Monday, 2014-07-07

*** eghobo has quit IRC01:19
openstackgerritlifeless proposed a change to openstack/ironic: Simplify error handling  https://review.openstack.org/10505001:19
*** nosnos has joined #openstack-ironic01:37
*** ccrouch has quit IRC02:36
*** pcrews has quit IRC02:36
openstackgerritA change was merged to openstack/ironic: Simplify error handling  https://review.openstack.org/10505002:42
*** ccrouch has joined #openstack-ironic02:45
*** ramineni has joined #openstack-ironic02:49
*** coolsvap|afk is now known as coolsvap03:05
*** aswadr has joined #openstack-ironic03:09
*** nosnos has quit IRC03:25
openstackgerritDaryl Walleck proposed a change to openstack/ironic: Updates the Ironic on Devstack dev documentation  https://review.openstack.org/10446903:35
*** takadayuiko has joined #openstack-ironic03:41
*** Poornima has joined #openstack-ironic03:44
takadayuikolifeless: is here?03:57
lifelessI am03:59
takadayuikolifeless: Hi, thank you for reviewing my patch 73121. Sorry, I cannot understand what you means. I would like to talk about it.04:00
*** coolsvap is now known as coolsvap|afk04:00
*** radsy has quit IRC04:01
*** nosnos has joined #openstack-ironic04:02
lifelesstakadayuiko: I haven't reviewed 73121 - DB details leaking to API on error04:03
takadayuikoLifeless, Oops, I mistook. 104738.04:03
*** eghobo has joined #openstack-ironic04:07
lifelesstakadayuiko: ok, which bits are unclear?04:12
takadayuikolifeless: thank you. Do you mean that it's better to write neutron endpoint url as "http://192.168.0.10:9696" instead of "http" + "://" * "192.168.0.10" + ":" 9696 ?04:23
lifelesstakadayuiko: if it has to be written at all, yes.04:25
lifelesstakadayuiko: but thats opinion; if there is a consensus elsewhere, or a good reason to change - thats fine04:27
lifelesstakadayuiko: but - it needs to be backwards compatible; and it really shouldn't need to be set at all as keystone knows the neutron URL.04:27
takadayuikolifeless: OK,  I understood what you said. I myself think this method is better same as glance(glance_host, glance_port, glance_protocol)04:29
takadayuikolifeless: By changing the method of writing URL, backwards compatible will be lacked, right?04:31
*** eghobo has quit IRC04:37
lifelesssorry, whats the relevance of glance here?04:39
lifelessyes, if we stop supporting the url format its a backwards incompatible change04:39
lifelesswe should have a good reason to do that04:39
*** eghobo has joined #openstack-ironic04:41
takadayuikolifeless, Ah, there is no relevance. But the way of writing of Glance URL looks smart.04:41
takadayuikoI understood the importance of backwards incompatible. And my main purpose is setting my_ip as neutron URL instead of localhost. so the writing style is out of scope. I think I'll update my patch.04:44
*** k4n0 has joined #openstack-ironic04:45
takadayuikoMmm... but Lucas took this patch as the agenda of the next meeting. I think it's better to discuss about it together with everyone.04:47
*** sabah has joined #openstack-ironic04:49
*** ajc_ has joined #openstack-ironic04:52
*** geekyogi has joined #openstack-ironic04:58
lifelesstakadayuiko: I think the default should be to have no url set at all and read neutron out of keystone.04:58
openstackgerritAnusha Ramineni proposed a change to openstack/ironic-specs: Out-Of-Band firmware update for Baremetal Servers  https://review.openstack.org/10084204:58
takadayuikolifeless, about this your idea, I'm not sure, sorry. We should discuss about it.05:00
takadayuikoAnd I have to leave here. thank you lifeless! :)05:01
*** ramineni has quit IRC05:04
*** takadayuiko has quit IRC05:05
*** amitpp has joined #openstack-ironic05:07
*** coolsvap|afk is now known as coolsvap05:30
*** killer_prince is now known as lazy_prince05:35
*** geekyogi has quit IRC05:55
*** bvivek has joined #openstack-ironic06:00
openstackgerritOpenStack Proposal Bot proposed a change to openstack/ironic: Imported Translations from Transifex  https://review.openstack.org/10329106:02
*** lazy_prince is now known as killer_prince06:06
*** eghobo has quit IRC06:15
*** eghobo has joined #openstack-ironic06:15
*** eghobo has quit IRC06:16
*** pradipta_away is now known as pradipta06:23
*** bvivek has quit IRC06:39
*** bvivek has joined #openstack-ironic06:40
*** killer_prince is now known as lazy_prince06:43
*** shardy_afk is now known as shardy06:56
*** jcoufal has joined #openstack-ironic07:15
openstackgerritHaomeng,Wang proposed a change to openstack/ironic: Sync Oslo notifier module to Ironic  https://review.openstack.org/10507607:34
