Thursday, 2013-12-12

« Wednesday, 2013-12-11 Index Friday, 2013-12-13 »
*** lexx has quit IRC00:33
NobodyCamgah that was a lot longer then I had expected00:38
*** hemna_ is now known as hemnafk00:47
*** anniec has quit IRC01:01
*** jcooley_ has quit IRC01:12
NobodyCamok updated the dib walkthru to set the compute driver.01:14
*** anniec has joined #openstack-ironic01:15
*** jcooley_ has joined #openstack-ironic01:33
*** nosnos has joined #openstack-ironic01:35
devananda:)01:35
* devananda sends emails to ML, then signs off for the evening01:36
NobodyCamnight devananda01:44
*** kui has joined #openstack-ironic01:49
Haomenggood morning/evening Ironic:)01:50
NobodyCamgood mornig Haomeng01:50
HaomengNobodyCam: :)01:50
NobodyCam:)01:50
HaomengNobodyCam: I guess, now is your afternoon?01:51
NobodyCamnight time here just about 6 pm01:51
HaomengNobodyCam: :)01:51
*** rongze has joined #openstack-ironic01:52
HaomengNobodyCam: it is time for your Dinner:)01:52
HaomengNobodyCam: enjoy your Dinner:)01:52
NobodyCamjust about .. its in the oven01:52
HaomengNobodyCam: )01:53
NobodyCam:)01:53
HaomengNobodyCam: Beef and Bread?:)01:54
HaomengNobodyCam: Delicious:)01:54
NobodyCampork chops :)01:54
HaomengNobodyCam: :)01:54
openstackgerritA change was merged to openstack/ironic: Add JsonPatchType class  https://review.openstack.org/6019402:14
*** arata has joined #openstack-ironic02:19
openstackgerritA change was merged to openstack/ironic: Rework patch validation on Ports  https://review.openstack.org/5698402:22
openstackgerritA change was merged to openstack/ironic: Clean up duplicated change-building code in objects  https://review.openstack.org/6114302:22
*** anniec has quit IRC02:22
openstackgerritA change was merged to openstack/ironic: Set the provision_state to DEPLOYFAIL  https://review.openstack.org/6002402:24
*** jcooley_ has quit IRC02:24
*** rongze has quit IRC02:28
*** rongze has joined #openstack-ironic02:29
*** vkozhukalov has joined #openstack-ironic02:38
*** viktors has quit IRC03:12
*** viktors has joined #openstack-ironic03:13
*** rloo has quit IRC03:18
*** Haomeng has quit IRC03:20
*** Haomeng has joined #openstack-ironic03:20
*** jcooley_ has joined #openstack-ironic03:20
*** jcooley_ has quit IRC03:26
*** rongze_ has joined #openstack-ironic03:33
*** rongze has quit IRC03:36
*** anniec has joined #openstack-ironic03:37
*** anniec has quit IRC03:42
*** anniec has joined #openstack-ironic04:37
*** anniec_ has joined #openstack-ironic04:41
*** anniec has quit IRC04:41
*** anniec_ is now known as anniec04:41
*** anniec has quit IRC04:50
*** rongze_ has quit IRC04:52
*** rongze has joined #openstack-ironic05:22
*** rongze has quit IRC05:27
*** prekarat has joined #openstack-ironic05:28
*** vetalll has joined #openstack-ironic05:45
openstackgerritJenkins proposed a change to openstack/ironic: Imported Translations from Transifex  https://review.openstack.org/5962706:00
*** jcooley_ has joined #openstack-ironic06:06
*** rongze has joined #openstack-ironic06:20
*** jcooley_ has quit IRC06:30
*** jcooley_ has joined #openstack-ironic06:30
openstackgerritHaomeng,Wang proposed a change to openstack/ironic: Command call should log the stdout and stderr  https://review.openstack.org/6166406:30
*** jcooley_ has quit IRC06:31
*** jcooley_ has joined #openstack-ironic06:31
*** lifeless has quit IRC06:48
*** vkozhukalov has quit IRC06:51
*** vetalll has quit IRC06:58
*** lifeless has joined #openstack-ironic07:15
*** jcooley_ has quit IRC07:18
*** jcooley_ has joined #openstack-ironic07:22
*** urulama has joined #openstack-ironic07:42
*** jcooley_ has quit IRC07:46
*** jcooley_ has joined #openstack-ironic07:50
*** ifarkas has joined #openstack-ironic07:56
*** jcooley_ has quit IRC07:58
*** rongze has quit IRC08:02
*** rongze has joined #openstack-ironic08:06
*** lexx has joined #openstack-ironic08:08
*** rongze has quit IRC08:11
*** lexx has quit IRC08:24
*** vkozhukalov has joined #openstack-ironic08:28
*** prekarat has quit IRC08:35
*** ifarkas has quit IRC08:46
*** r-mibu has quit IRC08:48
*** ifarkas has joined #openstack-ironic08:50
openstackgerritHaomeng,Wang proposed a change to openstack/ironic: Command call should log the stdout and stderr  https://review.openstack.org/6166408:55
