaianshu_ has joined #openstack-nova09:03
*** tesseract has joined #openstack-nova09:03
*** jaianshu has quit IRC09:05
*** harlowja has quit IRC09:07
*** lpetrut_ has joined #openstack-nova09:08
*** yamahata has quit IRC09:10
*** yamamoto has joined #openstack-nova09:10
*** lpetrut has quit IRC09:10
*** cdent has joined #openstack-nova09:11
*** itlinux has joined #openstack-nova09:12
*** purplerbot has quit IRC09:12
*** ilyashakhat has quit IRC09:15
*** alexchad_ is now known as alexchadin09:15
*** ilyashakhat has joined #openstack-nova09:21
mdboothstephenfin: You were holding off on a +2 for melwitt ?
*** psachin` has quit IRC09:24
*** lpetrut_ has quit IRC09:25
*** psachin has quit IRC09:25
*** purplerbot has joined #openstack-nova09:25
*** purplerbot has quit IRC09:26
*** purplerbot has joined #openstack-nova09:26
*** ralonsoh has joined #openstack-nova09:27
*** markvoelker has joined #openstack-nova09:28
*** psachin has joined #openstack-nova09:32
*** ratailor_ has joined #openstack-nova09:34
*** ratailor has quit IRC09:35
*** ragiman has quit IRC09:39
*** derekh has joined #openstack-nova09:39
*** stakeda has quit IRC09:39
*** sridharg has joined #openstack-nova09:49
*** yangyapeng has quit IRC09:50
*** sdague has joined #openstack-nova09:50
*** yangyapeng has joined #openstack-nova09:51
*** matrohon has joined #openstack-nova09:53
*** ragiman has joined #openstack-nova09:53
*** yangyapeng has quit IRC09:55
*** tuanla____ has quit IRC09:58
bauzasmorning folks09:58
maciejjozefczykbauzas: hiho09:58
bauzasmy presence is a bit off, given I'm in a conference09:58
bauzasjust ping me directly if you need me09:58
maciejjozefczykbauzas: good to know cause I have something for you09:59
*** yamamoto has quit IRC09:59
*** ralonsoh_ has joined #openstack-nova10:01
*** maciejjozefczyk is now known as maciejjozefczyk|10:01
*** maciejjozefczyk| is now known as maciejjozefczykb10:01
*** maciejjozefczykb is now known as maciejjozefczyk_10:01
*** markvoelker has quit IRC10:02
*** lajoskatona has quit IRC10:03
*** ralonsoh has quit IRC10:04
openstackgerritnalini proposed openstack/nova master: Modify show aggregate to display 'updated_at' value
*** avolkov has joined #openstack-nova10:11
*** slaweq has joined #openstack-nova10:11
*** jcosmao has left #openstack-nova10:14
*** acormier has joined #openstack-nova10:16
*** yamamoto has joined #openstack-nova10:16
*** jaosorior has quit IRC10:16
*** jcosmao has joined #openstack-nova10:17
*** jaosorior has joined #openstack-nova10:17
*** lajoskatona has joined #openstack-nova10:19
*** dtantsur|afk is now known as dtantsur10:19
*** acormier has quit IRC10:20
*** sambetts|afk is now known as sambetts10:20
*** dikonoo has quit IRC10:21
openstackgerritSylvain Bauza proposed openstack/nova master: Avoid suspending guest with attached vGPUs
*** Tom-Tom has quit IRC10:24
*** jaianshu__ has joined #openstack-nova10:29
*** jaianshu_ has quit IRC10:32
*** lucas-afk is now known as lucasagomes10:33
*** dikonoo has joined #openstack-nova10:33
*** abhishekk has quit IRC10:34
*** rcernin has joined #openstack-nova10:34
*** gszasz has joined #openstack-nova10:35
*** matrohon has quit IRC10:37
*** edand__ has quit IRC10:37
*** yangyapeng has joined #openstack-nova10:39
*** phuongnh has quit IRC10:40
*** yankcrime has quit IRC10:41
*** slaweq has quit IRC10:42
*** slaweq has joined #openstack-nova10:43
*** slaweq has quit IRC10:48
*** dikonoo has quit IRC10:48
*** _nick has joined #openstack-nova10:49
*** _nick is now known as yankcrime10:49
*** mdnadeem|lunch has quit IRC10:50
SpazmoticToo full.. will die10:53
Spazmoticor burp.. one of those things10:53
*** AlexeyAbashkin has quit IRC10:53
*** AlexeyAbashkin has joined #openstack-nova10:54
*** ragiman has quit IRC10:55
*** alexchadin has quit IRC10:56
*** yangyapeng has quit IRC10:58
*** markvoelker has joined #openstack-nova10:58
*** liuzz has quit IRC10:59
*** yangyapeng has joined #openstack-nova10:59
*** alexchadin has joined #openstack-nova10:59
*** gaoyan has joined #openstack-nova11:01
*** gaoyan has quit IRC11:06
*** gaoyan_ has joined #openstack-nova11:06
*** ragiman has joined #openstack-nova11:07
*** chyka has joined #openstack-nova11:15
*** gongysh has quit IRC11:16
*** dtantsur is now known as dtantsur|brb11:19
*** sapd_ has quit IRC11:19
*** sapd_ has joined #openstack-nova11:19
*** psachin has quit IRC11:19
*** chyka has quit IRC11:20
openstackgerritAlex Xu proposed openstack/nova master: placement: using the dict format for the allocation in claim_resources
openstackgerritAlex Xu proposed openstack/nova master: placement: enable required traits from the flavor extra specs
alex_xugibi: stephenfin, ^ sorry for I missed that two unittests, just fix them11:22
*** itlinux has quit IRC11:23
*** gaoyan_ has quit IRC11:23
*** gaoyan_ has joined #openstack-nova11:23
*** gaoyan_ has quit IRC11:23
gibialex_xu: no worries, I was pulled into something internally so I have to go back to your patches11:24
alex_xugibi: no problem, thanks for your time11:25
openstackgerritAlex Xu proposed openstack/nova master: Fix nits in support traits on allocation candidates API
*** gszasz has quit IRC11:29
*** markvoelker has quit IRC11:32
*** edand__ has joined #openstack-nova11:34
*** bhujay has quit IRC11:37
*** itlinux has joined #openstack-nova11:38
*** bhujay has joined #openstack-nova11:39
rgerganovcdent, the issue I had yesterday:
rgerganovcdent, you were right that the problem is in update_from_provider_tree :)11:44
*** mvenesio has joined #openstack-nova11:45
cdentrgerganov: ah, interesting, nice sleuthing. Was the trait failure because you were using a not-allowed trait?11:48
rgerganovyes  :)11:49
rgerganovspeaking of that, how do I add a new trait?11:49
cdentrgerganov: are you wanting to establish a new official trait, or set a CUSTOM trait in the local deployment?11:52
cdentif the latter:
cdentif the former:
openstackgerritStephen Finucane proposed openstack/nova master: Don't filter out sibling sets with one core
openstackgerritStephen Finucane proposed openstack/nova master: trivial: Update tests to set 'NUMACell.siblings'
openstackgerritStephen Finucane proposed openstack/nova master: Ensure emulator threads are always calculated
openstackgerritStephen Finucane proposed openstack/nova master: Always pass 'NUMACell.siblings' to _pack_instance_onto_cores'
stephenfinsahid: Fancy taking a look at those ^ ?11:53
rgerganovcdent, thanks, I believe it is the latter11:53
*** jaianshu__ has quit IRC11:56
*** dave-mccowan has joined #openstack-nova12:09
*** zhurong has joined #openstack-nova12:10
*** smatzek has joined #openstack-nova12:17
*** mvk has quit IRC12:19
*** slaweq has joined #openstack-nova12:24
*** ratailor_ has quit IRC12:28
*** slaweq has quit IRC12:29
*** lpetrut_ has joined #openstack-nova12:29
*** markvoelker has joined #openstack-nova12:29
*** mvk has joined #openstack-nova12:31
*** blkart has joined #openstack-nova12:34
*** jpena is now known as jpena|lunch12:36
*** stvnoyes has joined #openstack-nova12:36
*** gszasz has joined #openstack-nova12:39
*** lajoskatona has quit IRC12:55
*** mvk has quit IRC12:56
efried_back_wedrgerganov Haven't caught up fully; what news?12:57
*** efried_back_wed is now known as efried12:58
*** lajoskatona has joined #openstack-nova12:58
*** READ10 has joined #openstack-nova12:58
rgerganovefried, hi, I found why the local tree was missing the nested RP, see my comment here:
rgerganovefried, my driver was trying to create an incorrect trait and update_from_provider_tree silently removed the nested RP12:59
*** gszasz has quit IRC13:02
efriedrgerganov Hum, yeah, that code will remove the entire sub-branch from the tree (but only remove its root from the association cache - yet another leak in that cache, though it's resolved a couple patches up the series).13:02
*** markvoelker has quit IRC13:03
efriedrgerganov But that's *supposed* to be okay, because we should wind up restoring it next time we refresh_associations13:03
*** acormier has joined #openstack-nova13:04
rgerganovefried, this is not the case, we don't restore the nested RP and when we try to recreate it, we fail13:04
*** acormier has quit IRC13:05
*** acormier has joined #openstack-nova13:05
efriedrgerganov So the bug we need to track down is *why* we don't restore it in the cache.13:05
efriedrgerganov Oh, is all of this happening in a single invocation of update_from_provider_tree ?13:05
*** lpetrut_ has quit IRC13:05
rgerganovefried, yes13:06
*** lpetrut_ has joined #openstack-nova13:06
rgerganovefried, in one invocation we create the nested RP in placement and mess up the local tree13:06
rgerganovand on the invocation we try to create it again and then fail13:06
rgerganovand on the next invocation we try to create it again and then fail13:06
efriedI would have expected us to refresh at some point before that second invocation13:07
rgerganovthat's why I was thinking that is a good idea13:07
*** lpetrut_ has quit IRC13:07
rgerganoveven if the root node exists, refresh the tree before calling the virt driver for changes13:08
*** acormier_ has joined #openstack-nova13:08
*** lpetrut_ has joined #openstack-nova13:08
*** chyka has joined #openstack-nova13:08
efriedrgerganov As written, that change will *not* refresh the tree, though13:09
*** lpetrut_ has quit IRC13:09
efriedIf you want a simple try-out, change to force=True13:10
*** lpetrut_ has joined #openstack-nova13:10
rgerganovefried, this loop doesn't run at all: for u in self._provider_tree.get_provider_uuids(uuid):13:11
*** lpetrut_ has quit IRC13:11
rgerganovbecause there are no children in the local tree13:11
efriedrgerganov That call should at least return `uuid` itself.13:11
rgerganovah, correct13:11
*** lpetrut_ has joined #openstack-nova13:11
rgerganovmy bad13:11
efriedAnd then the _refresh_associations call should re-grab all the tree-associated providers13:12
efriedBut possibly only if the cache timeout has occurred.13:12
*** acormier has quit IRC13:12
efriedWhich may actually be the issue.13:12
efriedSo we'll refresh the root there, but then on the recursive part of the call, we'll check the cache timeout map13:12
efried...which will still have stale entries because we only removed the nodes from the ProviderTree; we didn't remove the cache timeout entry for the descendants.13:13
*** chyka has quit IRC13:13
efriedSo force=True as mentioned above would work; but the more correct solution would be to remove all those other association refresh times when we invalidate that root from the cache.13:14
rgerganovefried, do you want me to try it?13:14
efriedrgerganov If you've got a setup all ready to go, please do.  It would take me some time to get to that point.13:14
efriedrgerganov Meanwhile, let me come up with the other solution real quick...13:14
*** acormier_ has quit IRC13:15
*** acormier has joined #openstack-nova13:15
rgerganovefried, with force=True it refreshed the associations for the root provider but it didn't add the nested RP in the tree13:17
*** amoralej is now known as amoralej|lunch13:19
efriedrgerganov At a glance, I don't see (and I don't offhand remember) how tree-associated providers get refreshed.  While I poke at that, would you please try this:
*** liverpooler has joined #openstack-nova13:20
rgerganovefried, ok, trying13:21
efriedOkay, that's not gonna work either.13:21
efriedBut try it anyway, for grins.13:21
openstackgerritTakashi NATSUME proposed openstack/nova master: Transform instance.resize_prep notification
*** tiendc has quit IRC13:21
*** edmondsw has joined #openstack-nova13:21
*** edmondsw_ has joined #openstack-nova13:22
efriedrgerganov and @639 are the things that *should* be repopulating the _provider_tree from placement.13:24
rgerganovefried, AttributeError: 'SchedulerReportClient' object has no attribute 'association_refresh_time'13:25
efriedrgerganov Oh, are you sitting on the very top of the series?13:25
efriedOr did I forget the underscore :)13:25
*** lajoskatona has quit IRC13:25
*** edmondsw has quit IRC13:26
openstackgerritGao Fei proposed openstack/nova master: Update VMWare vSphere link address
rgerganovefried, my patches are based on top of "Move refresh time from report client to prov tree"13:27
gibiis it only me or gerrit is slow like hell today13:27
rgerganovgibi, yes, it's really slow13:28
efriedgibi Not just you.  Slow as hell.  Started grinding yesterday.13:28
efriedMaybe ask -infra to bounce it.13:28
gibithanks for the info13:28
efriedrgerganov Okay, that patch got rid of _association_refresh_time in the report client, and would have made my suggested fix moot anyway.13:28
efriedrgerganov So what I can't figure out is why, in _ensure_resource_provider, the _get_providers_in_tree => populate_from_iterable isn't restoring the descendants to the cache.13:29
efriedAre you pdb'ing?13:30
rgerganovefried, yes13:30
*** lpetrut_ has quit IRC13:30
efriedCan you break at at make sure rps_to_refresh contains all the descendants?13:30
rgerganovwith or without my patch?13:31
efriedShouldn't matter.13:31
*** alexchadin has quit IRC13:31
rgerganovok, tracing13:31
efriedBecause as it turns out, _refresh_associations isn't where we pick up tree-associated providers.  It's in fact in this chunk of _ensure_resource_provider.13:32
*** sean-k-mooney has quit IRC13:32
rgerganovefried, it does matter because without my patch we don't get there13:33
efriedOh, because 'return uuid'...13:34
efried...but that .exists(uuid) should be False!13:34
efriedBecause we removed that guy.13:34
*** Tom-Tom has joined #openstack-nova13:34
rgerganovno, we pass the parent uuid here13:34
rgerganovand it is there13:34
artomAm I doing this right?