*** mkerrin has quit IRC07:41
*** mitz_ has quit IRC07:41
*** mkerrin has joined #openstack-ironic07:45
*** rakesh_hs has joined #openstack-ironic07:48
*** ndipanov has joined #openstack-ironic07:51
*** derekh_ has joined #openstack-ironic08:11
*** bvivek has quit IRC08:14
*** mitz_ has joined #openstack-ironic08:23
*** bvivek has joined #openstack-ironic08:26
*** lucasagomes has joined #openstack-ironic08:27
*** athomas has joined #openstack-ironic08:36
*** Poornima has quit IRC08:44
*** romcheg has joined #openstack-ironic08:51
*** coolsvap is now known as coolsvap|afk08:52
*** geekyogi has joined #openstack-ironic08:54
*** amitpp has quit IRC09:00
*** coolsvap|afk is now known as coolsvap09:04
*** lazy_prince is now known as killer_prince09:12
*** pelix has joined #openstack-ironic09:18
*** amitpp has joined #openstack-ironic09:23
*** martyntaylor has joined #openstack-ironic09:30
*** andreykurilin has quit IRC09:32
*** killer_prince is now known as lazy_prince09:32
romchegGood morning Ironic!09:38
*** athomas has quit IRC09:39
*** athomas has joined #openstack-ironic09:39
*** lazy_prince has quit IRC09:52
*** subah has joined #openstack-ironic10:08
*** sabah has quit IRC10:08
*** k4n0 has quit IRC10:13
*** romcheg has quit IRC10:16
*** sabah has joined #openstack-ironic10:22
*** subah has quit IRC10:22
*** k4n0 has joined #openstack-ironic10:26
*** k4n0 has quit IRC10:38
*** lazy_prince has joined #openstack-ironic10:38
*** overlayer has joined #openstack-ironic10:38
*** overlayer has quit IRC10:48
*** bvivek has quit IRC10:50
*** Haomeng|2 has joined #openstack-ironic10:51
*** Haomeng has quit IRC10:52
*** k4n0 has joined #openstack-ironic11:01
*** coolsvap is now known as coolsvap|afk11:08
*** pradipta is now known as pradipta_away11:15
*** coolsvap|afk is now known as coolsvap11:16
*** romcheg has joined #openstack-ironic11:29
*** geekyogi1 has joined #openstack-ironic11:29
*** k4n0 has quit IRC11:31
*** geekyogi has quit IRC11:32
*** lucasagomes is now known as lucas-hungry11:34
*** subah has joined #openstack-ironic11:43
openstackgerritGhe Rivero proposed a change to openstack/ironic: Initial oslo.i18n migration  https://review.openstack.org/10513211:44
*** subah has quit IRC11:46
*** subah has joined #openstack-ironic11:46
*** sabah has quit IRC11:46
*** coolsvap is now known as coolsvap|afk11:46
*** subah has quit IRC11:51
*** krtaylor_away is now known as krtaylor11:51
*** Haomeng has joined #openstack-ironic11:54
*** Haomeng|2 has quit IRC11:55
*** bvivek has joined #openstack-ironic12:03
*** jcoufal has quit IRC12:12
*** geekyogi1 has quit IRC12:12
*** ccrouch has left #openstack-ironic12:16
*** jcoufal has joined #openstack-ironic12:17
*** bvivek has quit IRC12:20
*** Shrews has quit IRC12:24
*** Shrews has joined #openstack-ironic12:25
*** lucas-hungry is now known as lucasagomes12:26
*** mgagne has quit IRC12:28
*** aweeks has quit IRC12:29
*** wendar has quit IRC12:29
*** wendar has joined #openstack-ironic12:29
*** coolsvapl has joined #openstack-ironic12:30
*** GheRiver1 has joined #openstack-ironic12:30
*** BadCub_ has joined #openstack-ironic12:30
*** coolsvap|afk has quit IRC12:30
*** antonym has quit IRC12:30
*** GheRivero has quit IRC12:31
*** yuriyz has quit IRC12:31
*** BadCub has quit IRC12:31
*** aweeks_ has joined #openstack-ironic12:31
*** mgagne has joined #openstack-ironic12:31
*** coolsvapl has quit IRC12:31
*** coolsvapl has joined #openstack-ironic12:31
*** aweeks_ has quit IRC12:31
*** aweeks_ has joined #openstack-ironic12:31
*** yuriyz has joined #openstack-ironic12:32
*** antonym has joined #openstack-ironic12:32
*** ajc_ has quit IRC12:33
*** k4n0 has joined #openstack-ironic12:44
*** k4n0 has quit IRC12:47
*** rakesh_hs has quit IRC12:56
*** nosnos has quit IRC13:02
openstackgerritLucas Alvares Gomes proposed a change to openstack/ironic: Add set_spawn_error_hook to TaskManager  https://review.openstack.org/10095713:04
openstackgerritLucas Alvares Gomes proposed a change to openstack/ironic: Fix nodes left in an incosistent state if no workers  https://review.openstack.org/10095813:04
*** amitpp has quit IRC13:05
openstackgerritLucas Alvares Gomes proposed a change to openstack/ironic: Fix nodes left in an incosistent state if no workers  https://review.openstack.org/10095813:08
*** jbjohnso_ has joined #openstack-ironic13:18
* romcheg written some changes to grenade13:21
romchegBut I have no idea what I just did, because there's no docs for this stuff :(13:23
openstackgerritOpenStack Proposal Bot proposed a change to openstack/ironic: Updated from global requirements  https://review.openstack.org/10516713:28
*** dhellmann_ is now known as dhellmann13:29
*** matty_dubs|gone is now known as matty_dubs13:43
NobodyCamgood morning ironic13:46