*** jistr has joined #openstack-ironic09:03
*** derekh has joined #openstack-ironic09:05
*** ekarlso has quit IRC09:17
*** ekarlso has joined #openstack-ironic09:17
*** lucasagomes has joined #openstack-ironic09:28
openstackgerritA change was merged to openstack/ironic: Change default API auth to keystone-based  https://review.openstack.org/6156809:42
openstackgerritLucas Alvares Gomes proposed a change to openstack/ironic: Handle DBDuplicateEntry on Ports with same address  https://review.openstack.org/5995509:46
*** kui has quit IRC09:47
*** max_lobur_afk is now known as max_lobur10:00
*** martyntaylor has joined #openstack-ironic10:09
*** romcheg has joined #openstack-ironic10:12
*** nosnos_ has joined #openstack-ironic10:15
*** nosnos_ has quit IRC10:16
*** nosnos_ has joined #openstack-ironic10:16
*** ndipanov has joined #openstack-ironic10:17
*** nosnos has quit IRC10:17
*** ndipanov has quit IRC10:18
*** ndipanov has joined #openstack-ironic10:19
*** nosnos_ has quit IRC10:20
*** ndipanov has quit IRC10:20
openstackgerritHaomeng,Wang proposed a change to openstack/ironic: Command call should log the stdout and stderr  https://review.openstack.org/6166410:38
*** rongze has joined #openstack-ironic10:41
openstackgerritLucas Alvares Gomes proposed a change to openstack/ironic: API: Validate and normalize address  https://review.openstack.org/6089010:43
*** rongze has quit IRC10:46
*** ndipanov has joined #openstack-ironic11:26
*** anniec has joined #openstack-ironic11:33
*** anniec_ has joined #openstack-ironic11:37
*** anniec has quit IRC11:37
*** anniec_ is now known as anniec11:37
openstackgerritLucas Alvares Gomes proposed a change to openstack/ironic: Convert trycmd() to oslo's processutils  https://review.openstack.org/6170911:46
*** kui has joined #openstack-ironic11:47
Haomenggood morning/evening Ironic:)11:54
*** homegrown has left #openstack-ironic11:57
*** kushi_ has joined #openstack-ironic12:12
*** arata has left #openstack-ironic12:21
kushi_I encounter an error while setup devtest env: https://bugs.launchpad.net/tripleo/+bug/126029212:23
kushi_I paste it on #tripleo too, any suggestion ?12:24
Haomengkushi_: welcome12:25
kushi_Haomeng: Good evening !12:25
Haomengkushi_: for such "error: cannot create file '/var/lib/libvirt/network/brbm.xml.new': Permission denied" persission issue, I suggest you work around to set it first12:25
Haomengkushi_: looks like current user has no permision to create file12:26
Haomengkushi_: what user you used? stack or root?12:26
Haomengkushi_: check /var/lib/libvirt/network folder permisions12:26
kushi_I use common user to run devtest.sh,  the setup-network will raise this error.12:26
*** ndipanov has quit IRC12:27
kushi_change the dir permission is a workaround, but this issue should be fixed.12:27
Haomengkushi_: yes12:30
Haomengkushi_: should be devtest.sh issue I think12:30
Haomengkushi_: you can raise defect to track such issue12:31
Haomengkushi_: in tripleo launchpad12:31
kushi_yes, I have filed one:   https://bugs.launchpad.net/tripleo/+bug/126029212:32
Haomengkushi_: cool12:35
Haomengkushi_: and you can try to fix I suggest:)12:35
kushi_Haomeng: It is a pity that Ubuntu 12.04 is not supported well.  I will change to Ubuntu 13.0412:36
Haomengkushi_: ok12:37
Haomengkushi_: :)12:37
*** rongze has joined #openstack-ironic12:37
*** rongze has quit IRC12:41
openstackgerritYuriy Zveryanskyy proposed a change to openstack/ironic: Check connectivity in SSH driver 'validate' method  https://review.openstack.org/6087612:42