rgerganovefried, _ensure_resource_provider is called with the uuid of the root RP13:35
artomIt says 0 hits, but we clearly got some here:
efriedrgerganov I see, and you were trying to add the bogus trait to the child?13:35
rgerganovefried, correct13:36
efriedSo we invalidated the child, but not the root.13:36
efriedOkay, it's coming together.13:36
rgerganovcould you please tell why my patch won't refresh the tree?13:36
efriedrgerganov No I can't.  It should.13:37
*** jpena|lunch is now known as jpena13:38
*** Tom-Tom has quit IRC13:39
*** bhujay has quit IRC13:39
*** alexchadin has joined #openstack-nova13:40
*** Tom-Tom has joined #openstack-nova13:40
*** alexchadin has quit IRC13:41
efriedrgerganov So with your patch, is rps_to_refresh populated at that breakpoint?13:42
rgerganovefried, yes13:43
efriedrgerganov And after populate_from_iterable, is the _provider_tree populated properly?13:43
rgerganovit contains both parent and child13:43
rgerganovas far as I can see, yes13:44
mvenesioHi guys i'm trying to set nova to use SSL for the database connection, i set the mysql connection as well as i set it for the rest of the projects like cinder and glance, but for nova it does not work and i got an OperationalError. Any idea about how to set it right ?13:44
efriedmvenesio I think that's a better question for #openstack (see channel topic).13:45
efriedrgerganov So if the child is in the cache... where do we run into a problem?13:46
openstackgerritLee Yarwood proposed openstack/nova master: rbd: flatten images when creating/unshelving an instance
*** pchavva has joined #openstack-nova13:46
*** ragiman has quit IRC13:46
lyarwoodmelwitt: ^ I wonder if this is something we could land by the rc13:46
*** AlexeyAbashkin has quit IRC13:47
mvenesioefried: ok i'll do it, thanks13:47
rgerganovefried, the child gets in the cache with my patch and then we don't have a problem (at least not now :) )13:47
*** acormier has quit IRC13:47
*** rcernin has quit IRC13:47
efriedrgerganov I must have misunderstood "could you please tell why my patch won't refresh the tree?"13:48
*** acormier has joined #openstack-nova13:48
rgerganovefried, I asked that because you said "As written, that change will *not* refresh the tree, though"13:49
efriedrgerganov Okay, yeah, I was wrong.  I had missed the 'return uuid' part and was misunderstanding my own code :(13:50
efriedrgerganov So your solution will work, but it's skipping a nontrivial optimization that I would like to keep if possible.13:51
efriedrgerganov One possible alternative would be for that cache invalidation to in fact kill the whole tree.  But the implications of that are probably too far-reaching to be practical.13:52
*** acormier has quit IRC13:52
efriedrgerganov Other than that patch, do you have any local code on the series?13:52
efriedrgerganov In particular, we should definitely commit whatever test case you're running to hit this problem.13:53
*** yamamoto has quit IRC13:53
openstackgerritMerged openstack/nova master: Transform instance.resize_confirm notification
efriedrgerganov Though I would like to do it as part of the update_from_provider_tree patch.13:53
rgerganovefried, so basically the test case should be virt driver adding an incorrect trait13:54
efriedrgerganov a child13:54
*** dtantsur|brb is now known as dtantsur13:54
efriedrgerganov How close to your EOD are you?13:55
rgerganovefried, I will head out in 2 hours13:55
*** mvk has joined #openstack-nova13:56
*** esberglu has joined #openstack-nova13:56
*** esberglu has quit IRC13:56
efriedrgerganov So I've *almost* got this code path in my test_report_client work in progress.13:57
*** vladikr has joined #openstack-nova13:57
*** alexchadin has joined #openstack-nova13:57
efriedrgerganov I've got one piece setting a bogus trait on the root.  And another setting inventory in a bogus *resource class* on a descendant.  The latter of which should *probably* have the same effect.13:57
efriedrgerganov But since we've identified this exact issue, it wouldn't go amiss to have a small, isolated test case that's separate from that big one.13:58
efriedrgerganov Do you have the time/inclination/know-how to write that test case before you leave?13:58
*** eharney has quit IRC13:58
rgerganovefried, I will give it a try13:58
efriedrgerganov Thanks!13:59
rgerganovefried, do you want me to update an existing patch or start a new patch?13:59
*** lpetrut_ has joined #openstack-nova13:59
*** markvoelker has joined #openstack-nova14:00
*** weshay|rover is now known as weshay|mtg14:00
*** AlexeyAbashkin has joined #openstack-nova14:00
mriedemalex_xu: i'm fine with checking the version in the Selection object or just checking if the allocation request is the list or dict format and adjusting properly, up to you, but i think we have to handle both cases in queens, we can then probably remove that check later in rocky14:00
artommriedem, 'morning - did we figure out if the libvirt connection reset errors were a real bug or not?14:02
efriedrgerganov New one.  We can always slot it into the series, or squash it into the existing commit.14:02
efriedrgerganov And that way I can continue working on these test cases locally in parallel.14:02
artomAlso - I think I'm using logstash wrong - is turning up nothing for the past month14:03
rgerganovefried, ok cool14:03
bauzasmriedem: when you said "sneaky" in, you meant okay for that or not ?14:03
Roamer`actually, yeah, I've been meaning to ask - is there some documentation on using logstash somewhere?  I've seen people compose nice queries, like "a message that looks almost like this in this set of files", but I'd like to know more :)14:04
mriedemartom: message:"Connection reset by peer" AND tags:"screen-n-cpu.txt"14:04
mriedembauzas: it's ok14:04
mriedemjust sneaky14:04
Roamer`ah... that's more or less exactly it14:04
artommriedem, cheers :)14:04
*** jaosorior has quit IRC14:05
mriedemRoamer`: artom:
Roamer`mriedem, thanks!14:05
alex_xumriedem: for handle both cases in queens, do you means check both the list or dict and version at sametime?14:05
alex_xumriedem: for checking version, i have done today, it looks like this
mriedemalex_xu: we do'nt need to check the version and the type, just one or the other14:05
Roamer`mriedem, and thanks again for the +2 yesterday; unfortunately 140733 has had a bad case of "the same spurious totally unrelated test failure showing in a different job on every recheck" all day today :(14:06
alex_xumriedem: ok, I done that, I choice checking the version14:06
*** krtaylor has quit IRC14:07
*** krtaylor has joined #openstack-nova14:08
artomHrmm, so it started all of a sudden on Jan 16th14:09
bauzasI'm around for reviewing14:09
bauzasmriedem: which priority changes should I be doing ?14:09
mriedemartom: keep in mind that logstash only holds 10 days worth of logs, so that's getting close to the cutoff14:10
bauzasnested RPs ?14:10
*** yamamoto has joined #openstack-nova14:10
mriedembauzas: actually i'd really like to get a few more osc-placement changes merged before we do the first 1.0.0 release this week,14:10
mriedemi've got a +2 on a change here
mriedemand there is an easy cleanup and docs series starting here
mriedemstephenfin: ^14:10
mriedemthis change had a +2 from jaypipes before a rebase
stephenfinmriedem: Sure, I can take a look14:11
mriedembauzas: and this is an easy libvirt volume driver add
stephenfinSpeaking of jaypipes, wonder where he's at. I'd like some eyes on
mriedem"work conference" i was told14:12
stephenfinVegas it is14:12
mriedemheh maybe14:12
bauzasokay looking14:13
*** sapd has joined #openstack-nova14:13
bauzasFWIW, I'm also in a conference now for the next 3 days14:13
bauzasorganizing it, so I have time :p14:13
*** amoralej|lunch is now known as amoralej14:14
*** lyan has joined #openstack-nova14:14
*** yamahata has joined #openstack-nova14:14
*** links has quit IRC14:15
*** avolkov has quit IRC14:15
mriedemdidn't that ski conference already happen a couple of weeks ago?14:16
mriedemis that a monthly conference?14:16
*** hshiina has quit IRC14:17
*** smatzek has quit IRC14:17
*** esberglu has joined #openstack-nova14:17
*** zhurong has quit IRC14:18
artommriedem, ah - heh, it's kinda misleading that you can search for a month back though14:18
artomWell, 16th is 8 days14:19
mriedemartom: fwiw, some other stuff start randomly exploding around the 16th too in the ironic multinode grenade job14:20
edleafealex_xu: commented on
mriedemjroll was investigating that - random segfaults14:20
artommriedem, ah, interesting14:20
edleafealex_xu: I am concerned that the change you made could break if placement changes in the future14:20
bauzasmriedem: huh, unfortunately not, only every year ;)14:21
bauzaswe need snow14:21
stephenfinmriedem: Done. Only have comments for
*** acormier has joined #openstack-nova14:23
mriedemstephenfin: thanks14:24
*** yamahata has quit IRC14:24
alex_xuedleafe: if we change the allocation_request format in the future, we must do another patch just like 536083. that parameter 'version=allocation_request_version' can't do a magic let us upgrade to the new format14:24
alex_xuedleafe: we can say: the reason for including the allocation_request_version in the selection obj is so that claim_resources can know how to handle alloc_reqs.14:25
mriedem"the reason for including the allocation_request_version in the selection obj is so that claim_resources can know how to handle alloc_reqs." - not exactly, but that is a side effect14:26
*** markvoelker has quit IRC14:27
mriedemsince we have client side code that needs to know what format the thing is in14:27
*** markvoelker has joined #openstack-nova14:27
mriedemthe point of the version in the selection object, though, is so the client on a reschedule makes the same PUT /allocations request in the version/format that the scheduler initially created the allocation request (from GET /allocation_candidates)14:27
*** crushil has joined #openstack-nova14:27
edleafealex_xu: if placement is upgraded to a new version that changes the AR format, the way you changed it will force it to be posted to placement as 1.12, which would be wrong14:30
mriedemalex_xu: edleafe: "if we change the allocation_request format in the future" - if that happens, i think we'll have to add an AllocationRequest versioned object to nova to deal with the version differences getting passed over rpc14:30
alex_xumriedem: edleafe yes, but I don't want to implement the claim_resources method and the sub method to support two version format, that makes the code hard to read, I want to convert the format to consistent in the begining of claim_resources method14:30
mriedemedleafe: if placement is upgraded to a new version, it shouldn't affect the client side code since the client side code is requesting a specific microversion14:30
mriedemwhich shouldn't change14:30
edleafemriedem: allocation requests are *supposed* to be opaque14:30
edleafemriedem: we are violating that here to fix a bug14:30
edleafeOnce we are in Rocky, the need for this modification goes away14:31
mriedemi agree once we are in rocky this can go away,14:31
mriedemthe modification here is for upgrades14:31
edleafeand all this doubling code can be removed14:31
edleafeI understand14:31
alex_xuedleafe: when placement upgrade, our client won't use the lastest version, we have specified the version in the client
edleafeI just don't want all requests to be posted at 1.1214:32
*** jackie-truong has joined #openstack-nova14:32
edleafealex_xu: the client should pass the AR and AR_version without inspecting the contents14:32
alex_xuand that is the rule of using micorverion in the client, never use the latest version, and specified a version explicitly14:32
mriedemedleafe: you said, "it would be better if you also modified the allocation_request_version  to 1.12 when you modify the allocation_request in the block starting on  L1163." - if you change allocation_request_version='1.12' anywhere it's going to post all requests at 1.12 regardless14:33
mriedemwe have to inspect the contents in this case14:33
*** AlexeyAbashkin has quit IRC14:33
mriedemand to do that, we need to know what format it's in14:33
openstackgerritAmeed Ashour proposed openstack/nova master: detach instance volumes when VM creation fails
edleafemriedem: no, that would be inside the 'if' block on L1162, so only <1.12 would be affected14:34
mriedemedleafe: true, like i said in the comment, i'm fine with that14:34
mriedemand i think makes sense14:34
edleafemriedem: if it's >=1.12, then it won't get changed14:34
mriedemsure i'm ok with that14:34
mriedemalex_xu: ^ want to just make that change?14:34
*** lucasagomes is now known as lucas-hungry14:35
alex_xumriedem: make the version=allocation_request_version?14:35
mriedemif the version is < 1.12 and you modify ar, then set allocation_request_version='1.12'14:35
alex_xumriedem: ok, no problem14:35
mriedemand use allocation_request_version as before when claim_resources does it's PUT reuest14:35
*** avolkov has joined #openstack-nova14:36
alex_xumriedem: edleafe got the point, will update soon, thanks14:36
*** gszasz has joined #openstack-nova14:36
edleafealex_xu: thanks14:37
edleafealex_xu: and don't forget my nit on while you're at it :)14:37
alex_xuedleafe: yes sir!14:38
*** acormier has quit IRC14:39
edleafealex_xu: :)14:39
mriedemartom: jroll: one thing i was wondering was if there was a new package version of something in the Pike UCA around 1/1614:39
mriedemi'm not sure if there is a package change log somewhere for the pike UCA though14:39
*** Eran_Kuris has quit IRC14:40
*** crushil has quit IRC14:42
mriedemcoreycb: ^?14:42
*** cleong has joined #openstack-nova14:42
coreycbmriedem: artom: i can check. nova package right?14:43
mriedemcoreycb: no, just looking for a changelog for the pike cloud archive14:43
mriedemCI results started going wonky since ~1/1614:44
*** eharney has joined #openstack-nova14:44
mriedemso wondering about changes to distro packages for things like qemu/libvirt/httpd, et14:44
coreycbmriedem: i don't know if anything is available externally but i can at least check dates internally14:44
*** zhaochao has quit IRC14:45
*** Tom-Tom has quit IRC14:47
*** jobewan has joined #openstack-nova14:47
openstackgerritRadoslav Gerganov proposed openstack/nova master: Add update_from_provider_tree() negative test
rgerganovefried, ^^^14:48
efriedrgerganov Looking (if gerrit will ever load)14:48
coreycbmriedem: for pike specifically we haven't had anything go into pike-updates since december. it's possible though that something in the base xenial packages changed though.14:49
coreycbmriedem: for base xenial packages you can find dates on Launchpad if you know the package name, ie.