ShrewsNobodyCam: morning13:47
NobodyCam:) morning Shrews13:47
lucasagomesmorning romcheg NobodyCam Shrews13:51
*** dhellmann is now known as dhellmann_13:51
romchegMorning NobodyCam Shrews lucasagomes!13:51
Shrewso/13:51
matty_dubsHowdy fellers13:51
lucasagomesmorning matty_dubs13:51
* romcheg is about to start killing people around13:59
romchegWhen I was struggling with infra integration, I didn't know that there is Grenadeā€¦ :13:59
lucasagomeslol?13:59
romcheg:)13:59
NobodyCammorning lucasagomes romcheg and matty_dubs13:59
lucasagomesromcheg, ah thought it was neutron :P14:00
* matty_dubs steps back from IRC, out of romcheg's view14:00
lucasagomesromcheg, btw, it's now working for me14:00
lucasagomesromcheg, don't ask me what I didnt, I rebuilt my devstack env14:00
lucasagomesand now neutron works14:00
*** rloo has joined #openstack-ironic14:05
romcheglucasagomes: It keeps failing in random places for me :)14:06
lucasagomesromcheg, heh14:11
*** jgrimm has joined #openstack-ironic14:20
*** geekyogi has joined #openstack-ironic14:26
*** dhellmann_ is now known as dhellmann14:26
*** lazy_prince has quit IRC14:37
BadCub_Morning :-p14:38
NobodyCammorning BadCub_14:43
NobodyCamlucasagomes: love the var cat="meow"14:43
lucasagomesNobodyCam, lol14:43
NobodyCam:)14:44
NobodyCamlucasagomes: do you see _set_states_on_error getting used else where?14:50
lucasagomesNobodyCam, apart from deploy and tear down?14:51
* lucasagomes thinking14:52
lucasagomesNobodyCam, I don't think so because that's touch the provisioning state part14:53
NobodyCamI was woundering if it would be worth a name change, maybe something like: _set_node_state_after_error14:53
NobodyCam???14:53
NobodyCambut ... maybe not14:53
NobodyCamnot really sure if thats more clear or not14:54
NobodyCamgot a conf call.. now... bbiab14:55
lucasagomesNobodyCam, right, yeah naming things is not my thing really :(14:58
NobodyCam:) its a good name .. but may be a bit unclear14:59
NobodyCamI'm trying to think15:00
openstackgerritA change was merged to openstack/ironic: Updates the Ironic on Devstack dev documentation  https://review.openstack.org/10446915:01
*** dwalleck has joined #openstack-ironic15:05
*** killer_prince has joined #openstack-ironic15:13
*** killer_prince is now known as lazy_prince15:14
*** overlayer has joined #openstack-ironic15:19
*** jcoufal has quit IRC15:22
*** jcoufal has joined #openstack-ironic15:24
jrollmornin y'all15:29
matty_dubsHowdy jroll15:29
romcheglucasagomes: Did grenade fail for you because of version mismatch for six?15:30
NobodyCammorning jroll15:34
NobodyCamhave a good weekend15:34
jrollheya NobodyCam15:34
jrolltwas great, you?15:34
*** coolsvapl is now known as coolsvap15:37
*** overlayer has quit IRC15:41
*** aswadr has quit IRC15:42
NobodyCamwas nice...15:42
NobodyCam:)15:42
lucasagomesromcheg, where?15:43
romcheglucasagomes: grenade.sh when installing icehouse15:43
lucasagomesdidn't try it :/15:43
lucasagomeslemme check15:43
*** rwsu has joined #openstack-ironic15:47
matty_dubsInstanceDeployFailure: Disk partitioning failed on device /tmp/tmpXqo2W6/tmp24f0J7. Fuser exited with "Cannot stat file /proc/21639/fd/0: Permission denied15:55
*** Nisha has joined #openstack-ironic15:55
matty_dubsHaven't seen that before. Do I maybe need to update root_wrap?15:56
lucasagomesmatty_dubs, yes, ifarkas patch to use fuser to verify if the disk is not been held by another process was merged last friday15:57
lucasagomesso fuser should be in ur rootwrap15:57
matty_dubsOh, I see. I just need to update what's in git15:59
*** pcrews has joined #openstack-ironic15:59
*** matty_dubs is now known as matty_dubs|lunch16:01
*** jcoufal has quit IRC16:05
*** blamar has joined #openstack-ironic16:09
*** dwalleck has quit IRC16:11
*** coolsvap is now known as coolsvap|afk16:16
NobodyCambbt... brb16:22
*** derekh_ has quit IRC16:28
*** datajerk has quit IRC16:30
*** jcoufal has joined #openstack-ironic16:32
*** dwalleck has joined #openstack-ironic16:34
*** datajerk has joined #openstack-ironic16:35
*** martyntaylor has quit IRC16:37
*** ndipanov has quit IRC16:41
*** ramineni has joined #openstack-ironic16:42
*** romcheg has quit IRC16:45
*** ramineni has quit IRC16:48
*** martyntaylor has joined #openstack-ironic16:51
NobodyCamno distributions at all found for lxml >=2.3 hummmmm.....16:52
Shrewsjroll: the more i look into this concept of "hypervisor uptime" and how it relates to ironic makes me want to kick puppies16:53
jrollShrews: :)16:56
*** lucasagomes is now known as lucas-afk16:58
*** harlowja_away is now known as harlowja17:01
devanandamorning, all17:03