*** prekarat has joined #openstack-ironic12:56
*** urulama has quit IRC13:07
*** ndipanov has joined #openstack-ironic13:09
*** rongze has joined #openstack-ironic13:12
*** kushi_ has quit IRC13:12
*** kui has quit IRC13:17
*** arata has joined #openstack-ironic13:24
*** arata has left #openstack-ironic13:25
*** ndipanov has quit IRC13:30
*** ndipanov has joined #openstack-ironic13:30
*** linggao has joined #openstack-ironic13:31
*** jdob has joined #openstack-ironic13:40
*** prekarat has quit IRC13:41
*** prekarat has joined #openstack-ironic13:41
*** romcheg1 has joined #openstack-ironic13:56
*** rloo has joined #openstack-ironic13:56
*** romcheg has quit IRC14:00
*** rongze_ has joined #openstack-ironic14:01
*** blamar has quit IRC14:01
*** rongze has quit IRC14:04
openstackgerritLucas Alvares Gomes proposed a change to openstack/ironic: API: Add sample() method on Port and PortCollection  https://review.openstack.org/6175214:18
rloohi lucasagomes.14:19
lucasagomesrloo, hey ya :) morning14:19
rloo:-)14:19
lucasagomesrloo, how r u?14:19
rloolucasagomes, things are good here!14:20
*** rongze has joined #openstack-ironic14:20
rloolucasagomes, i just saw the bug you filed. which reminded me. i was thinking i'd like to understand the api part a bit more.14:20
rloolucasagomes. client -> library parts.14:21
rloolucasagomes - off the top of your head, is there anything you'd like me to look into? bug or whatever?14:21
lucasagomesrloo, hehe I opened that and made that ^ review as a sample, it's a good work for someone who's just starting in Ironic to get some patches in14:23
*** rongze_ has quit IRC14:24
rloolucasagomes: just looked at it and it is assigned to matt wagner.14:26
lucasagomesrloo, yea he's just starting in ironic so I assigned it to him... lemme think if there's something else14:26
lucasagomesrloo, I think there's a bug on the limit variable of the pagination14:26
lucasagomesrloo, https://bugs.launchpad.net/ironic/+bug/125080414:27
lucasagomessee if something you'd feel like fixing14:27
rlooah, ok, i'll take a look at it. thx.14:27
rloolucasagomes. sorry, another question. the importance is 'undecided'. does it matter or does someone decide/change it?14:29
lucasagomesrloo, you can change that. actually there's a guide to help decided what's the importance14:30
lucasagomeslemme try to find the link14:30
lucasagomesrloo, https://wiki.openstack.org/wiki/BugTriage14:31
rloolucasagomes. seems like there is a guide to almost everything :-) is there a top-level page where you search for such links? some openstack page?14:31
lucasagomesrloo, not that I know, if you find it please lemme know :D... These links I found at random places or someone pointed me to them and I just bookmarked it14:32
rloothx, lucasagomes. I see that there's a search box at wiki.openstack.org, so I might try to search from there the next time. So I don't bother you all too much :-)14:33
*** jdob has quit IRC14:34
*** jdob has joined #openstack-ironic14:34
lucasagomesrloo, it's grand :) yea the search box actually works alright14:35
rloolucasagomes: now I'm tempted to go through all the new bugs and try to confirm them, but I won't :-)14:36
*** arata has joined #openstack-ironic14:36
lucasagomesrloo, :P heh, that's actually needed14:36
* lucasagomes should look at the bugs more often14:36
rloolucasagomes: I'll keep it in mind for the future ;)14:37
lucasagomes:) thanks14:38
rloolucasagomes: there are only 4 with status 'new', and I'm going to change one of them soon, so that leaves 3 :-)14:39
lucasagomesrloo, hah! magic :D14:39
*** rongze_ has joined #openstack-ironic14:43
*** rongze has quit IRC14:44
*** blamar has joined #openstack-ironic14:45
*** prekarat has quit IRC14:49
*** rloo has quit IRC14:57
*** martyntaylor1 has joined #openstack-ironic14:57
*** rloo has joined #openstack-ironic14:57
*** jistr has quit IRC14:58
*** martyntaylor has quit IRC14:59
*** arata has left #openstack-ironic14:59
*** jistr has joined #openstack-ironic15:02
*** jcooley_ has joined #openstack-ironic15:05
*** matty_dubs|gone is now known as matty_dubs15:08
*** mdurnosvistov has joined #openstack-ironic15:14
NobodyCamgood morning iRonic15:27