mriedemok was trying to find just a global list like in but no dice14:51
*** slaweq has joined #openstack-nova14:51
ameedajaypipes: please don't forget to check this for me14:52
coreycbmriedem: yeah unfortunately i don't think there's a list for stable updates14:52
mriedemdon't see any changes in 2018 to libvirt or qemu though so it's not that14:52
mriedemthanks anyway14:52
mriedemOH SNAP14:53
mriedemjroll: ^14:53
mriedemreleased 1/1814:54
*** avolkov has quit IRC14:54
*** acormier has joined #openstack-nova14:54
*** markvoelker has quit IRC14:55
mriedemshould compare the versions of the python2.7 package in the failing ironic CI jobs14:55
*** slaweq has quit IRC14:56
mriedembauzas: thanks for hitting those osc-placement changes14:57
*** tbachman_ has joined #openstack-nova14:58
*** markvoelker has joined #openstack-nova14:59
*** awaugama has joined #openstack-nova14:59
*** sridharg has quit IRC14:59
efriedrgerganov So as written, this test will fail?14:59
rgerganovefried, yes14:59
*** AlexeyAbashkin has joined #openstack-nova15:00
efriedrgerganov Good deal.  If you don't mind, as I work through this today, I think I'm going to incorporate your test and (some version of) your fix into the update_from_provider_tree patch.  You'll get co-author credit, of course :)15:01
*** edand__ has quit IRC15:01
*** tbachman_ is now known as tbachman15:01
rgerganovefried, fine with me :)15:02
cdentefried: I tried to do another recheck run through your changes this morning (and some others) but I'm not sure how much impact it had15:04
*** acormier has quit IRC15:05
efriedcdent Seen and appreciated sir.  Silver lining: it takes *some* of the pressure off getting the top of the series perfect.15:05
*** acormier has joined #openstack-nova15:06
efriedAt this rate, even things +W'd by FF will take a week to merge.15:06
*** kwathore has quit IRC15:07
*** slaweq has joined #openstack-nova15:07
*** Nil_ has joined #openstack-nova15:08
*** kwathore has joined #openstack-nova15:08
*** felipemonteiro has joined #openstack-nova15:08
openstackgerritAlex Xu proposed openstack/nova master: placement: using the dict format for the allocation in claim_resources
openstackgerritAlex Xu proposed openstack/nova master: placement: enable required traits from the flavor extra specs
openstackgerritAlex Xu proposed openstack/nova master: Fix nits in support traits on allocation candidates API
*** amodi has joined #openstack-nova15:09
alex_xuedleafe: mriedem stephenfin, thanks for the review, ^ addressed all the comments15:09
*** felipemonteiro_ has joined #openstack-nova15:10
*** yangyapeng has quit IRC15:11
*** avolkov has joined #openstack-nova15:11
*** armax has joined #openstack-nova15:11
edleafealex_xu: thanks - will review soon15:11
alex_xuedleafe: thanks15:12
*** yangyapeng has joined #openstack-nova15:12
mriedemalex_xu: were you going to put up a patch for the placement api-ref docs changes?
efriedcdent mriedem Sorry for not understanding how placement version lockstepping works; are we now ready to address these comments?
*** smatzek has joined #openstack-nova15:13
*** rgerganov has quit IRC15:13
*** felipemonteiro has quit IRC15:13
cdentthat may be out of date by now, since it's been sitting there for quite a while15:14
efriedcdent Thanks, at least it's on the radar.15:14
* efried adds self15:14
mriedemefried: yeah we require >=1.14 now really, or that's what nova-status is checking for and the compute requires for root providers,15:15
mriedemefried: but at this point might as well hold that off until rocky - the cleanup i mean15:15
mriedemdansmith: that reminds me - we havent bumped major compute rpc api versions in a few years; that's historically something you've been the master of, are you interested in doing one of those, since i think it has to happen shortly before RC1 yeah?15:16
*** itlinux has quit IRC15:16
mriedemwe have a lot of "drop this compat code when we bump major versions"15:16
*** gszasz has quit IRC15:16
efriedokay.  I must also not have a good handle on what kind of patch will be accepted after FF.  This isn't a feature, so...?15:17
*** yangyapeng has quit IRC15:17
*** lucas-hungry is now known as lucas-brb15:17
dansmithmriedem: yeah I should probably do that15:17
*** itlinux has joined #openstack-nova15:18
*** edand__ has joined #openstack-nova15:19
*** alexchadin has quit IRC15:19
mriedemRoamer`: looking at the latest cells job failure on your patch
mriedemit's timing out waiting for a volume backup to complete,15:20
mriedemlooking at the node it's running on:
mriedemprovider: rax-dfw15:20
mriedemthose have been really really slow since the intel kernel patcharoo15:20
*** dave-mccowan has quit IRC15:22
openstackgerritMerged openstack/osc-placement master: CLI for aggregates (v1.1)
*** stvnoyes has left #openstack-nova15:23
openstackgerritStephen Finucane proposed openstack/nova master: objects: Remove 'NUMATopologyLimits.obj_from_db_obj'
openstackgerritStephen Finucane proposed openstack/nova master: objects: Remove legacy '_to_dict' functions
openstackgerritStephen Finucane proposed openstack/nova master: objects: Remove legacy '_from_dict' functions
openstackgerritTakashi NATSUME proposed openstack/nova master: Transform instance.resize_prep notification
-openstackstatus- NOTICE: gerrit has been suffering from a full disk, some mails may have been lost in the last couple of hours. we will now restart gerrit to address ongoing slowness, too15:24
*** tidwellr has joined #openstack-nova15:26
*** slaweq has quit IRC15:26
edmondsw_gibi here's your requested daily reminder to look at if you get a chance :)15:27
*** bhujay has joined #openstack-nova15:28
*** mvenesio has quit IRC15:29
*** Guest28399 is now known as mgagne15:33
*** mgagne has joined #openstack-nova15:33
gibiedmondsw_: thanks :)15:33
efriedcdent edleafe Not loving the fact that DELETE RP inventory returns 204 and the client code is assuming the RP generation can just be incremented by one.15:34
*** edmondsw_ is now known as edmondsw15:34
cdentif you've deleted an RP why would you need to know it's generation?15:34
cdentit's gone15:34
openstackgerritMerged openstack/osc-placement master: Add missing runtime requirements
*** hongbin has joined #openstack-nova15:34
efriedcdent Deleted its inventory, not the RP itself.15:35
edleafeefried: what would you love more?15:35
cdentthen GET the /rp/{uuid}, that's what it's there for15:35
cdentthis notion that it is bad to make requests is unproven. just make requests15:36
efriededleafe Either DELETE returns a payload, which would look like { 'resource_provider_generation': N, 'inventories': {} }, or we always use PUT with {}15:36
efried...which does that15:36
efriedcdent It's not quite that simple.15:37
openstackgerritMerged openstack/osc-placement master: Address comments from original inventory patch
efriedcdent The point of getting the generation in the response is that I know it's the generation right after the inventory was deleted.15:37
openstackgerritMatt Riedemann proposed openstack/osc-placement master: Usage docs and initial release note for osc-placement
mriedemstephenfin: done ^15:37
cdentIf you are depenendent on that, you're not rally making a very resilient system?15:37
efriedcdent Whereas if I do a subsequent GET, I don't know if an intervening operation incremented the generation as well.15:37
cdentyes that's _always_ the case15:37
cdentwhether you did the the GET or not15:38
efriedcdent No, not when I get the generation in the response.15:38
cdentyou have no idea, ever, what any other client or thread might be doing with the same resource provider15:38
Roamer`mriedem, yes, I saw it was rax-dfw on the previous runs, too, that's why I haven't complained to -infra (I guessed somebody else with more authority would complain again sooner or later), but thanks for raising it in -infra right now15:38
cdentwhen you get the generation in the response, something else, immediately, might have changed it15:38
cdentthat's always true15:38
cdentwhat ever mechanism you are using to track generations15:38
efriedcdent Yes, but I don't need to care about that until the next time I go to update something.15:38
cdentwhich should always be the case15:39
cdentyou should always be prepared to get a 40915:39
cdentand act accordingly15:39
edleafegotta agree - I don't see what getting the generation in the response buys you15:39
efriedI need to be able to get a 409 reliably, though.15:39
efriedLook look.15:39
cdentthe generation should be utterly meaningless15:39
cdentsorry, I gotta go fetch the cat, will catch up in 30'15:40
efriedThat's exactly why I shouldn't be assuming it gets incremented by 115:40
efried(the meaningless thing, not the cat thing)15:40
*** itlinux has quit IRC15:40
edleafeif the delete succeeds, then it will be incremented by 1. You can certainly assume that15:40
mriedemRoamer`: see -cinder15:40
Roamer`mriedem, yes, looking15:40
efriededleafe No way.15:40
Roamer`I mean watching15:40
mriedemlooks like cinder made a change in the last 24 hours that is slowing down volume backup15:40
edleafeWhat you can't assume is that the next time you do something, it won't have been changed by something else15:41
efriededleafe That's not part of the contract.  It just happens to be the way placement is implemented.15:41
efriedThis is exactly my point15:41
Roamer`mriedem, ah, you even figured out which one!  I just mentioned it a couple of times there, but it seemed nobody was awake yet...15:41
*** bhujay has quit IRC15:41
efriedIt is crucial that the generation I have in my local cache matches the data I have in my local cache.15:41
edleafeefried: do you mean if generation was a hash of the current state, you wouldn't be able to make any assumption?15:42
efriededleafe Or any other implementation that's not a monotonically increasing integer.15:42
efriededleafe I mean that the format of the generation is not part of the API contract at all, and the client side should be assuming absolutely nothing about it.15:42
edleafeSo what would getting the generation buy you? It *still* could have been changed by the time you go to use it15:43
efriedYes, in which case I will legitimately get a 409 and need to refresh/redrive my change.  That's the good path.15:43
efriedThe bad path is this:15:43
*** mlavalle has joined #openstack-nova15:44
efriedI have generation 5.  I do my DELETE.  Then you do a GET (so you have generation 5) and a PUT to create some inventory.  You're now at gen 6.  Now I do my GET as cdent suggested and find out the gen is 6.  So I think gen 6 corresponds with empty inventory when in fact it does not.  Now I decide to add some inventory back.  I do a PUT with that new inventory and gen 6, which will *succeed* because the server has gen 6.  But15:45
efriedI just blew away your inventory update.15:45
mriedemRoamer`: for rechecks i guess15:46
openstackLaunchpad bug 1745168 in Cinder "volume backup tests timing out since 1/23" [Undecided,New]15:46
mriedemmelwitt: dansmith: et al ^ if you see gate timeouts15:46
*** dave-mccowan has joined #openstack-nova15:46
openstackgerritStephen Finucane proposed openstack/nova master: objects: remove pagesize from __init__ of InstanceNUMATopology
efriededleafe Great, it's defined as an integer, in the schema, for the current microversion.  There's still nothing in the contract that tells me it's monotonically increasing starting from zero.15:47
openstackgerritStephen Finucane proposed openstack/nova master: objects: remove related pinning from __init__ of InstanceNUMATopology
openstackgerritStephen Finucane proposed openstack/nova master: objects: remove cpuset_reserved from __init__ of InstanceNUMATopology
edleafeefried: "Then you do a GET (so you have generation 5)" - that's not correct. You'd get 6 in that case, since the DELETE succeeded15:47
Roamer`mriedem, thanks, I'll use it if needed... let's hope it's not needed much longer, because there's not much time left till... what, tomorrow? :)15:48
efriededleafe Sorry, right, add 1 to all the gen numbers after the first one.15:48
efriededleafe Or whatever.  The logic is the same.15:48
mriedemRoamer`: eod tomorrow yeah15:48
edleafeefried: so you'd think you're at 6, but meanwhile someone did an update to move it to 7. You then PUT the inventory at 6 (since you don't know it's been changed), and get a 40915:48
edleafeNothing blown away15:49
Roamer`mriedem, aaaaand if failed again.... and I can't do a recheck before it finishes failing...15:49
efriededleafe No, I don't think I'm at 6 unless I assume the generation increments by 115:49
efriededleafe I'll think I'm at 7 because I did my GET (to discover the "correct" generation) after you did your PUT.15:49
dansmithbauzas: are you around today to address my minor comments here?