*** rloo has quit IRC17:11
jrollmorning devananda17:13
*** romcheg has joined #openstack-ironic17:18
*** lazy_prince is now known as killer_prince17:19
NobodyCamgood mornign devananda17:22
ShrewsFYI, I plan to ask for advice re: this bug during today's meeting (https://bugs.launchpad.net/ironic/+bug/1337460), if people want to read it ahead of time.  :)17:22
*** matty_dubs|lunch is now known as matty_dubs17:26
adam_ganyone know whats up with my tox py27 environment? http://paste.ubuntu.com/7761142/17:26
*** eghobo has joined #openstack-ironic17:35
devanandaShrews: that's fun. Im curious what the nova team will think17:37
devanandaShrews: it's returning the uptime for each compute host, right?17:38
jrollis that actually useful info, though?17:39
jrollfor any of the hypervisor-* commands, not just -uptime17:40
Shrewsdevananda: best as i can tell (from this http://docs.openstack.org/api/openstack-compute/2/content/ext-os-hypervisors.html), it's supposed to. But not even the libvirt driver does it properly17:41
devanandajroll: stats and list are, though they don' tmean the normal things17:41
matty_dubsIt feels like the 'right' thing is to report Ironic's uptime, but I'm not sure what one might do with that information.17:41
devanandaor i'm thinking of something else17:42
JayFif you define it that way, you rapidly get into 'what does uptime mean' :)17:42
devanandayea17:42
devanandait seems like returning the instance uptime is the wrong thing here17:42
jrolldevananda: welllll, those display node info, not n-cpu info17:42
devanandajroll: right17:43
devanandajroll: which can be useful :)17:43
jrollseems like hypervisor-uptime would be equivalent to 'node uptime'17:43
devanandabut that's really equivlane to instance uptime17:43
jrolltoday.17:43
jroll:)17:44
devanandaah, heh17:44
jrollidk. maybe it should be 'time since provisioned' for now... think about 'ramdisk uptime' later17:45
devanandaso this is a good example where ironic doesn't really fit in nova's API17:45
devanandarather than shoehorn it in very awkardly17:45
devanandaand confuse users17:45
jrollright17:45
jrolljust don't implement17:45
jroll(imo)17:45
*** hemna has quit IRC17:45
devanandaI think we should punt to Nova17:45
ShrewsI mean, we could just return an empty string17:45
devanandais there a way to indicate that "this is intentionallynot implemented" ?17:45
*** hemna has joined #openstack-ironic17:45
Shrewsdevananda: raise NotImplementedException17:45
Shrews:)17:46
devanandaor, alternately, we have these only return the info for the n=cpu host itself17:46
devanandawhich is less useful from *our* POV but probably the correct thing from Nova's POV17:46
*** pelix has quit IRC17:46
devanandaShrews: right. and then tempest complains :)17:46
Shrewswell, a <space> then  :-P17:46
devanandaheh17:47
Shrewsbut like i said, not even libvirt does the "right thing"17:47
Shrewsseems like an api someone once thought might be useful, but turns out it wasn't, so it's forgotten about17:48
devanandaexcept it's not forgotten about by tempest17:48
rameshg87hello devananda17:49
devanandarameshg87: hi17:49
Shrewsdprince doesn't think any driver does it right: https://github.com/openstack/nova/blob/master/nova/virt/libvirt/driver.py#L4834-L483517:50
*** rloo has joined #openstack-ironic17:50
rameshg87devananda: a question regarding ilo power driver spec https://review.openstack.org/#/c/97455/17:51
*** rloo has quit IRC17:51
rameshg87devananda: do we need approval from one more core reviewer for this spec ?17:51
*** rloo has joined #openstack-ironic17:52
devanandaShrews: hm, is "host" there the "hypervisor_hostname" ?17:54
NobodyCamI have a spec question: RE: spec dependenceies. as it pertains to my last comment on https://review.openstack.org/#/c/10095117:54
devanandaShrews: if so, then most hypervisors will be ignoring it at that layer, since only baremetal/ironic/vmware support >1 hypervisor_hostname per n-cpu17:54
Shrewsdevananda: i believe so, from the compute_nodes table17:55
devanandaShrews: right. then that's why it's ignored there.17:55
*** pelix has joined #openstack-ironic17:56
Shrewsdevananda: vmware does... return 'Please refer to %s for the uptime' % CONF.vmware.host_ip17:57
Shrewsso they defer the work.. hehe.. nice17:57
*** rloo has quit IRC17:58
*** rloo has joined #openstack-ironic17:58
*** rloo has quit IRC18:00
devanandarussell_h: ping18:00
jrolldevananda: russell_h is out for the next two weeks18:01
jrollgetting married / honeymoon18:01
devanandajroll: ahh. seems to be taht time of year :)18:01
jrollheh18:02
jrollanything I can help with?18:02
devanandajroll: specs-core team has 2/5 members out for two weeks ...18:02
jrollgah18:02
