Monday, 2015-02-16

*** kragniz_ is now known as kragniz00:00
*** smoriya has joined #openstack-ironic00:02
*** romcheg has quit IRC00:13
*** naohirot has joined #openstack-ironic00:27
openstackgerritNaohiro Tamura proposed openstack/ironic: Fix ml2_conf.ini settings  https://review.openstack.org/15609200:29
*** Marga_ has joined #openstack-ironic01:12
*** maurosr has quit IRC01:27
*** zz_jgrimm has quit IRC01:27
*** oomichi has joined #openstack-ironic01:28
*** achanda has joined #openstack-ironic01:28
*** zz_jgrimm has joined #openstack-ironic01:31
*** maurosr has joined #openstack-ironic01:32
*** achanda has quit IRC01:33
*** Marga__ has joined #openstack-ironic01:52
*** Marga_ has quit IRC01:55
*** lazy_prince has quit IRC02:00
*** anderbubble has joined #openstack-ironic02:06
*** anderbubble has quit IRC02:28
openstackgerritYuiko Takada proposed stackforge/ironic-discoverd: Support i18n  https://review.openstack.org/15610402:37
*** ramineni has joined #openstack-ironic02:42
*** achanda has joined #openstack-ironic02:52
*** jerryz has joined #openstack-ironic03:09
*** lazy_prince has joined #openstack-ironic03:15
*** mjturek1 has quit IRC03:19
*** achanda has quit IRC04:00
openstackgerritYuiko Takada proposed stackforge/ironic-discoverd: Support i18n part2  https://review.openstack.org/15611504:07
openstackgerritYuiko Takada proposed stackforge/ironic-discoverd: Support i18n part3  https://review.openstack.org/15611904:30
*** pradipta has joined #openstack-ironic04:40
*** pensu has joined #openstack-ironic04:54
*** achanda has joined #openstack-ironic05:00
*** achanda has quit IRC05:06
openstackgerritYuiko Takada proposed stackforge/ironic-discoverd: Support i18n part2  https://review.openstack.org/15611505:45
openstackgerritNisha Agarwal proposed openstack/ironic: Automate boot iso creation with in ironic for iscsi-ilo  https://review.openstack.org/15590005:58
openstackgerritMerged stackforge/proliantutils: Add hpssa module to proliantutils  https://review.openstack.org/15351906:02
*** oomichi has quit IRC06:13
*** achanda has joined #openstack-ironic06:13
*** marios_ has joined #openstack-ironic06:27
*** rameshg87 has joined #openstack-ironic06:32
openstackgerritOpenStack Proposal Bot proposed openstack/ironic: Imported Translations from Transifex  https://review.openstack.org/15597006:32
openstackgerritShivanand Tendulker proposed openstack/ironic: Ilo drivers sets capabilities:boot_mode in node  https://review.openstack.org/15573106:32
*** faizan has joined #openstack-ironic06:33
openstackgerritNisha Agarwal proposed openstack/python-ironicclient: enhanced node-set-provision-state  https://review.openstack.org/14880406:42
*** yog__ has joined #openstack-ironic06:42
*** achanda has quit IRC06:45
*** lazy_prince has quit IRC06:49
*** dlpartain has joined #openstack-ironic06:49
dtantsurMorning Ironic!06:49
*** killer_prince has joined #openstack-ironic06:50
*** killer_prince is now known as lazy_prince06:50
*** dlpartain has left #openstack-ironic06:53
*** achanda has joined #openstack-ironic07:04
*** andreykurilin_ has joined #openstack-ironic07:13
*** Marga__ has quit IRC07:17
*** pradipta has quit IRC07:31
*** andreykurilin_ has quit IRC07:35
openstackgerritDmitry Tantsur proposed stackforge/ironic-discoverd: Don't wait for too long for IPMI credentials update  https://review.openstack.org/15614407:51
*** chlong has quit IRC07:51
*** yog__ has quit IRC07:57
*** kbs has quit IRC08:07
*** jistr has joined #openstack-ironic08:08
*** jcoufal has joined #openstack-ironic08:10
*** kbs has joined #openstack-ironic08:10
*** yog__ has joined #openstack-ironic08:11
*** achanda has quit IRC08:18
*** gillllliard is now known as gilliard08:18
*** yuanying has quit IRC08:24
*** kbs has quit IRC08:27
*** achanda has joined #openstack-ironic08:28
*** achanda has quit IRC08:35
*** vdrok_afk is now known as vdrok08:36
vdrokmorning ironic08:36
vdrokmorning dtantsur08:36
*** kbs has joined #openstack-ironic08:38
yuriyzmorning Ironic08:39
*** achanda has joined #openstack-ironic08:40
*** ifarkas has joined #openstack-ironic08:47
openstackgerritAnusha Ramineni proposed openstack/ironic: Update Ilo drivers to use REST API interface to iLO  https://review.openstack.org/15440308:50
openstackgerritVladyslav Drok proposed openstack/ironic: Support for non-Glance image references  https://review.openstack.org/13674108:52
openstackgerritDmitry Tantsur proposed stackforge/ironic-discoverd: Also set IPMI address if it's not set already  https://review.openstack.org/15615608:55
*** jiangfei has quit IRC08:57
dtantsurifarkas, hey, could you review this patch chain ^^^ ?08:57
*** openstackgerrit has quit IRC08:57
*** openstackgerrit has joined #openstack-ironic08:58
ifarkasdtantsur, hi, sure, I will do it probably later today08:58