dansmithbauzas: if not I can do it, but I'd like to retain my impartiality to be able to vote on it :)15:49
edleafeefried: Huh? I thought you said "Then you do a GET" after the delete15:50
mriedemRoamer`: i'll recheck it throughout the day15:50
edleafeWhere are you assuming anything15:50
*** mriedem has quit IRC15:50
*** david-lyle has joined #openstack-nova15:50
efriededleafe Just so.15:50
Roamer`mriedem, so yeah, I'll keep on rechecking... oh?  You will?  Thanks a lot!  (again... it seems that I'm thanking you on every other line...)15:50
*** bhujay has joined #openstack-nova15:51
efriededleafe If I'm not assuming the generation increments by 1, and I have to do a GET to discover what the generation is after I do my DELETE, that's broken.15:51
*** mriedem has joined #openstack-nova15:51
edleafeefried: I still don't see the issue15:51
*** gszasz has joined #openstack-nova15:51
efriedHere it is again with the correct generations:15:52
edleafeefried: you can make the assumption, and if something else changes it, you get a 409 and *then* do the GET to see what the current state is15:52
efriedI have generation 5.  I do my DELETE.  Server has gen 6, I have gen 5 still.  Then you do a GET (so you have generation 6) and a PUT to create some inventory.  You and server are now at gen 7.  Now I do my GET and find out the gen is 7.  So I think gen 7 corresponds with empty inventory when in fact it does not.  Now I decide to add some inventory back.  I do a PUT with that new inventory and gen 7, which will *succeed* bec15:52
efriedause the server has gen 7.  But I just blew away your inventory update.15:52
efriededleafe Perhaps the disconnect is where I'm doing that GET to discover the generation.  At that point I'm only doing a GET /resource_provider/{uuid} - I really *don't* want to go getting its traits, aggs, and inventories as well.15:53
edleafeefried: if you do a GET and ignore that there is now inventory and then proceed to blow it away, that's not placement's fault15:53
*** yangyapeng has joined #openstack-nova15:53
efriedThen there's no point having a cache.15:54
efriedWe have this cache that we're using successfully for most of the operations, but if we run any DELETE API, we have to blow it away and start over?  That seems not okay.15:54
edleafeefried: especially if you have two different systems claiming to be authoritative on what the inventory for an RP is15:54
*** jose-phillips has quit IRC15:54
*** jose-phi_ has joined #openstack-nova15:55
efriedThat's the entire reason we have this generation business in the first place.  To manage sync issues when more than one thread claims authority over a provider.15:55
efriedAnd I'm not saying it's placement's fault that this happens.  It's a matter of cache management by the client.  Which IMO we're not doing correctly in this scenario.15:56
edleafeefried: yes, but it is to prevent *accidental* conflicts. If you know you have a conflict and proceed ahead anyway, there isn't a system that will make that OK15:56
efriedBut I don't know I have a conflict.15:57
efriedPut another way, a client shouldn't use DELETE for inventories (or traits or aggs if it comes to that) if there's no response containing the generation.15:57
edleafeefried: your cache has RP, generation, and inventory. You go to update that and get a 409. It is now important that you invalidate the cache before you proceed15:57
efriedI don't get a 409.  That's the crux of the problem.15:58
edleafeefried: you don't do the GET *before* trying to update15:58
efriedHow else do I discover the "new" generation?15:59
efried(the one I think is post-DELETE, but which is in fact post-my-DELETE-and-your-PUT)15:59
edleafeYou update based on what the cache says. If the cache is current, the generation it has will work. If something else has changed in placement, you invalidate the cache, and do a GET to refresh it15:59
efriedand how do I know "something else has changed in placement"?15:59
edleafethat's where the assumption that after the delete, the generation is original gen + 116:00
efriedRight.  Which works, today, because we're dirtily exploiting what we know to be the internals of the placement service.16:00
edleafe"dirtily" - heh16:01
edleafeit's documented16:01
edleafeGive me a second to write up the flow as I understand it.16:01
efriededleafe Where is it documented?16:01
*** bhujay has quit IRC16:02
efriededleafe In the API ref, all we get is: "A consistent view marker that assists with the management of concurrent resource provider updates." and "409 Conflict if the resource_provider_generation doesn’t match with the server side." <== note "match"16:04
efriedAnd it's not mentioned in the overview doc ( at all16:05
*** bhujay has joined #openstack-nova16:05
*** AlexeyAbashkin has quit IRC16:05
edleafeefried: I'll have to do some doc digging, but I'm sure it's there16:05
edleafeefried: here's the flow I see for your inventory delete/update:16:06
edleafeYou have a cached RP with generation 5. You call DELETE for the inventory, and it succeeds. You now update your cache with generation 6, and no inventory.16:06
edleafeYou now want to PUT some inventory. So you send the PUT with the data and generation 6. Something else has changed the RP in the meantime, so your PUT returns 409. At that point, your cache is not accurate. You don't know what has changed, so you get rid of anything for that RP, and call GET with the rp_uuid to refresh everything in your cache for that RP. You then compare the inventory you16:06
edleafeexpected with what is there (which you would have done before the original PUT), and if it is still correct to update inventory, you re-PUT the inventory with the new generation. Rinse and repeat.16:06
efriededleafe That's a good flow, but it's not the one I described.16:06
*** bhujay has quit IRC16:07
efriededleafe But before we go on, how does "You now update your cache with generation 6" happen?16:07
*** bhujay has joined #openstack-nova16:07
edleafethat was the assume that if your DELETE succeeds at gen5, it is now at gen616:07
efriedAssuming gen+1?16:07
melwittmriedem: ack16:08
edleafeefried: yes.16:08
efriedOkay, then I agree it works, but is illegal API consumption.  Because the API doesn't tell us we can assuming monotonically increasing generation counter per operation.16:08
*** bhujay has quit IRC16:09
edleafeefried: if it doesn't, then that's a doc bug16:09
*** bhujay has joined #openstack-nova16:09
efriedThis is how we do it today.  Which is (part of) the reason we don't have a "bug" per se.  (The other part is that we don't actually have any paths that do concurrent updates yet.)16:09
efriededleafe I don't agree with that.16:10
edleafeefried: let's get jaypipes and cdent to weigh in on this16:11
efriededleafe But maybe.  We should find out what Jay thinks.  If we add words to those DELETE 204s along the lines of, "You may now assume the resource provider generation has been incremented by 1," I'll stfu.16:11
efriedyeah, what you said.16:11
edleafeI think we understand each other, and are not in agreement16:11
efriedDoes dansmith have skin in this game?16:12
edleafeI think dansmith is pretty busy with other things :)16:12
edleafeefried: the assumption is that after anything that modifies anything about a RP, the generation is +1 from what it was16:13
edleafenot just DELETE16:13
dansmithclient-side ever predicting the generation is wrong16:14
dansmithyou would never do gen+=1 locally16:14
dansmithyou have no idea if your generation 6 is the same as the server side16:14
dansmithyou think so because you posted with 5 and it succeeded, but you don't know16:14
edleafedansmith: which is why you always need to expect a 40916:15
edleafeotherwise, what's the point of caching anything?16:16
efriedWhich is actually another good point.  DELETE /rp/{u}/inventories doesn't accept a generation (because doesn't accept a payload).16:16
*** lpetrut_ has quit IRC16:17
edleafeefried: ah, now *that's* a problem16:17
*** jobewan has quit IRC16:17
efriedHere's the other place we're making assumptions: -- that a newly-created provider has gen 0.16:19
efriedBecause the POST /resource_providers responds 201 with no content.16:20
edleafeefried: that's the same assumption as earlier. generation starts at 0, and each change increases it by one16:21
edleafeif you just created it, it's at zero16:21
efriededleafe Right.  Which isn't a good assumption, unless it's a documented part of the API, which today it ain't.16:22
*** slaweq has joined #openstack-nova16:22
edleafeIt's sounding more and more like a doc bug16:22
efriedIf in fact we want it to be part of the API.  Which I'm not convinced of.16:23
*** lucas-brb is now known as lucasagomes16:23
*** tidwellr_ has joined #openstack-nova16:27
*** tidwellr has quit IRC16:27
*** slaweq has quit IRC16:27
efriedI'm gonna say we definitely have a doc shortage with respect to how to *use* the generation in any case.  Other than that one spot in the API ref, there's not even anything that tells you you're supposed to send down the same generation you GET.16:29
*** felipemonteiro_ has quit IRC16:33
*** felipemonteiro_ has joined #openstack-nova16:34
cdentefried, edleafe: I definitely agree (with dansmith) that _in the presence of the provider tree caching mechanism_ we should never gen += 1 on the client side. Instead we should always be prepared to invalidate the _entire_ cache when a 409 happens. This is the reason for creating the PlacementAPIConflict base class: so any gen conflict can cause a complete reset.16:35
*** pcaruana has quit IRC16:35
openstackgerritMatt Riedemann proposed openstack/osc-placement master: CLI for resource classes (v1.2)
openstackgerritMatt Riedemann proposed openstack/osc-placement master: RP list: member_of and resources parameters (v1.3, v1.4)
openstackgerritMatt Riedemann proposed openstack/osc-placement master: RP delete inventories (v1.5)
openstackgerritMatt Riedemann proposed openstack/osc-placement master: CLI for traits (v1.6)
openstackgerritMatt Riedemann proposed openstack/osc-placement master: Resource class set (v1.7)
openstackgerritMatt Riedemann proposed openstack/osc-placement master: Usages per project and user (v1.8, v1.9)
openstackgerritMatt Riedemann proposed openstack/osc-placement master: CLI allocation candidates (v1.10)
openstackgerritMatt Riedemann proposed openstack/osc-placement master: [WIP] Get resource provider by uuid or name
jackie-truongmriedem: During the 1/4 meeting, you said that you weren't sure who should be reviewing the certificate validation feature and that you expect a lot of blueprints to get deferred. jaypipes has been reviewing the patch series since that meeting, but seems to be out this week. Do you think it's reasonable to try to get this patch series through before the feature freeze?16:36
efriedcdent No arguments there.16:36
efriedIn the scenario at hand, we don't get a 409, even though we should.16:36
cdentefried, edleafe; Another option is for the provider tree to do a full reset after any finished operation16:36
edleafecdent: so the cache will *always* be invalid, as we will never have agreement with the rp's data and its generation16:36
mriedemjackie-truong: i'm not going to be able to context switch to an entirely new thing the day before FF16:36
cdentedleafe: yeah, which is why I've always wondered why we have a cache?16:36
edleafecdnet: jinx-ish16:36
efriedcdent At some point you started work on the "updating generation when aggregates changed" - where is that?16:37
edleafecdent, not cdnet16:37
efriedcdent I thought you started writing a spec, but it must not have been in nova-specs.16:37
mriedemjackie-truong: unfortunately it's probably going to get deferred to rocky16:37
*** fanzhang has quit IRC16:37
*** fanzhang has joined #openstack-nova16:37
cdentefried: the scenario at hand is using logic that pre-dates the provider tree and was an acknowledged hack that was deemed safe-ish in the world of _not_ nested16:37
mriedemjackie-truong: i'd like it if we could do some version of the 'slots' or 'runways' concept so these blueprints that keep getting deferred from release to release actually become a review priority early in the next cycle before we take on new work16:38
cdentefried: mriedem told me that wasn't going to get considered until rocky, so I lowered it off my priority list in favor of rechecking lots of things over and over16:38
cdentthat == aggregates with generations16:38
mriedemespecially before the people grinding that contribution don't burn out and move on16:38
*** janki has joined #openstack-nova16:38
jackie-truongmriedem: Yeah, that'd be nice16:38
efriedcdent Okay, did you not already start writing words about it, though?16:38
mriedemjackie-truong: i'll add something to the dublin ptg agenda for discussion16:39
jackie-truongmriedem: If I can find jaypipes, I'll see if he has any time to look at the patches. Otherwise, I'll plan to wait for rocky16:39
*** janki has quit IRC16:39
efriedcdent I think we need a launchpad-something (bug?  blueprint?) for that.16:39
*** janki has joined #openstack-nova16:39
efriedcdent And I want to point to it from the bug I'm opening now, titled "Placement client cache consistency is broken"16:40
edleafecdent: did you see my comments starting here?