max_loburmorning NobodyCam , Ironic :)15:31
NobodyCammorning max_lobur :)15:31
openstackgerritAlexandra Fomina proposed a change to openstack/python-ironicclient: Reuse Oslo exceptions functionality  https://review.openstack.org/6177115:31
NobodyCamlucasagomes: looks like 59955 was approved then you uploaded another patch15:33
lucasagomesNobodyCam, I had to rebase it15:34
NobodyCamahh15:34
NobodyCamwill +a it again15:34
lucasagomesNobodyCam, cheers :)15:34
lucasagomesNobodyCam, and good morning, how r u?15:34
NobodyCam:) doing good so far.. not as cold today and I already have my coffee :)15:35
lucasagomesNobodyCam, nice! yea man, here as well it's pretty warm today (judging by the winter standard)15:36
lucasagomesit's like 15C, very warm15:36
NobodyCam:) it was down to like 32 F here last night... Thats too cold for /me15:38
NobodyCamlucasagomes: question on: https://review.openstack.org/#/c/61664/315:38
lucasagomesyup15:38
NobodyCamyou +2'd it15:38
lucasagomesyea15:39
openstackgerritMikhail Durnosvistov proposed a change to openstack/ironic: Rid of specific `dict` methods in object model  https://review.openstack.org/6177615:39
openstackgerritMikhail Durnosvistov proposed a change to openstack/ironic: Rid of specific `dict` methods in object model  https://review.openstack.org/6002515:39
NobodyCamI'm not sure _("blah : %s" st[in|out|err] is correct15:39
NobodyCamdo we translate the entire error message15:40
NobodyCamor just the "blah" bits15:40
lucasagomesNobodyCam, just the blah bits15:41
lucasagomese.g15:41
lucasagomesLOG.debug(_('Command line is "%s"'), ' '.join(cmd))15:41
lucasagomesonly 'Command line is "%s"' will be translated15:41
lucasagomesthe commands/outputs are dynamic15:42
NobodyCamok15:43
lucasagomesNobodyCam, btw quick q as u'r using/tesing the ironic nova driver15:47
lucasagomesNobodyCam, when registering nodes you do it directly on the ironic api right? nova won't mediated that right?15:48
NobodyCamlucasagomes: correct15:50
lucasagomesNobodyCam, thanks15:50
*** vkozhukalov has quit IRC15:51
openstackgerritA change was merged to openstack/ironic: Handle DBDuplicateEntry on Ports with same address  https://review.openstack.org/5995515:55
NobodyCambrb quick morning walkies15:57
NobodyCammikal: are https://review.openstack.org/#/c/60025/ and https://review.openstack.org/#/c/61776/ both the same16:02
devanandag'morning, all16:04
NobodyCamgood morning devananda :)16:05
devanandarloo, lucasagomes: unfortunately I don't know of any top-level wiki index page either. site search and google search are helpful16:09
devanandarloo: as far as bugs, did you see my email yesterday?16:10
rloodevananda: thx, I'm going to try site search.16:10
devanandaactually, ^ is a general question for everyone ....16:10
rloodevananda: email to ?16:10
devanandaas a project, we hven't used the openstack-dev ML nearly enough16:10
devanandalet's change that :)16:10
rloodevananda: so I guess the answer is no.16:10
devanandahehehe16:10
rloodevananda. oh, not the dreaded openstack-dev ML!16:11
devanandamuahahaha! yes! you must all start reading the dreaded dev list! Hahahahah!16:11
rloodevananda: i'm only 756 emails behind :-(16:11
devanandarloo: filters!16:11
rloodevananda: ok, i'll add a filter for ironic then :-)16:11
NobodyCamlucasagomes: such real looking samples in 61752...16:12
lucasagomesdevananda, I did, I'm taking a look at some bugs today16:12
lucasagomesNobodyCam, yea :D16:13
devanandalucasagomes: awesome, thanks!16:13
linggaoMorning all.16:13
lucasagomesNobodyCam, so I created that review more as a sample for the bug it's partially fixing16:13
rloomorning linggao16:13