*** viktors has joined #openstack-ironic08:58
*** ndipanov has joined #openstack-ironic08:59
*** jistr has quit IRC08:59
*** lucasagomes has joined #openstack-ironic09:01
*** achanda has quit IRC09:02
*** zhenzanz has quit IRC09:02
*** romcheg has joined #openstack-ironic09:04
*** kbs1 has joined #openstack-ironic09:07
dtantsurack09:08
*** kbs has quit IRC09:08
openstackgerritRamakrishnan G proposed openstack/ironic: Add driver interface for RAID configuration  https://review.openstack.org/15523009:09
*** lazy_prince is now known as killer_prince09:12
*** kbs1 has quit IRC09:12
*** achanda has joined #openstack-ironic09:12
*** kbs has joined #openstack-ironic09:16
*** lsmola has joined #openstack-ironic09:20
*** achanda has quit IRC09:23
*** kbs has quit IRC09:26
*** mgoddard has joined #openstack-ironic09:31
*** pas-ha has joined #openstack-ironic09:37
*** erwan_taf has quit IRC09:43
*** erwan_taf has joined #openstack-ironic09:43
*** kbs has joined #openstack-ironic09:44
*** athomas has joined #openstack-ironic09:46
*** pas-ha has quit IRC09:50
openstackgerritAnusha Ramineni proposed openstack/ironic: Add tests for ilo_deploy driver  https://review.openstack.org/14298509:52
*** sambetts has joined #openstack-ironic09:53
*** erwan_taf has quit IRC10:03
*** pas-ha has joined #openstack-ironic10:05
*** Nisha has joined #openstack-ironic10:08
Nishadtantsur, hi10:08
dtantsuro/10:08
Nishadtantsur, regarding InspectInterface's method inspect_hardware() return value10:10
dtantsurNisha, https://github.com/openstack/ironic/blob/master/ironic/conductor/manager.py#L1502-L1508 may answer your question :)10:10
dtantsurthat's what we do when we need to handle both in-band and OOB10:11
*** ifarkas has quit IRC10:11
Nishadtantsur, but in case is it INSPECTING, we dont have a wait state?10:12
dtantsurNisha, we can use INSPECTING for that. So if new_state == INSPECTING, just wait10:12
Nishadtantsur, means to say we dont have any verb as wait10:13
Nishawhat action shall i do when it is inspecting10:13
dtantsurNisha, just return10:13
dtantsurand then either driver will handle finishing or it will time out eventually10:13
*** ifarkas has joined #openstack-ironic10:14
Nishaplus it means that we will either have to remove the transition from INSPECTED -> INSPECTING or add a transition from INSPECTING->MANAGEABLE10:14
Nishahow does conductor knows and put the state correctly in provision_state10:15
Nishasay if driver handles it, then conductor must know that it has successfully done10:16
Nishato correctly reflect the state10:16
dtantsurNisha, why&10:16
dtantsur?10:16
dtantsur(and there should not be transition from INSPECTED -> INSPECTING, only the opposite)10:17
Nishasorry that was a typo....10:17
Nishaok if conductor just returns in case it is inband, what it shall set in provision state and target provision state10:18
*** erwan_taf has joined #openstack-ironic10:18
Nishaor it should leave the states as it is?10:18
dtantsurNisha, yes, please leave them. Driver must handle transitions in this case.10:19
Nishaohk. and i was updating the last_inspected in case of successful inspection when transiting from INSPECTING to INSPECTED10:20
Nishawill the driver take care of it also?10:21
dtantsuryes, it has to10:22
*** pradipta has joined #openstack-ironic10:25
*** MattMan has joined #openstack-ironic10:29
*** killer_prince is now known as lazy_prince10:30
*** pelix has joined #openstack-ironic10:41
*** Nisha has quit IRC10:45
openstackgerritDmitry Tantsur proposed stackforge/ironic-discoverd: Update functest to new ramdisk code  https://review.openstack.org/15617910:49
rameshg87dtantsur, hi, one question10:58
dtantsuryep10:58
rameshg87dtantsur, https://review.openstack.org/#/c/155230/4/ironic/api/controllers/v1/node.py10:58
rameshg87dtantsur, why not use body=types.jsontype instead of body={wtypes.text: types.jsontype}10:59
rameshg87dtantsur, is body={wtypes.text: types.jsontype}  better than  body=types.jsontype ?10:59
dtantsurrameshg87, if you're ok with body being e.g. a list - yes :)10:59
rameshg87dtantsur, oh11:00
rameshg87dtantsur, okay .. got it11:00
rameshg87dtantsur, thanks .. :)11:00
dtantsurnp)11:00
rameshg87dtantsur, to a certain level schema does the same validation too11:01
dtantsuryeah, you can rely on it too11:01
rameshg87dtantsur, don't know how much we want to catch the error in wsme11:01
rameshg87dtantsur, because schema error messages seem to make more sense11:01
rameshg87dtantsur, schema can call out 'logical_disks' is missing :)11:02
dtantsurI'm find with both, though assumption that top-level is a dict sounds fine to me too11:02
rameshg87dtantsur, okay. i think i will go with body=types.jsontypes for now ..11:02
dtantsurI mean, if user provides a list, it's useless to do schema validation at all :)11:02
dtantsurack11:02
rameshg87dtantsur, yeah it would require conductor to do it for you11:03