*** slaweq has joined #openstack-nova16:40
edleafeThat's my understanding how caching was supposed to work with generations16:40
cdentefried: I hope you'll revise that to say ProviderTree cache consistency, not Placement. Placement is fine.16:40
cdentedleafe: looking. I tried to read the whole backlog but am way sick, so faking having a brain16:41
efriedcdent "Placement client"  not "Placement".16:41
cdentit's not even a placement client, it's the provider tree16:41
cdentwhich uses a placement client16:41
edleafecdent: so IOW, normal for you :-P16:41
cdentedleafe: yes16:41
cdentif the day of the week ending in 'y': chris sick16:41
efriedp'tayta p'tata, but okay.16:42
mriedemjackie-truong: L102
efriedAny placement client wishing to use a cache is broken16:42
cdentI disagree16:43
cdentAny placement client wishing to use a unified cache is broken16:43
cdentI dispute the need for a cache16:43
mriedemgibi: have time to go over this again before your eod?
jackie-truongmriedem: Awesome. Thanks for adding that16:44
cdentefried: but it is the road we went down16:44
cdentand I think we can make it work16:44
cdentthe things edleafe says at [l 4liQ] seem sane16:44
efriedcdent That part has never been under dispute.16:45
edleafecdent: we seem to have an underwhelming amount of docs about using the generation16:45
cdentI don't doubt that16:45
cdentefried: which part is "that part"?16:45
efriedWhere we should get a 409 if we PUT something with a generation that's not the latest.16:46
cdentthe problems that seem possible from edleafe's flow is that appears to assume a granularity of interaction that it's not clear if the provider tree enables?16:47
edleafecdent: re: generation doc amount:
*** slaweq has quit IRC16:48
efriedcdent I didn't follow that.16:48
* cdent loves me some EC16:48
*** yamamoto has quit IRC16:49
cdentWhat is it that we want/need the generation docs to say?16:50
*** damien_r has quit IRC16:50
efriedcdent Certainly they need to say, "You need to include whatever generation came in your GET when you do your PUT.  If something else updated the provider in between your GET and your PUT, you'll get a 409."16:51
efriedcdent One part of the doc says sort of the second half of that, for one API.16:51
edleafeBasically that it is a monotonically-increasing integer, and is increased with each modification of an RP16:51
cdentdo we _want_ people to know it is monotonic?16:51
openstackgerritMatt Riedemann proposed openstack/osc-placement master: Address review comments from allocations patch
openstackgerritMatt Riedemann proposed openstack/osc-placement master: Usage docs and initial release note for osc-placement
cdentthat point seems to not have agreeement16:52
mriedemstephenfin: ok updated the docs one more time to fix a link and to include the resource class stuff which is approved ^16:52
efriedcdent Beyond that, they *either* need to say, "We start at zero and increment by 1 whenever something changes about the provider," *or*, "You can't count on how we deal with the generation; it's an opaque value that you need to turn around and send back to us."16:52
efriedcdent Yes, I agree that point does not have agreement.16:52
efriedIf we document the former, then the code as it sits is (almost) okay.16:53
cdentefried why would we want to say the first point above? I think of it as the second.16:53
efriedcdent I'm with you.16:53
efriededleafe is on the other side.16:53
efriedWe don't know what jaypipes thinks16:53
cdentI think edleafe is being pragmatic, more than anything? edleafe ?16:53
efriedAnd I *think* dansmith is also on the "opaque" side.16:53
edleafeif our code makes that assumption, we should document that16:53
efriedwhoah, we should document something about the API because of something our client code assumes?  That's cray-cray.16:54
dansmithefried: yes16:54
dansmithefried: in fact, we could use some random sequence of numbers if we had one and that should be fine. or a uuid. "here is my proof that I'm making a change to the most recent copy" is all we should assume from the client side about the meaning of the generation field, IMHO16:55
efriedI'd be willing to bet the client assumptions were made out of expediency to get function merged.  At the time it didn't matter because a) single point of control, b) no nested, c) no traits/aggs to complicate.16:55
edleafeefried: the code uses it because it was understood to be the case. If it is going to be the case, it should be documented. If it isn't, the code that assumes that should change16:55
efrieddansmith ++16:55
efriededleafe ++16:55
cdentefried: yes, expediency is what I said above16:55
cdentgeneration are ints because jay likes ints in databases16:56
edleafeWe went with a monotonic integer scheme, even though there were better options available. I don't recall why it won out16:56
cdentbut that they are ints is incidental16:56
efriedSo we need to make a call.  Course, Jay should be included in the "vote" (if it's a democracy in the first place).16:56
* edleafe thinks that assumption is cute16:57
efried"opaque" will require some nontrivial redesign and rework16:57
cdentIs the call as narrow as "should we += 1 in the client" or something more broad than that?16:57
efriedcdent The call is, do we document +1 behavior or do we document opacity.  What flows from the former is fairly narrow.  What flows from the latter is broad.16:58
*** gyee has joined #openstack-nova16:59
efriedAnd not to distract, but we still definitely have a problem with DELETE not sending a generation down, no matter which way we decide to go.16:59
dansmithefried: I'm not sure we fully have a problem with delete17:00
dansmithefried: I should be able to delete allocations without a generation without being in violation of anything17:00
dansmithbecause I know the thing I'm deleting for is never coming back17:00
dansmithwhat else do we delete for? inventory maybe, but I bet the same argument works for that17:00
cdentefried: I think we can continue to fake on +1 in the ProviderTree if that's what we need to do, and _not_ document it. Other (to be developed) clients won't need to care and should respond to documentation that says what it already says "A consistent view marker"17:01
efrieddansmith Inventory, yeah, and agree it's the same argument.17:01
dansmithefried: same argument, but you think we _do_ need a generation on delete right?17:02
efrieddansmith Considering the multiple-points-of-control thing, if I think I'm ready to delete the RP's inventory, but you think you're happily editing it...17:02
dansmithefried: but inventory is and should only be managed by the thing owning the inventory17:03
*** gszasz has quit IRC17:03
efrieddansmith If that's the case, we don't need generations at all.17:03
dansmithefried: I guess the only case that we might hit would be:17:03
efrieddansmith I think the problem comes in when deleting the whole inventory is actually a result of "I want to remove inventory in this resource class, but that happens to be the last resource class".17:03
dansmithefried: I'm going to _allocate_ against a RP, right when the RP is deleting the inventory17:03
efrieddansmith I think yours is okay.  The DELETE will bounce if the allocations happened first; and the allocations will bounce if the DELETE happened first.17:04
*** dhellmann has joined #openstack-nova17:05
*** jackie-truong has quit IRC17:05
dhellmannsdague : are you interested in this testing patch for the nova doc redirects?