*** rloo_ has joined #openstack-ironic18:03
*** rloo_ has left #openstack-ironic18:03
*** rloo has joined #openstack-ironic18:05
NobodyCamdevananda: woud you have a free minute to take a look at my last comment on https://review.openstack.org/#/c/100951 ?18:06
rameshg87devananda: will we need to wait for specs-core reviewers to be back ?18:07
*** jgrimm has quit IRC18:08
jrollrameshg87: there are still 3 spec cores around18:08
jroll2 is enough for approval18:08
rameshg87jroll: one basic questions, is the core reviewers list available somewhere ? :-)18:09
jrollrameshg87: yes, looking18:10
*** eghobo has quit IRC18:11
JayFIt's a gerrit group18:13
JayFbut I always have a ton of trouble 'discovering' gerrit groups, or finding the list, even knowing the name :x18:13
jrollyeah :/18:14
rameshg87devananda: please have a look at the ilo deploy spec https://review.openstack.org/#/c/97744/18:14
rameshg87devananda: i have revised it after your comments.18:15
*** chuckC has quit IRC18:15
jrollrameshg87: I can't find the list, sorry :/18:17
rameshg87jroll: np :-)18:18
*** rameshg87 is now known as rameshg87_afk18:22
*** eghobo has joined #openstack-ironic18:22
*** eghobo has quit IRC18:23
*** eghobo has joined #openstack-ironic18:24
NobodyCamrameshg87_afk: JayF: jroll: https://review.openstack.org/#/admin/groups/352,members18:24
rloorameshg87_afk, jroll : https://review.openstack.org/#/admin/groups/352,members18:24
NobodyCamlol18:25
rlooNobodyCam: it was a draw ;)18:25
NobodyCamhehehe18:25
NobodyCamgreat minds ....18:25
NobodyCam:-p18:25
jrollah ha18:25
rloothat too :D18:25
jrollthanks NobodyCam18:25
*** eghobo has quit IRC18:28
NobodyCambb.. b4 meeting18:31
*** dwalleck has quit IRC18:31
devanandare gerrit groups, the full list is browsable here18:31
devanandahttps://review.openstack.org/#/admin/groups/18:31
devanandajust type in "ironic-specs" in the filter and you'l find it18:32
*** foexle has joined #openstack-ironic18:32
rloojroll, NobodyCam, rameshg87_afk, devananda: I just added a link to that in our wiki: https://wiki.openstack.org/wiki/Ironic/Specs_Process#Ironic_Specs_Process18:32
jrollrloo: thanks18:32
JayFnice, thanks devananda  (and thanks to NobodyCam and rloo for the direct link)18:36
matty_dubsHow are core reviewers for the specs chosen?18:38
matty_dubs(Not that I have any issue with the current list; I'm just curious.)18:38
NobodyCammatty_dubs: they volunteer18:39
devanandamatty_dubs: i throw spaghetti at them18:39
devanandamatty_dubs: if it sticks, they're on the team18:39
matty_dubslol18:39
devananda:)18:39
NobodyCamlol18:39
NobodyCamspaghetti is better then a wet cat (which has also been throwen at /me)18:40
devanandaouch18:40
NobodyCamlol18:40
matty_dubsSpaghetti's a lot less likely to claw you. Also, a tasty snack.18:41
*** dwalleck has joined #openstack-ironic18:42
devanandalucas-afk: have time to review /approve some specs tomorrow?18:42
*** tatyana has joined #openstack-ironic18:43
lucas-afkdevananda, hey morning, yes!18:45
lucas-afkdevananda, tomorrow sounds good to me18:46
devanandalucas-afk: cool. lemme get some #'s for ya18:46
lucas-afkack18:46
*** lucas-afk is now known as lucasagomes18:46
devanandalucasagomes: https://review.openstack.org/102914 https://review.openstack.org/97455 https://review.openstack.org/9890418:49
devanandalucasagomes: those three look ready to me18:49
lucasagomesdevananda, ack, I will take a look tomorrow morning18:49
devanandathanks much!18:49
lucasagomesyvw18:50
NobodyCam10 minutes til meeting, last chance for coffee, soda, and other rest stop activities18:52
*** takadayuiko has joined #openstack-ironic18:53
takadayuikoHi, ironic :)18:53
NobodyCamgood morning takadayuiko18:53
takadayuikoNobodyCam, good morning!18:54
sirushtidevananda, Hi, could you please take a look at https://review.openstack.org/#/c/97150/? I've responded to your comments18:54
openstackgerritAdam Gandelman proposed a change to openstack/ironic: Specify rootfstype=ramfs deploy kernel parameter  https://review.openstack.org/10472818:54
NobodyCamlast chance to update the agenda18:55
*** eghobo has joined #openstack-ironic18:56
*** eghobo has quit IRC18:57
*** eghobo has joined #openstack-ironic18:58
* lucasagomes looks at the agenda18:58
lucasagomestakadayuiko, morning, wow it's quite late there no?18:58
lucasagomesor quite early...18:59
devanandasirushti: the claim in that paragraph is still false18:59
devanandasirushti: it is possible to deploy windows today using partition image and pxe boot18:59
devanandasirushti: * with some changes to the template18:59
sirushtidevananda, That sounds cool, is there a link for this?18:59
takadayuikolucasgomes, good morning! yes, it's 4:00 AM, super early :O18:59