rameshg87dtantsur, instead of catching it in api11:03
rameshg87dtantsur, that's a difference11:03
dtantsurok11:03
*** pensu has quit IRC11:04
*** marios has quit IRC11:06
*** ramineni has quit IRC11:10
rameshg87lucasagomes, can you please have a look at inband raid configuration spec: https://review.openstack.org/#/c/147803/11:14
*** marios has joined #openstack-ironic11:16
openstackgerritRamakrishnan G proposed openstack/ironic: Add driver interface for RAID configuration  https://review.openstack.org/15523011:17
*** rameshg87 has quit IRC11:18
*** athomas has quit IRC11:23
*** derekh has joined #openstack-ironic11:23
*** yuikotakada has joined #openstack-ironic11:29
yuikotakadaGood evening, Ironic :)11:30
openstackgerritVictor Sergeyev proposed openstack/ironic: Run tests in py33 environment  https://review.openstack.org/15619211:33
lucasagomesyuikotakada, g'evening11:38
yuikotakadalucasagomes, Hi, long time no see :D11:41
lucasagomesyuikotakada, long time indeed! How are you?11:44
*** athomas has joined #openstack-ironic11:45
dtantsuryuikotakada, g'evening!11:48
yuikotakadalucasagomes, so fine! and you?? you also look fine because you're contributing active!11:49
yuikotakadadtantsur, Hi! good evening! And thank you for reviewing my patch, I'm fixing now :D11:50
dtantsuryeah, thanks a lot for working on it. I would never find time :)11:50
sambettsMorning dtantsur, lucasagomes11:52
dtantsurmorning!11:53
lucasagomesyuikotakada, all is well over there (:11:53
lucasagomessambetts, yo, morning11:53
lucasagomess/there/here11:53
yuikotakadadtantsur, I will contribute more and more, so let's work hard together ;)11:54
yuikotakadalucasagomes, nice :) Don't you have any other job except for OpenStack contribution?11:57
*** andreykurilin_ has joined #openstack-ironic11:59
lucasagomesyuikotakada, I work mostly upstream on openstack, like 80% of the time I'd say11:59
dtantsuryuikotakada, awesome \o/12:01
dtantsur:)12:01
yuikotakadalucasagomes, Do you find your task by yourself? Or assigned by your company?12:04
lucasagomesyuikotakada, hmm both I would say, I try to align what needs to be done upstream to what the company needs too12:05
yuikotakadadtantsur, there are a lot of things which are first time to me, so I might trouble you, but I'll do my best12:05
dtantsurno problem at all :)12:08
yuikotakadalucasagomes, it's ideal :) I want to be so12:11
lucasagomesyuikotakada, yeah, sometimes can be trick too. But overall it's working fine12:12
yuikotakadadtantsur, is there any plan to merge ironic-discoverd to Ironic? I don't know the circumstances why ironic-discoverd is an independent stackforge project.12:14
dtantsuryuikotakada, I think people are pretty ok with them being separate. It's like that because back in Juno cycle we decided that it's too tricky to implement it in Ironic tree. Now we have plans to _use_ ironic-discoverd from within Ironic: http://specs.openstack.org/openstack/ironic-specs/specs/kilo/inband-properties-discovery.html12:15
*** pensu has joined #openstack-ironic12:17
dtantsurit's like Ironic was separated from Nova (it used to be nova-baremetal thingy): it's too complex on it's own already :)12:17
dtantsurlucasagomes, wdyt on ^^^12:17
lucasagomesdtantsur, hmm, haven't thought much about it. For now at least I think it's fine being a separated project (perhaps we should move it under openstack/ later on)12:20
dtantsuryeah, people tend to think too much about stackforge/ prefix :( while actually it has 0 essential difference from openstack/ one12:20
dtantsuranyway we have to ask devananda :)12:21
lucasagomesyeah, just a namespace12:24
*** smoriya has quit IRC12:25
*** andreykurilin_ has quit IRC12:25
*** andreykurilin_ has joined #openstack-ironic12:26
sambettsright now I think that the stackforge is being overused for some things tbh its getting pretty full, for instance the neutron plugin/driver decomposition things, I think should have a different namespace12:26
yuikotakadadtantsur, I'm not sure weather implementing ironic-discoverd in Ironic tree is tricky or not... but it looks like a plugin :)12:26
dtantsursambetts, do you think discoverd belongs to openstack/ or stackforge/ ?12:27
dtantsuryuikotakada, I guess the biggest concern was a way how discoverd enables PXE booting. Ironic is using Neutron, discoverd can't use Neutron here and uses firewall to rule statically configured DHCP.12:28
dtantsuryuikotakada, also now discoverd has it's own plugins, so it would be plugin with plugins :D12:28
sambettsdtantsur: I would like to see a different namespace tbh, like a related projects one if its not core but also not a experimental project either12:31
dtantsursambetts, I think it's exactly what stackforge is about :) experimental projects should leave in their author's github IMO12:32
yuikotakadadoes "experimental" mean "not stable"? if so, I understand. This feature itself is very useful and attractive, so everyone who are using Ironic want to use12:35