linggaomorning rloo.16:14
lucasagomesNobodyCam, there's some people starting in ironic so that would be a good bug to them to tackle16:14
lucasagomesand get some patches in16:14
linggaoNobodyCam and devananda, sorry I could not be able to contribute code because I am pulled to do something else.16:15
linggaoI am not good at multitasking.16:15
NobodyCamlucasagomes: ya. my comment was just that they look so real. any reason to not use 11111-222222-33333 and aa:bb:cc type samples16:16
devanandalinggao: no worries, that is life :)16:16
NobodyCamjust seems odd to real looking id's in a sample16:16
NobodyCam:-p s/to real/to have real/16:17
linggaoOne question on nova baremetal, I am setting up a cloud with nova baremetal, everything is on one node except the baremetal compute node16:17
linggaowhere should the scheduler run? on the compute node or the controller node?16:17
lucasagomesNobodyCam, ah, hmm well I looked at ceilometer as an example16:18
NobodyCam?? linggao is that by your design?16:18
lucasagomesthey use real stuff there16:18
lucasagomesseems so16:18
lucasagomesNobodyCam, https://github.com/openstack/ceilometer/blob/master/ceilometer/api/controllers/v2.py#L587-L60016:18
lucasagomesNobodyCam, https://github.com/openstack/ceilometer/blob/master/ceilometer/api/controllers/v2.py#L855-L86316:18
lucasagomesso I used some real data in ironic as well16:18
NobodyCamhumm16:18
lucasagomesbut maybe it's a bad practice16:19
lucasagomesdon't hesitate in -1 if you think it's not right to do it16:19
NobodyCamdevananda: any thoughts on ^^^16:19
linggaoNobodyCam, I have two nodes, one has the nova conductor, keystone, glance... etc, one has the nova-compute.16:19
NobodyCami am tempted to16:19
devanandalinggao: i dont think it matters on which node those run. they can all be on one node even16:20
NobodyCamlinggao: they can all be on the same node16:20
NobodyCamlike the OoO seed and undercloud are currently nova baremetal16:20
linggaoyes, but I would like to have the compute node separate. can16:21
devanandaNobodyCam: link to the example in questin?16:21
linggaoN16:21
linggaoo16:21
linggaon16:21
linggaoN16:21
linggaoo16:21
linggaos16:21
linggaoo16:21
linggaosorry16:21
linggaoNobodyCam and devananda, in fact I'd like to have 2 compute nodes, 1 for kvm 1 for baremetal. Is it possible?16:22
devanandalucasagomes: usually i discourage using real-looking data in unit tests16:23
devanandalinggao: there has been some work on allowing multiple hypervisors in the same region. the last time I tried (~6mo ago) it was not possible, but that may have cvhanged16:23
linggaodevananda, thanks for the info.16:24
lucasagomesdevananda, yea, that's not unit tests, they are sample data that will be exposed in out API documentation16:24
lucasagomesour*16:24
devanandaahhh16:24
devanandalink pls :)16:24
linggaodevananda, the wiki says that we need to set scheduler_host_manager = nova.scheduler.baremetal_host_manager.BaremetalHostManager on the compute host16:25
devanandalinggao: correct16:25
lucasagomesdevananda, https://review.openstack.org/#/c/61752/16:25
linggaodoes that mean the scheduler has to be on the compute host?16:25
devanandalinggao: this is why you can't run two compute hypervisors -- baremetal needs a different scheduelr than kvm16:25
devanandalinggao: no16:25
devanandalinggao: scheduler can be anywhere. it's just a module specific to how baremetal is scheduled16:25
NobodyCamdevananda: https://review.openstack.org/#/c/61752/16:25
linggaodevananda, if the scheduler is on the controller, should I also set scheduler_host_manager = nova.scheduler.baremetal_host_manager.BaremetalHostManager on the controller?16:26
devanandalinggao: i dont understand that question. that setting only affects the scheduler16:28
devanandaNobodyCam, lucasagomes: for example, http://api.openstack.org/api-ref-compute.html16:29
devanandai have no objection to real-looking data in the API samples. that seems actually quite useful16:29