efrieddansmith But for mine: let's say for whatever reason two threads own inventory in different resource classes on the same RP. Thread A decides there's no longer any CUSTOM_FOO, so he goes to remove all the CUSTOM_FOO.  He does a GET and sees that the inventory comprises *only* CUSTOM_FOO, so he gets ready to do a DELETE.17:05
*** yamamoto has joined #openstack-nova17:05
cdentefried: I thought the point of the provider tree itself was to lock over that kind of operation?17:06
cdentthe tree can only be manipulated by one thread at a time17:06
efrieddansmith Meanwhile, thread B decides it's time to add some CUSTOM_BAR.  He does his GET, sees some CUSTOM_FOO in there, adds his CUSTOM_BAR, and PUTs the whole shebang back.  Now on the server, there's CUSTOM_FOO and CUSTOM_BAR.17:06
dansmithefried: but nothing other than the thing that owns the inventory should collapse a PUT to a DELETE because it thinks it's the last thing, right?17:06
dansmithefried: and the PUT in your case would have included a generation which is wrong now because the DELETE bumped it, no?17:07
*** yamamoto has quit IRC17:07
*** yamamoto has joined #openstack-nova17:07
*** yamamoto has quit IRC17:07
efriedNo, because the PUT happens first.17:07
efriedThat's the point.17:07
dansmithno I meant the second case17:07
dansmithin the first case, the non-owner is collapsing a PUT to a DELETE which isn't legit It hink17:08
cdentefried: is [t 1zzs] not true?17:08
purplerbot<cdent> the tree can only be manipulated by one thread at a time [2018-01-24 17:06:26.123234] [n 1zzs]17:08
dansmithcdent: it doesn't matter,17:08
efrieddansmith I don't disagree.  But that's what the code is doing today.17:08
dansmiththere could be multiple trees on multiple hosts I think is his point17:08
dansmithefried: I'm saying that's probably wrong, for the non-owner17:09
cdentdansmith: I though we were concluding that the compute node owns the inventory17:09
cdentbut I guess some things might share nested bits17:09
cdentwhich is17:09
*** tidwellr_ has quit IRC17:09
*** chyka has joined #openstack-nova17:09
dansmithcdent: we do, I'm just saying I think efried is talking about a more general case of multiple external actors needing to interact with one set of inventory or allocations, using generations as the locking mechanism17:10
efriedsean-k-mooney has been coming up with all sorts of interesting use cases where it may very well be the case that e.g. nova owns the VFs, neutron owns the VIFs, and some other thing owns the bandwidth - all on the same RP.17:10
dansmithI'm only half paying attention to this conversation so maybe I should keep quiet :)17:10
*** tidwellr has joined #openstack-nova17:10
dansmithefried: I'm not sure I agree that such a case is legit17:11
efrieddansmith So far you've only been reaffirming my position, so by all means continue half paying attention.17:11
cdentefried: in which case a cache that is local the compute node is going to fraught with challenges beyond the one you have identified today17:11
efrieddansmith Once you start disagreeing, then yeah...17:11
openstackgerritStephen Finucane proposed openstack/nova master: Don't filter out sibling sets with one core
openstackgerritStephen Finucane proposed openstack/nova master: trivial: Update tests to set 'NUMACell.siblings'
openstackgerritStephen Finucane proposed openstack/nova master: Ensure emulator threads are always calculated
openstackgerritStephen Finucane proposed openstack/nova master: Always pass 'NUMACell.siblings' to _pack_instance_onto_cores'
*** bhujay has quit IRC17:14
*** bhujay has joined #openstack-nova17:14
*** gyee has quit IRC17:15
cdentefried: are you blocked in the immediate sense, or "merely" the medium term sense?17:17
efriedcdent I'm not blocked, other than the fact that I'm discussing and writing up this issue rather than working on the update_from_provider_tree patch.17:18
efriedI think we're broken long-term, but again, it's not going to matter until we have real-world situations that can do concurrent updates to the same RPs.17:18
efriedWhich won't be Q.17:18
cdentis "broken if we want a cache" or "broken"?17:19
cdentWhat I mean is, if later we decided "screw it, let's strip out the cache", would that be a thing?17:19
cdent(just trying to make sure I'm grokking)17:20
mriedemalex_xu: some comments in which can be addressed in the follow up nit cleanup patch - and also a request for a functional test, which would be done in a follow up patch17:20
efriedcdent If we get rid of the cache, and never DELETE allocations/inventories unless we know we're the sole owner thereof, then I think we would be okay.17:20
efriedcdent I don't think the placement API is broken.  I just think it has some holes for consumers wishing to do good caching.17:21
efriedcdent And certainly some holes in documentation.17:22
cdentfor allocations that's (sole owner thereof, when deleting by consumer id) supposed to be true, but for inventories there's probably work to do17:22
mriedem"And certainly some holes in documentation." - i will say, all things aside to this discussion, the placement stuff has a lot of good documentation compared to what we've historically had for stuff in nova17:22
cdentFrom my perspective placement wasn't designed for caching. It was designed to be capable of tell you the truth, right now.17:23
cdentefried: if it can't do that fast enough, that's the bug17:23
cdentfast -> effectively, efficiently, low latency, reliably17:23
efriedcdent I continue to be vehemently not on board with the theory that it's okay to do lots of calls to placement if that can be avoided.17:24
cdentwe can agree to disagree on that and I respect your opinion.17:24
efriedcdent Because I don't care how fast we get it, it's still a "call over the wire" (even if localhost).17:24
*** slaweq has joined #openstack-nova17:24
*** sapd has quit IRC17:24
efried...which is always going to be an order of magnitude higher overhead than a local cache index.17:24
cdentefried: if you want to avoid calls over the wire I'd suggest looking at RPC :)17:25
cdentwhich is ripe for adjustments17:25
efriedEven RPC17:25
efriedoh, that's not what you meant.17:25
edleafeand what good is a local cache if you aren't sure it's current?17:25
efriededleafe Or more precisely, you can't detect reliably when it's not.17:25
edleafethat's what the 409 handling is about17:26
efriedwhich is what started me off on this whole thing.17:26
cdentefried: we can (and perhaps should) put a placement web service (with a memcached) near the compute nodes17:26
efriedI don't know what that means.17:26
cdenthave hundreds of them17:26
efriedBut it's still ultimately socket(), nah?17:26
efriedHell, even if that socket() is to a fifo, it's still an order of magnitude heavier than talking to the cache.17:27
cdentsure, but a) that's such the example of early optimization, b) if talking to placement turns out to be the consistent high expense in spawning a server, I'd be surprised.17:27
cdentSince we don't know that b is or is not a problem17:27
cdentit is17:27
cdenttoo soon17:27
cdentto be optimizig17:27
efriedyet here we are.17:27
cdentindeed, wtf?17:27
edleafeI was just typing the same thing - optimizing when we don't know that that's the problem17:28
cdentbut the great thing is: it mostly works, we have a temporary workaround, and we can keep on improving stuff17:28
efriedso now on the table is a third option, even broader.17:28
efriedfor the long term17:28
efriedwhich is: rip out the cache17:28
*** felipemonteiro_ has quit IRC17:29
cdentefried, edleafe: so fairly productive conversation to some extent; more pages shared17:30
mriedemthis reminds me, we're GETing aggregates twice per update_available_resource periodic for 0 reason :)17:30
cdentyes _that_ is a bug17:30
cdentbut I think efried fixes it in ProviderTree?17:31
mriedemi was going to push a simple backportable patch to remove that unnecessary callin17:31
efriedI'm actually not sure any of the stuff I've done touches the legacy update_available_resource code paths.17:31
mriedemlike just remove the shit, and leave a comment saying "revert git hash xyz once aggregates are a thing the client cares about"17:31
mriedemthe ever growing todo list17:32
cdentmriedem: is it enough of a concern to do anything? I seem to recall someone reporting it as a problem?17:32
cdentoh, I get you, take it all out17:32
mriedemklindgren__ at godaddy was just noticing the number of placement REST API calls in a single periodic run w/o no changes otherwise17:32
efriedmriedem If aggregates were only useful for sharing providers, maybe.  Also, it won't be as simple as "revert this commit".  That's gonna be merge conflict central on the patch series in flight right now.17:32
mriedemi assume to start planning for the scale reqiurements17:32
mriedemefried: i realize, but it would be a breadcrumb to look at what existed before,17:33
mriedemidk, could just be simpler / better to remove it all and when needed, add the stuff back in fresh as needed17:33
*** slaweq has quit IRC17:33
cdentI gotta go before getting sucked into another thing, I feel like ass.17:33
efriedmriedem On what time frame?  Before FF?  Or between now and when we cut Q?17:33
mriedemefried: i should be backported, so whenever17:34
* cdent waves17:34
efriedBye cdent, thanks for the talk.17:34
*** cdent has quit IRC17:34
efriedmriedem So that backport is gonna be very different for Q and pre-Q, just sayin.17:35
*** mvk has quit IRC17:35
mriedemmelwitt: want to hit this cleanup patch and the docs one after it? - i've got the osc-placement 1.0.0 release dependent on this series17:35
mriedemefried: because the provider tree stuff changed everything?17:35
melwittmriedem: sure thing17:35
mriedemi haven't looked at things there in 2 weeks17:36
efriedmriedem Yes, changed some things, and added lots of things.17:36
mriedemanyway, whatever, it should be pretty straight-forward17:36
mriedemmelwitt: thanks17:36
*** pramodrj07 has joined #openstack-nova17:37
* mriedem goes to get osc-placement release notes published17:37
melwittmriedem: will this link be working after the change merges or does it need to be corrected now?
*** slaweq has joined #openstack-nova17:43
mriedemmelwitt: it'll work as a result of this change17:43
mriedemand once the docs get published17:43
mriedembauzas: what do you need from me wrt the libvirt gpu series?17:44
dansmithmriedem: :17:47
openstackgerritDan Smith proposed openstack/nova master: Avoid suspending guest with attached vGPUs
dansmithmriedem: I just fixed up all the things we commented on17:47
dansmithmriedem: I can try to find someone else to be the second +2 if my hands are too dirty.. I'll let you make that call17:48
*** gouthamr has joined #openstack-nova17:49
*** gouthamr has quit IRC17:49
*** slaweq has quit IRC17:49
*** mvenesio has joined #openstack-nova17:49
*** tidwellr has quit IRC17:51
*** Tahvok has quit IRC17:51
*** doude has joined #openstack-nova17:51
*** Tahvok has joined #openstack-nova17:52
*** mdrabe has quit IRC17:53
*** mdrabe has joined #openstack-nova17:53
*** bhujay has quit IRC17:54
openstackgerritDan Smith proposed openstack/nova master: Avoid suspending guest with attached vGPUs
dansmithmriedem: sorry, I forgot to replace one line in the test ^17:54
*** edand__ has quit IRC17:54
*** tidwellr has joined #openstack-nova17:54
*** gouthamr has joined #openstack-nova17:56
dansmithlbragstad: hey, mriedem told me you fixed the copious warnification about policy deprecativity recently17:59
dansmithbut I still experience said pain17:59
dansmithis there something we're doing that blocks us from experiencing intended test euphoria?18:00
*** xinliang has quit IRC18:00
lbragstaddansmith: oh - really?18:01
dansmithlbragstad: yah18:01
lbragstadi had a few patches in flight for a couple different issues, let me check to see where they are at18:01
*** sahid has quit IRC18:01
openstackgerritOpenStack Release Bot proposed openstack/os-traits master: Update reno for stable/queens
openstackgerritOpenStack Release Bot proposed openstack/os-vif master: Update reno for stable/queens
lbragstaddansmith: what version of oslo.policy are you experiencing this with?18:03
dansmithlbragstad: I just tox -r'd, but let me look18:03
dansmith(r/s/b/add-suppo)% grep policy requirements.txt18:04
dansmithoslo.policy>=1.30.0 # Apache-2.018:04
mriedemdansmith: +2 on - i think you're ok to +W18:04
lbragstad is the patch that should have fixed the issue you're seeing18:04
dansmithmriedem: rr thanks18:04
*** david-lyle has quit IRC18:04
*** xinliang has joined #openstack-nova18:04
lbragstadand that *should* be in oslo.policy 1.33.118:04
lbragstadwhich looks good here -
dansmithlbragstad: hmm, but I've got 1.30 in requirements.. am I missing how g-r overrides that or something?18:06
*** derekh is now known as derekh_afk18:07
lbragstadthat's a good question18:07
*** yamamoto has joined #openstack-nova18:07
mriedemdansmith: upper-constraints should pull in 1.33.118:07
lbragstaddansmith: do you know exactly which version of oslo.policy you have in your env?18:08
*** sambetts is now known as sambetts|afk18:08
melwittmriedem: can you remind me if you know if anything is unique to the cells v1 tempest job? in looking at the libvirt errors on stable, it's both stable branches and only the cells v1 job18:08
mriedemhe's got 1.3018:08
dansmithlbragstad: above I quoted what is in requirements.txt18:08
*** janki has quit IRC18:08
mriedemmelwitt: cellsv1 job shouldn't have anything unique about the compute / virt setup18:08
lbragstadmriedem: i thought he had >=1.30 in requirements18:08
*** tesseract has quit IRC18:08
mriedempip freeze your tox venv18:08
dansmithI hacked my requirements.txt, tox -r, and still get it18:08
mriedem1.33.1 is in u-c18:08
melwittmriedem: thanks18:09
dansmith(r/s/b/add-suppo)% .tox/py27/bin/pip freeze | grep policy18:09
dansmithafter my hack18:09
mriedembauzas: is this the final piece of the vgpu puzzle for libvirt ? i don't see any other libvirt patches for vgpu - so assuming we're done once that merges18:10
dansmithtrying again without the hack, but assume I'll get that same thing again18:10
mriedemdo we have any patch that has a feature support matrix update for vgpu support?18:10
lbragstaddansmith: do you have a policy file you're testing with locally, or are you just running nova tests?18:11
dansmithlbragstad: just nova tests18:12
mriedemdansmith: if it's always a handful of rules, it could be something in the policy fixture or a specific test18:12
*** jpena is now known as jpena|off18:12
lbragstadthe fix in 1.33.1 just makes it so a warning is logged iff a policy is deprecated *and* you're specifying it in a policy file or somewhere on disk18:13
dansmithmriedem: there are 11 warnings, seemingly regardless of what tests I run18:13
* lbragstad wonders if a fixture is setting up a policy on disk for those tests18:14
mriedemnova.tests.unit.policy_fixture.PolicyFixture maybe?18:15
mriedemloads things up from some fake rules18:15
mriedemfrom nova.tests.unit.fake_policy18:15
dansmithmriedem: wouldn't that only affect me if I'm running those tests though?18:15
dansmithI get this if I run a single-shot test18:15
lbragstadso - it is writing them to disk
mriedemdansmith: i think the base test case class loads up the policy fixture18:17
mriedemfor all tests18:17
lbragstadwhich would trip this case -
mriedemyup self.policy = self.useFixture(policy_fixture.PolicyFixture())18:17
mriedemit's loaded for all tests18:18