*** kbyrne has joined #openstack-ironic12:37
dtantsurright, a lot of people requested it to be implemented. anyway, OpenStack governance rules are being changed this cycle, we'll see how it ends up.12:38
sambettsthe stackforge is full of young/undeveloped projects around OpenStack, which I think make it difficult to find projects like ironic-discoverd12:38
dtantsurhmm... maybe. though e.g. Fuel lives in stackforge as well, and it must be a pretty mature project12:39
sambettsyeah, thinks like that I personally feel need a non-core but graduated namespace to live in12:41
sambettss/thinks/things12:41
sambettsbut I guess having one big area makes it much easier on infra12:42
*** dprince has joined #openstack-ironic12:42
dtantsurit's also a good point12:42
*** derekh has quit IRC12:46
*** derekh has joined #openstack-ironic12:46
*** dprince_ has joined #openstack-ironic12:47
*** Marga_ has joined #openstack-ironic12:47
*** dprince has quit IRC12:47
*** humble_ has joined #openstack-ironic12:51
*** pas-ha has quit IRC12:57
*** lazy_prince is now known as killer_prince13:03
*** pas-ha has joined #openstack-ironic13:09
*** jcoufal_ has joined #openstack-ironic13:10
*** lucasagomes is now known as lucas-hungry13:10
*** jcoufal has quit IRC13:13
openstackgerritSirushti Murugesan proposed openstack/ironic: Adds support for deploying whole disk images  https://review.openstack.org/15014213:14
*** dprince_ has quit IRC13:14
*** dprince has joined #openstack-ironic13:14
*** EmilienM is now known as EmilienM|afk13:20
*** mlsoule has joined #openstack-ironic13:21
*** yog__ has quit IRC13:27
*** coolsvap- is now known as coolsvap13:27
*** mlsoule has left #openstack-ironic13:34
*** mjturek1 has joined #openstack-ironic13:44
*** lucas-hungry is now known as lucasagomes13:50
*** mjturek1 has quit IRC13:51
*** mjturek1 has joined #openstack-ironic13:52
openstackgerritVictor Sergeyev proposed openstack/ironic: Run tests in py33 environment  https://review.openstack.org/15619213:54
*** andreykurilin_ has quit IRC13:56
*** andreykurilin_ has joined #openstack-ironic13:56
*** Marga_ has quit IRC13:59
*** pensu has quit IRC14:03
*** dprince has quit IRC14:08
openstackgerritShivanand Tendulker proposed openstack/ironic: Common changes for secure boot support  https://review.openstack.org/15397414:11
openstackgerritShivanand Tendulker proposed openstack/ironic: Secure boot support for pxe_ilo driver  https://review.openstack.org/15480814:14
openstackgerritShivanand Tendulker proposed openstack/ironic: Secure boot support for iscsi_ilo driver  https://review.openstack.org/15481414:17
*** faizan has quit IRC14:18
openstackgerritShivanand Tendulker proposed openstack/ironic: Secure boot support for agent_ilo driver  https://review.openstack.org/15481614:20
*** rameshg87 has joined #openstack-ironic14:23
*** vdrok is now known as vdrok_afk14:27
*** yuikotakada has quit IRC14:33
openstackgerritLucas Alvares Gomes proposed openstack/ironic: BaseAgentVendor: Invoke the driver vendor interface from task  https://review.openstack.org/15623314:34
rameshg87lucasagomes, hi14:42
rameshg87lucasagomes, regarding https://review.openstack.org/#/c/156233/114:42
lucasagomesrameshg87, hi there14:43
lucasagomeswhat's up?14:43
rameshg87lucasagomes, i had noticed some kind of a problem like this when calling passthrus14:44
rameshg87lucasagomes, i could find this part of the code had some problem14:44
rameshg87lucasagomes, https://github.com/openstack/ironic/blob/master/ironic/drivers/base.py#L445-L45714:44
rameshg87lucasagomes, specifically https://github.com/openstack/ironic/blob/master/ironic/drivers/base.py#L456-L45714:45
lucasagomesrameshg87, right, you found it out when working on some iLO passthrus? like 2 methods with the same name ?14:45
* rameshg87 is figuring out the code ..14:47
lucasagomesyeah, it took me 2 hours + to figure that out14:47
lucasagomesmaybe it's something we can fix on the @passthru14:48
* lucasagomes looks more into it14:48
rameshg87lucasagomes, yes, i actually know some sort of a fix .. even i took a long time figuring out14:48
lucasagomesrameshg87, ah, if you can point me I can test it here14:48
rameshg87lucasagomes, we were trying to have an extended version of AgentVendorInterface for some reason14:49
*** dprince has joined #openstack-ironic14:49
rameshg87lucasagomes, we were trying to extend the method reboot_to_instance() in our IloAgentVendorInterface14:49
lucasagomesrameshg87, yeah, that's what I'm trying to do with the BaseAgentVendor14:49
lucasagomesrameshg87, I see14:49
rameshg87lucasagomes, we wanted heartbeat() to call IloAgentVendorInterface.reboot_to_instance()14:50
rameshg87lucasagomes, but instead it was always calling AgentVendorInterface.reboot_to_instance() because 'self' was actually wrong14:50