devanandasirushti: I put a link in my comment on the spec last week19:00
sirushtidevananda, that's just chainloading19:00
devanandasirushti: correct19:00
sirushtinot really pxe-booting, right?19:00
devanandasirushti: let's continue after meeting19:00
sirushtidevananda, Sure19:00
NobodyCammeeting time19:01
*** foexle has quit IRC19:02
*** tatyana has quit IRC19:04
*** tatyana has joined #openstack-ironic19:09
*** jgrimm has joined #openstack-ironic19:09
*** martyntaylor has quit IRC19:21
*** mkerrin1 has joined #openstack-ironic19:28
*** mkerrin has quit IRC19:29
*** Nisha has quit IRC19:37
*** pelix has quit IRC19:38
*** chuckC has joined #openstack-ironic19:39
*** sseago has quit IRC19:49
*** Penick has joined #openstack-ironic19:49
jrollwanyen... isn't here.20:01
NobodyCamgreat meeting all20:01
BadCub_good meeting :-)20:01
NobodyCamlooks like he logged off20:01
jrollwell, he wasn't here before thay20:01
jrollthat20:01
*** wanyen has joined #openstack-ironic20:01
devanandaon the generic spec -> specific spec dependnecy20:01
devanandaI asked for things like hardware discuvery to be written up in a generic way specifically so that they are addressed in a way that works for more than one vendor20:02
devanandai do not want to approve the spec (or the code) for one vendor before the generic spec20:02
NobodyCam++20:02
devanandabecause it is likely to create technical debt -- we'll have to refactor it to make it fit the generic model we eventually adopt (or remove it)20:03
devanandaOTOH, that is the main purpose for the /vendor_passthru/ API20:03
devanandaso I have no objection to a vendor implementing it *there*20:03
devanandawhen ever they want to20:03
*** rloo has quit IRC20:04
jroll+120:04
*** rloo has joined #openstack-ironic20:04
NobodyCamdevananda: my concern is that we will approve a spec like this and then the dependent spce will change20:04
wanyenSo, I have concern that no other discovery code can be approved until the generic hw discovery bit code has been approved.  because the dependency is the inventory field should it just be dependencnt on the lading of tnat data structure?20:05
NobodyCamfor things like changing the datamodel20:05
NobodyCamie adding inventor field20:05
devanandaNobodyCam: that's one reason I dont think they should be approved until the common spec is approved20:05
*** tatyana has quit IRC20:05
jrollwanyen: the common spec also defines the API interfaces20:05
devanandawanyen: did you see my comments here just after the meeting?20:05
jrollfor discovery20:05
jrollwhat's the point of implementing "discovery on node-create" if the discovery code doesn't exist?20:06
devanandawanyen: http://paste.openstack.org/show/85613/20:06
lucasagomesdevananda, wanyen just approved the iLO power driver spec20:07
*** takadayuiko has quit IRC20:07
lucasagomesthanks for that, small and direct spec20:07
wanyendeva> yes.  I saw taht .20:07
devanandajroll: why is https://review.openstack.org/#/c/95551 a dependency for ipa driver?20:08
NobodyCamgah ... brb20:08
lucasagomesI will take a look at the other specs tomorrow, it's late here20:08
jrolldevananda: it shouldn't be, any more. did you get that from ipa-todos or did I forget to rebase?20:08
devanandalucasagomes: indeed it is :) have a good evening!20:08
lucasagomes(and I will still tired from the trip to the ceilometer midcycle)20:08
wanyen>deva> it seems to that the generic hw discovery bit focus on in-band discovery which can be very different from oob discovery20:08
devanandajroll: ipa-todos20:09
lucasagomesdevananda, thanks you too20:09
lucasagomeshave a great night everybody20:09
jrolldevananda: right, no longer a dep20:09
devanandajroll: great20:09
*** lucasagomes is now known as lucas-dinner20:09
devanandajroll: i'll stop reading it until ya'll update it :)20:09
jrollI just removed that line20:09
jrolldevananda: that section is accurate now20:09
jrolldevananda:20:09
jrolloops20:09
devanandaoh, reminder for folks that will be at the midcyucle (i should have pointed this out in the meeting)20:10
devanandaI'm tracking agenda topics here: https://etherpad.openstack.org/p/juno-ironic-sprint20:10
jrolldevananda: you can look at the reviews in that section... then the bottom needs to be updated for the branches coming off the main agent driver spec20:10
jrollor I can just ping you when I update :P20:10
openstackgerritA change was merged to openstack/ironic-specs: iLO Power Driver for Ironic  https://review.openstack.org/9745520:10
devanandajroll: i doubt i'll get much time for that today, but hopefully this week I can give them a read agaon20:11
devanandaagain20:11