mriedemdansmith: so you'd have to try and remove the deprecated policy rules from the fake fixture data18:18
*** yamamoto has quit IRC18:18
lbragstadbecause oslo.policy thinks it needs to inform an operator about a policy override that is deprecated (i assume contains some deprecated policies)18:18
dansmithmriedem: it doesn't seem to have a set of rules in the fixture though18:19
*** stvnoyes has joined #openstack-nova18:20
mriedemdansmith: what's one of the ones it's complaining about?18:20
dansmithmaybe it's in
mriedemthat's the problem,18:20
lbragstad^ that one is deprecated18:20
mriedemit's writing the fixture data to a temp file18:20
mriedemand read in by the policy fixture18:20
dansmithwhat does that mean though? there's no value set for that18:21
dansmithor do you mean I just need to delete it?18:21
mriedem"" means all i think18:21
mriedemdelete those entries18:21
*** jafeha__ has joined #openstack-nova18:21
lbragstadi'm working on a patch now18:21
mriedemi'm not sure why we actually even have that fake_policy file anymore18:21
mriedemlbragstad has grown bored with keystone18:22
*** jafeha has quit IRC18:24
dansmithlbragstad: I've got one already18:24
dansmithabout to push18:24
openstackgerritDan Smith proposed openstack/nova master: Remove deprecated policy items from fake_policy
dansmithlbragstad: mriedem ^18:24
*** avolkov has quit IRC18:24
dansmithI think I still get something18:25
dansmith /dan/nova/.tox/py27/lib/python2.7/site-packages/oslo_policy/ UserWarning: Policy "os_compute_api:os-extended-volumes":"rule:admin_or_owner" was deprecated for removal in 17.0.0. Reason: Nova API extension concept has been removed in Pike. Those extensions have their own policies enforcement. As there is no extensions now, "os_compute_api:os-extended-volumes" policy which was added for extensions is not needed any more.18:25
dansmith Its value may be silently ignored in the future18:25
dansmithI think the rule:admin_or_owner bit is new now18:25
*** imacdonn has joined #openstack-nova18:26
*** imacdonn has quit IRC18:26
*** imacdonn has joined #openstack-nova18:26
lbragstaddansmith: oh nice18:26
dansmithhmm, actually, no18:26
dansmithI backed out the change and I get the same exact thing18:27
lbragstadweird - that's still emitting
lbragstadit looks like it's writing everything to disk18:30
lbragstadregardless of it being in fake_policy or not18:30
*** andrewbogott has quit IRC18:30
*** andrewbogott has joined #openstack-nova18:30
dansmithper sdague's comment above I assume18:30
*** ralonsoh_ has quit IRC18:30
lbragstadthis must have been around before the default in code efforts18:31
dansmithI feel like it cropped up between when I left last year and popped up this year18:31
*** pramodrj07 has quit IRC18:31
*** Pramod has joined #openstack-nova18:31
*** fragatina has quit IRC18:32
lbragstaddansmith: trying a workaround hack with
lbragstadchecking if i can recreate locally18:36
dansmithah that seems nice18:36
lbragstad^ that fixes it for me18:37
lbragstad without the fix...18:38
lbragstader... "fix", no idea if that's how we want to work around it18:39
lbragstadwith the workaround
*** tssurya_ has joined #openstack-nova18:40
openstackgerritLance Bragstad proposed openstack/nova master: Remove deprecated policies from fake_policy
openstackgerritLance Bragstad proposed openstack/nova master: WIP: Reduce policy deprecation warnings in test runs
*** tssurya_ has quit IRC18:41
openstackgerritLance Bragstad proposed openstack/nova master: WIP: Reduce policy deprecation warnings in test runs
dansmithlbragstad: remove the WIP?18:42
openstackgerritLance Bragstad proposed openstack/nova master: Reduce policy deprecation warnings in test runs
dansmithlbragstad: thanks!18:43
lbragstaddansmith: no problem - thanks for the heads up... not sure if all those policies need to be there, but at least it doesn't spam as bad18:44
*** ralonsoh has joined #openstack-nova18:45
mriedemdansmith: do we still need ?18:45
mriedemoh i guess so18:45
*** tssurya_ has joined #openstack-nova18:49
mriedemstephenfin: melwitt: umm
mriedemcould we not have red hat cores approving red hat changes?18:56
*** Pramod has quit IRC18:56
*** felipemonteiro has joined #openstack-nova19:00
melwittmriedem: sorry about that. I have asked red hatters not to approve red hat changes that I've +2ed so I think this was just an oversight. downgrading my vote to +119:01
melwitter, I guess that doesn't help because I'm not the approver19:01
*** felipemonteiro_ has joined #openstack-nova19:02
dansmithyeah, it won't help :(19:02
*** tssurya_ has quit IRC19:02
dansmitha -2 will, but..19:02
dansmithit will also reset the gate if it's already there19:02
*** jackie-truong has joined #openstack-nova19:02
mriedemwe do'nt need to pull it out,19:02
mriedemi'm not going to review it in depth at this point19:03
mriedemthe commit message obviously shows there was thought put into it19:03
artomIt's that suave British accent, isn't it?19:04
mriedemwho wrote the commit message, lee or booth?19:04
mriedemthe live migration job runs with ceph i think,19:05
mriedembut i guess we don't enable volume-backed live migratoin do we,19:05
mriedemand even if we did, we don't have a tempest test for encrypted volume-backed live migration to run with ceph19:05
mriedembut we do have new enough libvirt and qemu in the pike UCA to test this19:05
*** felipemonteiro has quit IRC19:06
*** tssurya_ has joined #openstack-nova19:06
melwittmriedem: I have a patch up to enable it but it looks like it's hitting legit problems
*** lucasagomes is now known as lucas-afk19:10
mriedemmelwitt: is there supposed to be at least even a release note anywhere in this series?19:10
*** tidwellr has quit IRC19:10
mriedembecause given the commit message, it seems pretty complicated19:10
mriedemor some kind of advertisement anywhere about 'hey you can do this thing now with libvirt and encrypted volumes'19:10
openstackgerritMerged openstack/osc-placement master: CLI for resource classes (v1.2)
*** tidwellr has joined #openstack-nova19:10
melwittmriedem: I think there's no release note because it's automatic, that is, operators don't have to do anything, it will use native luks decryption if available. but now that you mention it, it might be helpful to mention that it will use native decryption if qemu and libvirt version combination support it19:14
*** mlavalle has quit IRC19:15
mriedemright, i mean, unless this is totally rock solid, and shit starts randomly failing, operators are probably going to want to know that something new is going on here19:15
melwittyeah, I agree with that19:16
artomHeh, we talked about this in the spec:
artomAnd then promptly forgot about it :(19:17
*** david-lyle has joined #openstack-nova19:17
*** mlavalle has joined #openstack-nova19:18
dhellmannmriedem : is of any interest or should I just abandon that?19:20
mriedemdhellmann: it is of the utmost interest19:23
dhellmannmriedem : ok. I'll leave it open then until someone has a minute to review it.19:23
mriedemdhellmann: already did19:24
mriedemmelwitt: didn't lee say he was going to be out a bit this week? do you think you could take a stab at a release note?19:24
mriedemsince you know the change19:24
mriedemas a patch on top of course19:24
openstackgerritEd Leafe proposed openstack/nova master: Add unit test for non-placement resize
mriedemedleafe: thanks for that19:24
edleafemriedem: np19:25
melwittmriedem: sure, will do19:25
mriedemmelwitt: thanks19:25
mriedemthen i guess we can close that one out for queens19:25
openstackgerritMerged openstack/osc-placement master: Address review comments from allocations patch
openstackgerritMerged openstack/osc-placement master: Usage docs and initial release note for osc-placement
*** pcaruana has joined #openstack-nova19:27
*** pcaruana has quit IRC19:27
mriedemthe cinder gate fix has been promoted btw19:28
mriedemso hopefully stuff starts merging again soon19:28
*** ralonsoh has quit IRC19:28
*** Guest50651 has quit IRC19:29
*** felipemonteiro_ has quit IRC19:32
*** muttley has joined #openstack-nova19:32
*** AlexeyAbashkin has joined #openstack-nova19:34
*** jackie-truong has quit IRC19:35
melwittthat's super news ++19:35
*** suresh12 has joined #openstack-nova19:36
*** harlowja has joined #openstack-nova19:38
openstackgerritMark Goddard proposed openstack/nova master: Unplug all VIFs from ironic nodes during tear down
*** yamahata has joined #openstack-nova19:38
*** AlexeyAbashkin has quit IRC19:38
*** fragatina has joined #openstack-nova19:40
*** slaweq has joined #openstack-nova19:45
*** dtantsur is now known as dtantsur|afk19:46
*** amoralej is now known as amoralej|off19:46
*** muttley has quit IRC19:48
*** tssurya_ has quit IRC19:49
dansmithmelwitt: mriedem tssurya: do we need to have a cells meeting today? I've got nothing new to talk abotu19:49
*** archit has joined #openstack-nova19:50
architmsg amodi identify titanic12319:50
dansmithwhoopsie :)19:51
*** tssurya_ has joined #openstack-nova19:51
archityep .:(19:51
*** amodi has quit IRC19:51
melwittdansmith: I don't have anything new either19:51
efriedswhy I always do those in the status channel19:51
mriedemdansmith: there are no major new bugs that i'm aware of,19:51
mriedemand the resize + alternate hosts change is approved19:51
efriedIs this cinder bug going to affect all patches?19:51
mriedemefried: anything that runs tempest api tests19:51
mriedemso all dsvm jobs19:52
*** slaweq has quit IRC19:52
efriedso all my nova patches19:52
* efried is going for record number of rechecks19:52
mriedemit's #13 in the gate right now19:52
mriedemdansmith: melwitt: so only other thing would be ptg topics, but that can wait until after FF19:53
*** archit is now known as amodi19:53
mriedemtssurya: tssurya_: do you have any status updates from cern's migration to cells v2?19:53
tssurya_mriedem : actually belmiro was going to participate in the cells meeting today, not sure though; but yea we are going to be doing in at the end of Feb19:54
dansmithdrat! :)19:54
mriedemtssurya_: doing it in prod? but you're all clear in pre-prod testing?19:55
tssurya_migrating to Ocata during the last days of this month19:55
dansmithtssurya: can you find out if he's going to be around?19:55
tssurya_and we will have to move to Pike in a few weeks19:55
mriedemok; would be good to know as soon as possible if there are blockers *before* rocky19:55
tssurya_yes in prod19:55
mriedembecause i think we still plan on dropping cellsv1 and nova-net in rocky19:55
tssurya_dansmith : yea will ping him and see if he will be joining19:56
dansmithtssurya_: okay thanks19:56
tssurya_mriedem : yea we have half out cloud on neutron19:56
tssurya_and trying to scale it to move out ot nova-net19:56
tssurya_out of *19:56
tssurya_our *19:56
*** amodi has quit IRC19:56
*** amodi has joined #openstack-nova19:57
*** tidwellr has quit IRC19:58
*** dave-mccowan has quit IRC19:58
mriedemthis needs another +2 - - moves the nova-tox-functional job in-tree so we can define our own files blacklist, so we don't have gate issues again with notification samples only patches not getting test20:01
*** tidwellr has joined #openstack-nova20:01
*** jangutter has joined #openstack-nova20:03
*** jangutter has quit IRC20:05
*** jangutter has joined #openstack-nova20:05
*** matrohon has joined #openstack-nova20:06
openstackgerritsean mooney proposed openstack/nova master: Change 'InstancePCIRequest' spec field
openstackgerritsean mooney proposed openstack/nova master: Add Neutron port capabilities to devspec in request
openstackgerritsean mooney proposed openstack/nova master: Format NIC features using os-traits definitions
openstackgerritsean mooney proposed openstack/nova master: Read Neutron port 'binding_profile' during boot
*** sean-k-mooney has joined #openstack-nova20:13
*** jaypipes has joined #openstack-nova20:15
sean-k-mooneystephenfin: if you have time tomorow rodoflos nic feature based schduling patch series is now all rebased with versioned objects
*** slaweq has joined #openstack-nova20:17
*** amodi has quit IRC20:20
*** dtruong has joined #openstack-nova20:20
openstackgerritmelanie witt proposed openstack/nova master: Add release note for QEMU native LUKS decryption
*** gouthamr has quit IRC20:26
*** READ10 has quit IRC20:26
*** slaweq has quit IRC20:29
mriedemmelwitt: some questions in that reno20:32
*** dhellmann has left #openstack-nova20:32
openstackgerritEric Fried proposed openstack/nova master: ProviderTree.get_provider_uuids: Top-down ordering
openstackgerritEric Fried proposed openstack/nova master: set_{aggregates|traits}_for_provider: tolerate set
openstackgerritEric Fried proposed openstack/nova master: WIP: SchedulerReportClient.update_from_provider_tree
openstackgerritEric Fried proposed openstack/nova master: WIP: Use update_provider_tree from resource tracker
openstackgerritEric Fried proposed openstack/nova master: Fix nits in update_provider_tree series
openstackgerritEric Fried proposed openstack/nova master: Move refresh time from report client to prov tree
openstackgerritEric Fried proposed openstack/nova master: WIP: New-style _set_inventory_for_provider
*** cleong has quit IRC20:37
*** slaweq has joined #openstack-nova20:39
*** slaweq has quit IRC20:39
*** slaweq has joined #openstack-nova20:40
*** dave-mccowan has joined #openstack-nova20:41
*** itlinux has joined #openstack-nova20:44
*** slaweq_ has joined #openstack-nova20:45
dansmithtssurya_: any word from belmiro?20:46
*** tidwellr has quit IRC20:47
*** slaweq has quit IRC20:47
tssurya_dansmith: no, nothing yet20:47
dansmithokay I guess we'll just go forward with it, see if he shows up and if not, bail :)20:48
tssurya_we had a couple of things to discuss, but nothing urgent;20:48
*** jangutter has quit IRC20:48
tssurya_in case he doesn't show up, we will just ask in the channel tomorrow20:49
*** slaweq_ has quit IRC20:50
*** belmoreira has joined #openstack-nova20:51
tssurya_oh cool he is here20:51
dansmithwell, lookie there20:51
dansmithbelmoreira: we were going to cancel the cells meeting to focus on other things, but tssurya_ rained on our parade and said you were coming with things to talk about20:52
*** jaypipes has quit IRC20:53
*** jackie-truong has joined #openstack-nova20:54
belmoreirahi dansmith. We will move to Ocata next week. Then start preparing Pike and cellsV220:54
*** tidwellr has joined #openstack-nova20:54
mriedembut ocata requires cellsv2...?20:54
mriedemi'm confused20:54
dansmithocata being single cell I assume20:55
belmoreiraI'm interested about the progress in listing instances if a cell goes down and the quota issue20:55
belmoreiramriedem: in Ocata we define only one cell but continue to use cellsV120:55
mriedembelmoreira: there hasn't been any progress on that,20:55
dansmithbelmoreira: tssurya_ was going to put up a straw man for us to discuss with api people20:55
mriedemit's a discussion item for the ptg20:56
jackie-truongmriedem: The cert validation patch series has been +2'd :-) Can I move it to the "Needs final +2 to complete blueprint" section in the nova-queens-blueprint-status etherpad?20:56
dansmithwhich I don't think has been posted, or at least that I saw20:56
mriedemjackie-truong: go for it20:56
jackie-truongmriedem: *\o/*20:56
tssurya_dansmith : we have been working on it and have some ideas/questions on how to proceed20:57
belmoreiratssurya_ and I started to check for some alternatives. The PTG for us may be to late20:57
mriedemjackie-truong: i don't understand why this isn't in the series before the API change?