lucasagomesyeah that's exactly what I'm doing... reboot_to_instance(), continue_deploy() and deploy_is_done() is overwrittable14:50
rameshg87lucasagomes, okay14:51
lucasagomesrameshg87, yeah, exactly14:51
rameshg87lucasagomes, let me see if i can dig up that patch14:51
rameshg87lucasagomes, i have it somewhere14:51
lucasagomesrameshg87, cool, I will take a look too14:51
rameshg87lucasagomes, i wanted to a file a bug, but then i lost it14:51
lucasagomes:(14:51
rameshg87lucasagomes, i mean i forgot to do it14:51
rameshg87lucasagomes, it was last week or something14:51
lucasagomesI see, yeah it's super tricky14:52
lucasagomesI spend my morning trying to figure out what was going on14:52
jrollI mean, that's just python14:52
jrollright?14:52
lucasagomespretty much14:52
lucasagomeswe are overwritten __new__ which is a bit blackmagicish14:53
jrolllike, you can't make a thing call into it's subclass without a giant hack14:53
lucasagomesyeah14:53
rameshg87lucasagomes, just give me some time, let me see if i can get a patch which worked for me ..14:53
lucasagomesrameshg87, no worries14:55
lucasagomesto be honest I wasn't expecting people would inherit from other's driver's vendor interface14:55
lucasagomesbut yeah...14:55
lucasagomesjroll, morning btw14:56
*** killer_prince is now known as lazy_prince14:56
*** yog__ has joined #openstack-ironic14:56
jrolloh hey morning lucasagomes :D14:56
*** jjohnson2 has joined #openstack-ironic15:01
dtantsurjroll, o/15:01
jrollmorning dtantsur rameshg87 :)15:02
rameshg87jroll, o/15:02
*** humble_ has quit IRC15:03
*** zz_jgrimm is now known as jgrimm15:04
*** ukalifon1 has joined #openstack-ironic15:08
openstackgerritRamakrishnan G proposed openstack/ironic: Vendorpassthru doesn't get correct 'self'  https://review.openstack.org/15624415:10
rameshg87lucasagomes, ^^^15:10
rameshg87lucasagomes, let me know what you think and if it solves the problem that you were facing15:10
lucasagomesrameshg87, cool, lemme test it15:11
rameshg87lucasagomes, the problem i found was that metadata['func'] was getting overwritten at some point later that is not in Ironic's control because (i think) we used up the reference that functools gave15:11
rameshg87lucasagomes, i just copied the dict to my own and modified that, and it seemed to solve the problem15:12
*** Nisha has joined #openstack-ironic15:12
rameshg87lucasagomes, but i am not sure why/where it was getting modified (again not inside Ironic, that's what i was able to find)15:13
*** beekneemech is now known as bnemec15:13
lucasagomesrameshg87, right, thanks for that :D15:13
jrollman, bmcs are the worst15:13
*** rameshg87 is now known as rameshg87-dinner15:14
lucasagomesjroll, lol whats up?15:15
jrolllucasagomes: not this BMC :D15:15
NobodyCamGood mornig Ironicer15:15
NobodyCamlol Ironicers even15:15
lucasagomesNobodyCam, yo15:15
NobodyCammorning lucasagomes :)15:15
jrollmorning NobodyCam15:15
NobodyCammorning jroll15:15
*** EmilienM|afk is now known as EmilienM15:16
NishaNobodyCam, morning15:16
NobodyCammorning Nisha :)15:16
lucasagomesrameshg87-dinner, as a first stab, ur patch seems to fix the problem yes15:17
lucasagomesrameshg87-dinner, it makes sense, because on the base class the _vendor_metadata will exist on the second time it's being instantiate15:19
lucasagomesso it will add that metadata on the new instance15:19
lucasagomesrameshg87-dinner, I will abandon my patch in favor of ur's15:20
*** rameshg87-dinner is now known as rameshg8715:22
lucasagomesbut the metadata comes from the first instance15:22
rameshg87lucasagomes, but this seems to handle that https://github.com/openstack/ironic/blob/master/ironic/drivers/base.py#L45615:23
rameshg87lucasagomes, pointing 'func' to the correct inherited method15:23
lucasagomesrameshg87, metadata is a dict15:24
lucasagomeswe are passing the reference  inst.vendor_routes.update({vmeta.method: vmeta.metadata})15:24
lucasagomesso if later on that dict is modified, it will modify it across all instances15:24
lucasagomesthat's why copy() makes sense15:24
*** naohirot has quit IRC15:24
lucasagomesso each vendor_routes will get a different copy15:24
rameshg87lucasagomes, but when is vmeta.metadata dict modified ?15:25
lucasagomesrameshg87, when the class is instantiated for the second time15:25
rameshg87lucasagomes, okay.15:26
rameshg87lucasagomes, correct me if i am wrong15:26
rameshg87lucasagomes, this is the first time we instantiate it, right ? - https://github.com/openstack/ironic/blob/master/ironic/drivers/base.py#L44615:26
openstackgerritRyan Moore proposed openstack/ironic: Correctly rebuild the PXE file during takeover of ACTIVE nodes  https://review.openstack.org/15546015:26
openstackgerritRyan Moore proposed openstack/ironic: Don't write PXE config file during takeover  https://review.openstack.org/15625015:26