jrolldevananda: sounds good :) I have specs to update too20:11
wanyenlucasagomes> ty very much for approving the ilo power spec.20:11
devanandajroll: i'm eagerly awaiting the day when you point me at a walkthrough to get devstack up and running with it :)20:11
jrolldevananda: soon (this week?). my plan is to point at our devstack docs :)20:12
* jroll brb20:13
sirushtidevananda, Hi, could we continue our discussion?20:16
*** dwalleck has quit IRC20:27
*** dwalleck_ has joined #openstack-ironic20:30
devanandasirushti: hi! i have a few minutes20:31
devanandasirushti: so it is pxe booting in that the server issues a DHCP BOOT request, and the response is controllable by Ironic20:32
sirushtidevananda, that should do :-) so, as I was saying one can't pxe-boot windows using partition images+chainloading20:32
sirushtidevananda, ok20:32
sirushtidevananda, so anything involving pxelinux.0 is assumed to be pxe-boot?20:33
sirushtiI thought pxe-boot is pxe acting as the bootloader and not something on the disk to be called pxe boot. Am I missing20:33
sirushtibecause chain-load actually gives its responsibility to the node and the node boots using the local bootloader20:34
*** Hefeweizen has joined #openstack-ironic20:35
sirushti*s/missing/missing something/20:36
*** foexle has joined #openstack-ironic20:40
openstackgerritSirushti Murugesan proposed a change to openstack/ironic-specs: Whole Disk Image Support for PXE Deploy Driver  https://review.openstack.org/9715020:41
devanandasirushti: as i understand it, this will work even if the local HDD has no boot flag set20:42
JayFI saw that assumption in the spec. It's not awesome to assume that HDD will boot if PXE times out (because that's not always true)20:43
JayFbut I don't think that's a good reason to disapprove the spec20:43
sirushtidevananda, this = the link you provided?20:43
JayFOpportunistically supporting boxes that /can/ have their bios set to fallback but have no IPMI is a great idea, and I actually loved seeing it in the spec20:43
JayFsounds like a good way to support other things (like the agent) on boxes without explicit IPMI support (but maybe having power_driver hooked up to a ups, for instance)20:44
sirushtiJayF, thats only until we get the changes into the ssh driver to change the boot order though20:44
JayFsirushti: I think it's a good fallback scenario to be aware of in any situation. Earlier in the life of onmetal, we had thought of doing just that: boxes that always pxe'd, and had a power strip hooked up with a power driver, and just letting the fallback work20:45
sirushtiJayF, ooh, nice :-)20:46
devanandasirushti: ssh power driver's ability to change the boot order on VMs will enable some better testing in virtual environments20:47
devanandasirushti: but does not cover other power drivers (like PDU) which will not have such an ability20:47
JayFI look forward to finding even funkier hardware to boot with Ironic20:48
devanandaso, to rephrase my objection to rev.1120:48
devanandathis statement is illogical: "Since Windows cannot be PXE-Booted via a certain25 kernel/ramdisk, there's no choice but to deploy whole disk Windows images.20:48
devananda"20:48
JayFI honestly thought of doing an ironic-provisioned cluster of raspberry pis once, but they don't have any pxe support whatsoever (without putting an OS on that chains into pxe)20:48
sirushtidevananda, hmm, perhaps, I should just remove that line altogether :-p I only wanted to emphasize that windows would be a good use case20:49
devanandaJayF: stick a uboot image on the pi's sdcard that chains into ipxe ;)20:49
JayFdevananda: yeah but then I can't have IPA write a full image to it :P20:50
JayFdevananda: what I really want to do is buy an Orange Box from Ubuntu, make it work with Ironic, and show them how well their box works with a /real/ openstack provisioning method :)20:50
devanandaJayF: right. well....20:50
devanandaJayF: the NUCs are available on amazon20:51
JayFwell yeah, and it'd be sweet, but for now I'm spinning up bigger boxes :P20:51
devananda:)20:51
devanandasirushti: i agree that windows is a good use case :)20:51
devanandasirushti: just not that it is impossible to use windows without the changes that spec describes20:51
devanandawhich is how I interpret that sentence20:52
Shrewsdevananda: about the tempest test for uptime... looks like we're just going to have to skip it if baremetal driver is enabled. We can't make it consider 501 a valid response w/o affecting the whole test suite.20:52
devananda:(20:52
Shrewswhich makes me "grrrrr"20:52
sirushtidevananda, I was actually looking out for resources that would let me pxe-boot w/ windows and still have preserve_ephemeral, not much luck really20:52