dansmithtssurya_: okay cool20:57
belmoreiraI would like to move to Pike in few weeks after Ocata. Meaning that we will need to cook something even it's only internal20:58
*** tidwellr has quit IRC20:59
jackie-truongmriedem: That patch implements the the certificate_utils module, but doesn't make any changes to the API. So it basically adds the functionality with no way to use it.20:59
jackie-truongmriedem: The API change adds the usability part20:59
mriedemjackie-truong: so technically here,20:59
mriedemwe could merge the API,21:00
mriedemw/o the utils merged to do the things the api says you can do, right?21:00
mriedemadding functionality that is eventually exposed out of the rest api is kind of how all new things work21:00
*** liverpooler has quit IRC21:00
jackie-truongThe API change works with the trusted_certs object added in
jackie-truongSo, techincally21:01
jackie-truongWe could merge the API change (which would require the trusted_certs object change to be merged)21:01
jackie-truongwhich would result in you being able to create and pass around a TrustedCerts object21:02
mriedemwhat does the cert utils module do then?21:02
jackie-truongWithout making use of it21:02
mriedemif i can't make use of the thing, what's the point of having the api before said thing works?21:02
jackie-truongthe cert utils module takes in the TrustedCerts object (which gets turned into a list of strings)21:02
jackie-truongI realize that it technically makes sense to have the dependencies work: TrustedCerts object -> API -> cert_utils module21:03
jackie-truongBut technically, that isn't necessary21:03
jackie-truongSince the API can be updated to pass around a useless object21:04
mriedemno i was saying object > util (make things do something) > API (expose the things that do something)21:04
jackie-truongYeah, that timeline makes sense too. Basically, both util and API changes depend on the object change21:05
jackie-truongbut util and API don't have any dependency on each other21:05
jackie-truongSo we used the depends-on tag to indicate what dependencies needed to build21:06
mriedemjackie-truong: i think you're asserting like code deps, and i'm asserting functionality deps21:06
melwittmriedem: replied21:06
jrollmriedem: our jobs are no longer segfaulting n-cond as of this morning, so nfi what was causing it :/21:06
*** tidwellr has joined #openstack-nova21:06
mriedemjroll: huh21:06
jrollproves "have a beer and try again tomorrow" is always a good plan21:07
jackie-truongmriedem: Correct. Should the depends-on tag be used for functionality deps, too?21:07
*** bpoulos has joined #openstack-nova21:07
*** esberglu has quit IRC21:08
*** esberglu has joined #openstack-nova21:09
sean-k-mooneyjroll: hum well it works. kind of hoped you would get to the bottom of that one21:09
jrollsean-k-mooney: yeah, me too, but is nice to have it off my back21:10
mriedemjackie-truong: you don't need a depends-on when it's in the same repo, just stack the changes in a series21:10
mriedemobject > utils > API21:10
mriedemthe API changes should always be the last thing to go21:11
sean-k-mooneymriedem: logic being dont expose the feature at the api until its fully done right21:11
mriedemsean-k-mooney: yes that's generally how things should work21:12
*** awaugama has quit IRC21:12
mriedembecause we support CD,21:12
mriedemso that API can be in the wild once it's merged21:12
mriedemand we have to support it21:12
*** suresh12 has quit IRC21:14
*** matrohon has quit IRC21:14
jackie-truongmriedem: Got it. bpoulos will fix shortly21:14
bpoulosmriedem: I'll update the API order.  I think we chose the way we did because that was the order it was listed in in the spec (at )21:14
*** suresh12 has joined #openstack-nova21:14
*** suresh12 has quit IRC21:19
*** jangutter has joined #openstack-nova21:21
*** smatzek has quit IRC21:21
*** yamahata has quit IRC21:21
*** slaweq has joined #openstack-nova21:21
*** slaweq has quit IRC21:22
*** dave-mccowan has quit IRC21:25
*** jangutter has quit IRC21:26
openstackgerritBrianna Poulos proposed openstack/nova master: Add trusted_certs object
openstackgerritBrianna Poulos proposed openstack/nova master: Implement certificate_utils
openstackgerritBrianna Poulos proposed openstack/nova master: Add trusted_image_certificates to REST API
*** pchavva has quit IRC21:33
*** weshay|mtg is now known as weshay|rover21:34
*** threestrands has joined #openstack-nova21:41
*** threestrands_ has joined #openstack-nova21:44
*** threestrands_ has quit IRC21:44
*** threestrands_ has joined #openstack-nova21:44
*** threestrands has quit IRC21:47
openstackgerritBrianna Poulos proposed openstack/nova master: Add trusted_certs object
openstackgerritBrianna Poulos proposed openstack/nova master: Implement certificate_utils
*** arvindn051 has joined #openstack-nova21:54
*** eharney has quit IRC21:54
openstackgerritBrianna Poulos proposed openstack/nova master: Add trusted_image_certificates to REST API
*** eharney has joined #openstack-nova21:56
*** smatzek has joined #openstack-nova21:56
*** eharney has quit IRC21:59
openstackgerritBrianna Poulos proposed openstack/nova master: Add trusted_certs object
openstackgerritBrianna Poulos proposed openstack/nova master: Implement certificate_utils
*** belmoreira has quit IRC22:00
openstackgerritBrianna Poulos proposed openstack/nova master: Add trusted_image_certificates to REST API
mriedemtssurya_: dansmith: melwitt: L74
*** smatzek has quit IRC22:01
mriedemupdated to summarize some of the meeting notes22:01
melwittthanks for doing that22:01
* dansmith has to run off now22:02
tssurya_mriedem : thanks for putting it all in one place,22:04
*** tssurya_ has quit IRC22:05
*** mvenesio has quit IRC22:05
jackie-truongmriedem: dependency orders are now fixed. sorry about that22:07
*** lyan has quit IRC22:11
*** amodi has joined #openstack-nova22:11
*** rcernin has joined #openstack-nova22:14
*** jangutter has joined #openstack-nova22:15
*** yangyapeng has quit IRC22:16
*** suresh12 has joined #openstack-nova22:18
mriedemlyarwood: is the native luks swap volume thing b/c of blockRebase? and if so, is that then also an issue for guest-assisted snapshot of encrypted volumes for volumes like NFS/gluster?22:19
*** jangutter has quit IRC22:19
mriedemi don't know if we even have a tempest test for creating a snapshot of an instance with an encrypted volume attached, probably not22:20
mriedemand if we did, it would only be run in the NFS CI22:20
*** suresh12 has quit IRC22:22
openstackgerritMatt Riedemann proposed openstack/nova master: Move remaining uses of parted to privsep.
*** amodi has quit IRC22:27
*** rcernin has quit IRC22:29
mriedemgmann: do you plan on doing anything with these WIP changes?
*** rcernin has joined #openstack-nova22:31
mriedemgmann: otherwise once the 2 patches from alex_xu are merged i'm going to complete the blueprint for queens22:31
*** jackie-truong has quit IRC22:31
mriedemor i can just defer to rocky22:32
mriedemefried: i assume we'll just defer this to rocky
efriedmriedem Yeah.  I deemed nrp more important and didn't spend the cycles to investigate why that one's... "special".22:34
mriedemthat's fine22:34
*** suresh12 has joined #openstack-nova22:35
bpoulosmriedem lyarwood: the latest encrypted volume tempest tests are at and don't involve any snapshots22:36
*** jaypipes has joined #openstack-nova22:39
mriedem#success osc-placement 1.0.0 released; you can now do things with resource providers/classes via OSC CLI now22:39
openstackstatusmriedem: Added success to Success page22:39
mriedemmnaser: ^22:39
mriedemyou don't have to hack the db directly as much!22:39
melwitt"as much" \o/22:41
openstackgerritOpenStack Release Bot proposed openstack/osc-placement master: Update reno for stable/queens
*** acormier has quit IRC22:43
*** edmondsw has quit IRC22:47
*** acormier has joined #openstack-nova22:49
*** bpoulos has quit IRC22:49
*** claudiub has quit IRC22:50
*** itlinux has quit IRC22:51
*** esberglu has quit IRC22:56
*** yangyapeng has joined #openstack-nova23:01
*** tidwellr has quit IRC23:02
*** itlinux has joined #openstack-nova23:03
*** suresh12 has quit IRC23:05
*** eharney has joined #openstack-nova23:08
*** jangutter has joined #openstack-nova23:09
*** jangutter has quit IRC23:13
*** yamahata has joined #openstack-nova23:17
*** mlavalle has quit IRC23:23
*** itlinux has quit IRC23:23
*** acormier has quit IRC23:24
*** acormier has joined #openstack-nova23:24
*** takashin has joined #openstack-nova23:28
*** suresh12 has joined #openstack-nova23:29
*** derekh_afk has quit IRC23:33
*** acormier has quit IRC23:40
efriedjaypipes Hi, didn't see you come in.23:40
*** armax has quit IRC23:43
openstackgerritMike Perez proposed openstack/nova master: Replace support matrix ext with common library
*** suresh12 has quit IRC23:50
*** suresh12 has joined #openstack-nova23:51
jaypipesefried: sorry been out of network access for a while23:52
efriedjaypipes You missed a rather hearty debate on the extent to which the internals of generation management should be a part of the placement API spec.23:53
efriedjaypipes TL;DR: does the client get to know that the generation is a monotonically increasing integer starting at zero; or should the generation be completely opaque to the client?23:54
*** yassine has joined #openstack-nova23:54
*** yassine is now known as Guest1409423:55
*** hongbin has quit IRC23:56

Generated by 2.15.3 by Marius Gedminas - find it at!