lucasagomesrameshg87, https://github.com/openstack/ironic/blob/master/ironic/drivers/agent.py#L4315:27
lucasagomeswe instantiate the same class many times when building the drivers15:27
rameshg87lucasagomes, oh15:27
lucasagomesbut, since they were always using the same func for all drivers15:27
lucasagomeswe never detected that15:27
rameshg87lucasagomes, so this was getting overwritten back to parent's func when it was instantiated next time15:28
lucasagomesbut in reallity they were all pointing to the func of the last instantiated object15:28
lucasagomesrameshg87, yes15:28
lucasagomesnow that we inherited from a base class15:28
lucasagomeswe can get that unexpected behavior15:28
lucasagomesbecause the func's are not the same anymore15:28
rameshg87lucasagomes, yeah, may be we need a strong test case for this ..15:29
rameshg87:)15:29
lucasagomesthat's why you could be accessing a func ref from driver X on driver Y15:29
lucasagomesrameshg87, yeah, that would be good15:29
rameshg87i mean unit test ..15:29
lucasagomesright, yeah, that would help15:29
lucasagomesif we instantiate a class twice, make sure the maps point to different references15:29
rameshg87lucasagomes, yeah .. and when we inherit checking if we actually call child's method (as opposed to parent's method)15:30
lucasagomesyeah15:30
rameshg87lucasagomes, i mean child's method gets called15:30
lucasagomesvery tricky :)15:30
rameshg87lucasagomes, indeed15:30
rameshg87:)15:30
lucasagomesbut cool to find those very hidden bugs :D15:30
*** anderbubble has joined #openstack-ironic15:31
lucasagomesapart from the part that I was thinking I was crazy15:31
rameshg87:)15:31
rameshg87lucasagomes, i will pursue this tomorrow then .. you are welcome to take over if you want to get something merged before that :)15:32
lucasagomesrameshg87, cool, yeah I'm still working on the IPA part. If I finish the tests etc for that today I take over15:32
lucasagomesrameshg87, thanks15:33
rameshg87lucasagomes, sure..thanks ..15:33
lucasagomesI will put a dependency on that patch anyway15:33
lucasagomesas a dependency*15:33
rameshg87okay15:34
rameshg87lucasagomes, NobodyCam, if you get some time, please have a look at inband raid configuration spec: https://review.openstack.org/#/c/147803/. it's parent got merged, waiting for the implementation of that interface :)15:37
*** Nisha has quit IRC15:38
*** rameshg87 has quit IRC15:47
*** Nisha has joined #openstack-ironic15:50
NobodyCamwill have a look.. but have to run and pick up the kids15:55
NobodyCambbiafm15:55
Nishajroll, hi15:56
jrollhey Nisha :)15:56
Nisha:)15:56
Nishajroll, please check this http://logs.openstack.org/04/148804/3/check/gate-python-ironicclient-python26/2485e51/console.html15:56
NishaI am sure this has nothing to do with my patch but my jenkins are failing even on latest branch.15:57
openstackgerritRamakrishnan G proposed openstack/ironic: Add driver interface for RAID configuration  https://review.openstack.org/15523015:58
*** jistr has joined #openstack-ironic15:58
jrollNisha: hrm15:58
* jroll takes a look15:59
*** david-lyle_afk is now known as david-lyle16:01
jrollNisha: only your new test is failing, you need to set args.config_drive = None in your test16:02
jrollNisha: otherwise args.config_drive is an instance of mock.Mock, which will pass the "if args.config_drive" check16:03
Nishaok16:04
Nishabut by defauly config_drive shall be none correct16:06
jrollin normal usage, yes, but since args is a mock it doesn't work that way16:07
*** jerryz has quit IRC16:08
*** pradipta has quit IRC16:09
*** jgrimm is now known as zz_jgrimm16:11
*** zz_jgrimm is now known as jgrimm16:12
lucasagomescheck-tempest-dsvm-ironic-parallel-nv is now voting!?!16:22
jrollo.o16:22
lucasagomeswtf16:22
lucasagomeshttps://review.openstack.org/#/c/156244/16:23
lucasagomesit has been failing for a long time, it def should be non-voting16:23
openstackgerritDmitry Tantsur proposed stackforge/ironic-discoverd: Getting ready for inclusion into projects.yaml  https://review.openstack.org/15626516:26
jrollI don't see why it would be voting in project-config :/16:26
*** zigo has quit IRC16:26
*** lazy_prince is now known as killer_prince16:28
*** zigo has joined #openstack-ironic16:31
*** jistr has quit IRC16:32
openstackgerritMerged stackforge/ironic-discoverd: Getting ready for inclusion into projects.yaml  https://review.openstack.org/15626516:37
*** Marga_ has joined #openstack-ironic16:37
lucasagomesjroll, using regex to determine what votes and what not across all projects sounds dangerous16:41
lucasagomeslol16:41
jroll:P16:41
* jroll bbl16:41
victor_lowtherGood morning, Ironic.16:44
*** pensu has joined #openstack-ironic16:45
*** absubram has joined #openstack-ironic16:46
*** davideagnello has joined #openstack-ironic16:46
openstackgerritNisha Agarwal proposed openstack/python-ironicclient: enhanced node-set-provision-state  https://review.openstack.org/14880416:47