sirushtibut there probably is a way, so I suppose it is not right to say that it is impossible :-)20:53
devanandasirushti: as it is written today, preserve ephemeral won't work at all with whole-disk image20:54
sirushtidevananda, right, that was just off topic trying to find a way getting windows to be partition-booted20:55
devanandalunch time, then i have a bunch of paperwork to do .... probably lurking the rest of the day, but not really here20:55
NobodyCamack enjoy lunch devananda20:55
*** dwalleck_ has quit IRC20:56
*** dwalleck has joined #openstack-ironic20:57
openstackgerritSirushti Murugesan proposed a change to openstack/ironic-specs: Whole Disk Image Support for PXE Deploy Driver  https://review.openstack.org/9715020:58
*** dwalleck_ has joined #openstack-ironic20:58
sirushtidevananda, JayF thanks for the input :-) Please do have a look https://review.openstack.org/9715020:58
*** dwalleck has quit IRC21:01
*** harlowja is now known as harlowja_away21:02
*** GheRiver1 is now known as GheRivero21:05
*** jbjohnso_ has quit IRC21:06
openstackgerritEllen Hui proposed a change to openstack/ironic: Add more INFO logging to ironic/conductor  https://review.openstack.org/9974421:11
jrollhow do I specify an option as None in ironic.conf? :/21:16
jrolloption=, option=None, option=<None> do not work21:16
jrollall produce strings21:16
*** jcoufal has quit IRC21:18
*** dwalleck_ has quit IRC21:19
Shrewsjroll: is there a default?21:23
*** dwalleck has joined #openstack-ironic21:23
jrollShrews: nope21:23
Shrewsif not, then don't specify it... or is that broken?21:23
jrolllooking at oslo.config21:23
jrolloh21:24
jrollcould work21:24
jrollhmm /me tries21:24
* NobodyCam thinks devananda's idea about food was a good thing, and goes off to look for soething to eat.. bbiaf21:30
*** foexle has quit IRC21:31
jrollShrews: that did it, thanks21:32
* jroll fights keystone now21:32
*** matty_dubs is now known as matty_dubs|gone21:33
Shrewsjroll: i iz a jeenys21:33
jroll:P21:33
Shrewsjeenyus... geez, i can't spell21:34
Shrews:)21:34
rlooShrews, do you have a few min to look at https://review.openstack.org/#/c/100957/? I want to make sure you're OK with it.21:49
*** dwalleck has quit IRC21:50
*** dwalleck has joined #openstack-ironic21:52
NobodyCamjroll: was looking at the tests for https://review.openstack.org/#/c/10036421:54
*** max_lobur has joined #openstack-ironic21:55
jrollwow, so many -1's21:55
NobodyCamseveral of the test have no assert at all?21:55
jrollNobodyCam: what's up?21:55
jrolluh21:56
NobodyCamhttps://review.openstack.org/#/c/100364/8/ironic/tests/drivers/test_ipmitool.py line 76821:56
jrollNobodyCam: which?21:56
jrolloh21:56
NobodyCam74721:56
jrolloops :)21:56
NobodyCamhehehe note I did not point that on the review21:57
NobodyCam:-p21:57
jrolloh wait21:57
jrollhrmmm21:57
*** mkerrin1 has quit IRC21:57
*** mkerrin has joined #openstack-ironic21:57
jrollNobodyCam: thanks, will fix21:58
NobodyCamI think thats actually okay. but I always try and assert something21:58
jrollright21:58
jrollactually, ok, I remember21:58
jrollNobodyCam: that just calls vendor.validate()21:59
jrollwhich just validates args or whatever21:59
jrollso, it not raising an exception is the only way you can tell it failed21:59
jrolls/failed/passed/21:59
jrollbrb22:00
*** harlowja_away is now known as harlowja22:04
*** romcheg has quit IRC22:10
*** dkehnx1 has joined #openstack-ironic22:16
NobodyCamjroll: when your back please have a look at https://review.openstack.org/#/c/103595 with reguards to https://review.openstack.org/#/c/10073422:26
jrollgah22:27
jrollit would be nice if that depended on my patch or vice versa22:28
NobodyCam:) we can poke lucas tomorrow22:29
NobodyCamhehehe22:29
jrollyeah22:29
jrollhe even mentioned it in a comment22:29
NobodyCamya22:29
jrollso much to do22:29
NobodyCamthat series of patches is quite long :-p22:30
openstackgerritA change was merged to openstack/ironic: Imported Translations from Transifex  https://review.openstack.org/10329122:30
jrollonly the first is getting any love22:30
NobodyCam:)22:34
NobodyCamI'm looking :)22:35
NobodyCamhehehe ,,, but gota run to the store... brb22:35
*** Penick has quit IRC22:41
*** steveh1 has joined #openstack-ironic22:45
Shrewsrloo: re: 100957, not the way i'd do it (i consider it a design anti-pattern), but not here to be preachy about such things. fine to +A if you guys want.  :)22:49
*** dwalleck has quit IRC22:51
rloothx Shrews.23:14
* NobodyCam wanders afk... /me need to touch up some paint on the house23:28
*** lucas-dinner has quit IRC23:41

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