victor_lowtherSo, I have this problem with node vendor passthru methods.16:48
victor_lowtherNamely, that there does not seem to be a way to have them return the HTTP status code of my choosing16:49
NobodyCammornign victor_lowther :)16:49
victor_lowtherThis is slightly annoying when I have to deal with an operation that can operate either synchronusly or asynchronusly debelding on the exact values it is operating on.16:50
openstackgerritNisha Agarwal proposed openstack/ironic: Add InspectInterface for node-introspection  https://review.openstack.org/14981716:50
victor_lowtherer, depending.16:50
NishaNeed reviews on inspection patches16:51
Nisha:(16:51
victor_lowtherifarkas: if you could give a review of https://review.openstack.org/#/c/150488/, that would be awesome.16:52
ifarkasvictor_lowther, will do!16:52
lucasagomesvictor_lowther, that may need some work, right now I think it only returns 200 or 202 depending if it's sync or async16:52
lucasagomesvictor_lowther, we could include another metadata on the @passthru decorators16:53
lucasagomesincluding the return code if success16:53
lucasagomesvictor_lowther, and morning :)16:53
victor_lowtherThat only gets me part of the way there16:55
lucasagomesvictor_lowther, https://github.com/openstack/ironic/blob/master/ironic/api/controllers/v1/node.py#L72516:55
lucasagomesvictor_lowther, https://github.com/openstack/ironic/blob/master/ironic/conductor/manager.py#L48516:56
*** ukalifon1 has quit IRC16:56
victor_lowtherMy specific issue is that some BIOS settings on Dell systems require a reboot (and hence require creating a config job to acutally set), and some can be set immediately.16:58
victor_lowtherSo the set_config function in https://review.openstack.org/#/c/150488/13/ironic/drivers/modules/drac/bios.py,unified needs to be able to indicate both conditions16:58
victor_lowtherI am doing that with a boolean that indicates whether a BIOS config job needs to be scheduled or not.16:59
victor_lowtherI would rather do it with real HTTP status codes17:00
lucasagomesright17:02
*** coolsvap is now known as coolsvap_17:14
*** athomas has quit IRC17:25
*** kbs has quit IRC17:25
*** pcrews has quit IRC17:25
*** dtantsur has quit IRC17:25
*** tonyb has quit IRC17:25
*** vdrok_afk has quit IRC17:25
*** Madasi has quit IRC17:25
*** JoshNang has quit IRC17:25
*** keekz has quit IRC17:25
*** Marga_ has quit IRC17:31
*** athomas has joined #openstack-ironic17:31
*** kbs has joined #openstack-ironic17:31
*** pcrews has joined #openstack-ironic17:31
*** dtantsur has joined #openstack-ironic17:31
*** tonyb has joined #openstack-ironic17:31
*** vdrok_afk has joined #openstack-ironic17:31
*** Madasi has joined #openstack-ironic17:31
*** JoshNang has joined #openstack-ironic17:31
*** keekz has joined #openstack-ironic17:31
*** mkerrin has quit IRC17:39
*** andreykurilin_ has quit IRC17:40
openstackgerritNisha Agarwal proposed openstack/ironic: Generic changes for Node Inspection  https://review.openstack.org/14982317:42
*** jrist has quit IRC17:46
*** romcheg has quit IRC17:49
*** romcheg has joined #openstack-ironic17:49
openstackgerritChris Krelle proposed openstack/ironic: Correctly rebuild the PXE file during takeover of ACTIVE nodes  https://review.openstack.org/15546017:51
openstackgerritPavlo Shchelokovskyy proposed openstack/ironic: Fix file permissions in project  https://review.openstack.org/15630517:52
ShrewsNobodyCam: anyone had a chance to test 155460 yet?17:54
*** pas-ha has quit IRC17:56
*** dtantsur is now known as dtantsur|afk17:56
*** jcoufal_ has quit IRC17:57
*** jrist has joined #openstack-ironic17:59
*** MattMan has left #openstack-ironic18:01
*** Marga_ has joined #openstack-ironic18:03
*** athomas has quit IRC18:05
*** EmilienM is now known as EmilienM|afk18:07
openstackgerritLucas Alvares Gomes proposed openstack/ironic: Add AgentVendorPasshtru for the PXE driver  https://review.openstack.org/15572818:14
*** coolsvap_ is now known as coolsvap18:25
*** romcheg has joined #openstack-ironic18:28
*** pensu has quit IRC18:29
sambettsnight ironic18:35
lucasagomessambetts, night18:36
lucasagomesI'm going too, see y'all tomorrow18:36
*** lucasagomes is now known as lucas-dinner18:36
NobodyCamnight lucas-dinner18:39
NobodyCam:)18:40
*** andreykurilin_ has joined #openstack-ironic18:46
*** ifarkas has quit IRC18:49
openstackgerritJarrod Johnson proposed stackforge/pyghmi: Handle concurrent session requests  https://review.openstack.org/15632718:50
*** anderbubble has quit IRC18:52
openstackgerritMerged stackforge/pyghmi: Handle concurrent session requests  https://review.openstack.org/15632718:54
*** anderbubble has joined #openstack-ironic18:56
NobodyCamShrews: fyi https://review.openstack.org/#/c/15625018:59
ShrewsNobodyCam: eh?19:03
NishaNobodyCam, hi19:08
NishaNobodyCam, i reposted https://review.openstack.org/#/c/141012/ after comment from nova reviewer. Could you rereview it?19:08
NobodyCamhi Nisha :)19:10
NobodyCamwill have a look19:10
NishaNobodyCam, thanks.19:10
NobodyCamShrews: thats related to 15546019:10
NobodyCam:)19:10
NishaNobodyCam, i need reviews on inspection patches also :)19:10
*** pelix has quit IRC19:24
*** coolsvap is now known as coolsvap_19:31
*** lucas-dinner has quit IRC19:36
*** sdake is now known as sdake_ws19:37
openstackgerritAdam Gandelman proposed openstack/ironic: Use separate set of credentials for keystone service user  https://review.openstack.org/15634419:38
openstackgerritNisha Agarwal proposed openstack/python-ironicclient: enhanced node-set-provision-state  https://review.openstack.org/14880419:40
*** romcheg1 has joined #openstack-ironic19:53
*** romcheg has quit IRC19:54
*** achanda has joined #openstack-ironic19:58
NobodyCamNisha: did you see your nova patch was blocked?20:02
Nishais it?20:03
*** dprince has quit IRC20:03
NobodyCamJohn Garbutt -2'd it20:03
*** EmilienM|afk is now known as EmilienM20:04
Nisha:(20:04
Nishais there no way we can get that in?20:05
Nishait is such a simple patch20:05
Nishathey are running thru the list now20:08
Nishaon irc20:08
Nisha:(20:08
*** achanda has quit IRC20:09
NobodyCam:(20:13
*** dprince has joined #openstack-ironic20:15
NobodyCambrb20:21
*** andreykurilin_ has quit IRC20:21
*** andreykurilin_ has joined #openstack-ironic20:21
mrdaMorning Ironic20:27
*** andreykurilin_ has quit IRC20:30
*** andreykurilin_ has joined #openstack-ironic20:31
*** achanda has joined #openstack-ironic20:45
*** devlaps has joined #openstack-ironic21:01
*** hongbin has joined #openstack-ironic21:05
hongbinHi, is anyone here familiar with the ironic nova scheduler?21:06
*** achanda has quit IRC21:06
jrollhongbin: someone surely is, ask your questions21:06
jrollmorning mrda :)21:06
mrdahey jroll21:07
hongbinjroll: I have a nova setup with several nova-compute. With one dedicated nova-compute for Ironic, the rest for VM21:08
hongbinI saw a VM request was unexpectedly scheduler to the Ironic host21:08
jrollhongbin: there's a thing you can do with host aggregates, but I don't know it well21:09
hongbinI checked the host aggregation filter21:09
hongbinIt seems I need to set the key of all flavors, which is suboptimal21:09
hongbinI wonder if there is a better way to achieve that21:10
jrollyeah, I'm not sure, sorry :(21:11
*** jgrimm is now known as zz_jgrimm21:12
hongbinjroll: NP :), Thanks for your response. I am awaiting for potential response from others21:12
jrollwelcome :)21:13
*** Nisha has quit IRC21:22
*** jcoufal has joined #openstack-ironic21:29
jrollwhee, nisha's thing got an ffe21:48
NobodyCam:) nice... jroll if it got a ffe you want to toss a +1 it way: https://review.openstack.org/#/c/14101221:52
jrolllol21:52
jrollI'll think about it :P21:52
NobodyCamhehehehe21:52
NobodyCam:-p21:52
*** andreykurilin_ has quit IRC21:56
*** achanda has joined #openstack-ironic22:34
openstackgerritJosh Gachnang proposed openstack/ironic: Implement Cleaning States  https://review.openstack.org/15344422:46
*** dprince has quit IRC22:47
*** absubram has quit IRC22:48
*** trown is now known as trown|outttypeww22:52
*** datajerk has quit IRC23:06
*** chlong has joined #openstack-ironic23:09
*** achanda has quit IRC23:21
NobodyCamJoshNang: jroll question? Was I wrong in thinking we were keeping deleted as a noop for k?23:26
JoshNangNobodyCam: heh i've heard both. i asked devananda and he said (correct me if i'm wrong) we'll implement them in L23:27
NobodyCam:/23:28
*** Marga_ has quit IRC23:28
NobodyCamI recall use chatting at the office23:28
*** Marga_ has joined #openstack-ironic23:28
NobodyCamI thought it was keep as noop.. but I'll try ask at deva at the meeting tonight (if he makes it)... he is still recovering from travel plage as i understand it23:30
*** anderbubble has quit IRC23:30
*** hongbin has left #openstack-ironic23:30
JoshNang:( ugh travel plague23:30
JoshNangbut ok! no-op'ing it is as easy as adding some more process_event()'s and adding to states.py23:31
NobodyCambefore you do it lets ask23:31
JoshNangdefinitely23:32
NobodyCambut good to know its a easy one ot add23:32
JoshNangyup! i had it in the cleaning review at one point :P23:32
NobodyCam:(23:33
NobodyCam:-p23:33
*** Marga_ has quit IRC23:35
*** Marga_ has joined #openstack-ironic23:35
*** romcheg has joined #openstack-ironic23:38
*** romcheg1 has quit IRC23:42
NobodyCamJoshNang: added as an agenda item23:52
*** Marga_ has quit IRC23:52
*** Marga_ has joined #openstack-ironic23:53
JoshNangNobodyCam: thanks!23:54
JoshNangi'll be there23:54
* NobodyCam hopes to be awake for it ... lol... 23:55
*** jcoufal_ has joined #openstack-ironic23:59

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