linggaodevananda, that's what I thought, but the wiki said put that setting on the baremetal compute host. But my scheduler is on the other node.16:30
devanandawhereas in unit tests, i find it good to quickly see what was fake data and where an actual generate_uuid was called16:30
lucasagomesdevananda, makes sense16:30
devanandalinggao: ah. that is poor wording in the wiki16:30
linggaodevananda, got it. thanks.16:31
*** martyntaylor1 has quit IRC16:44
*** martyntaylor has joined #openstack-ironic16:44
*** martyntaylor has quit IRC16:49
*** ekarlso has quit IRC16:50
*** jistr has quit IRC16:50
*** yuriyz has joined #openstack-ironic16:50
*** ekarlso has joined #openstack-ironic16:51
NobodyCamlucasagomes: 60299 looks good I will +a but its going to need a rebase16:53
NobodyCamit's raining here :-p16:56
lucasagomesNobodyCam, :D lemme take a look16:56
lucasagomesNobodyCam, lemme rebase it16:57
lucasagomesNobodyCam, ohh actually, I will have to update it's dependency so rebasing now won't make sense :(16:57
* lucasagomes I'm answering to the reviews on that dependency right now16:58
NobodyCam:)16:58
NobodyCam:)16:58
yuriyzMorning Ironic16:58
NobodyCammorning yuriyz :)17:00
* NobodyCam brb making fresh coffeee17:00
*** anniec has quit IRC17:01
*** rongze_ has quit IRC17:01
lucasagomesmorning yuriyz17:02
yuriyzlucasagomes, about expose node reservation to users and POST a node. I think "readonly" parameter for WSME attributes needed for us. https://review.openstack.org/6179717:04
lucasagomesyuriyz, ah nice! But if it's readonly you shouldn't patch it17:05
*** mdurnosvistov has quit IRC17:05
*** martyntaylor has joined #openstack-ironic17:05
devanandawow, lots of reviews to catch up on ....17:07
yuriyzlucasagomes, patch uses Node __init__, not input json data17:07
*** matty_dubs is now known as matty_dubs|lunch17:08
yuriyzand Node api obj and NodePatchType are different17:10
lucasagomesyuriyz, right... so Node is the document, and it's going to expose 'reservation' as a field in that document17:14
devanandaNobodyCam: when you have time, https://review.openstack.org/58940 will interest you17:14
lucasagomesbut to break the node lock you have to update (patch) that document to set reservation to None17:14
lucasagomesyuriyz, but if the 'reservation' field is readonly, it sounds a bit contradicting to allow users to update it (set it to none)17:15
lucasagomesyuriyz, but I think that patch and the ability to have a readonly attribute great! for e.g to use in the 'links' attribute17:16
NobodyCamdevananda: :) I knew it...17:16
NobodyCamlol17:16
NobodyCam:)17:16
devanandayuriyz: re https://review.openstack.org/#/c/60876/, see my latest comment17:16
*** rloo has quit IRC17:17
NobodyCamdevananda: does that mean we need to add vmware to our ssh too17:17
*** rloo has joined #openstack-ironic17:17
devanandaNobodyCam: yep. I'm going to file a bug to track it17:18
*** max_lobur is now known as max_lobur_afk17:19
NobodyCamI may -1 that patch.. I need to test something17:20
devanandaNobodyCam: also, there are some changes to virtual_power_driver.py that i think you should see. he's removing "" in some of the string matching, for example17:20
*** rloo has quit IRC17:20
devanandayea17:20
NobodyCamya17:20
*** rloo has joined #openstack-ironic17:20
NobodyCamthat will cause sub name matching errors I think17:20
NobodyCamthough he also changes in to ==17:21
devanandayea17:21
yuriyzlucasagomes, this patch for exemple only prohibit changes from body=Node and not from Node __init__ (as in your 'patch' code)17:21
*** rloo has quit IRC17:22
*** rloo has joined #openstack-ironic17:23
devanandayuriyz: added another comment on ssh validate patch.17:23
devanandayuriyz: caching the ssh_obj is a reasonable thing to do BUT is going to add considerable complexity17:24
devanandayuriyz: you'll need a dict to store many of them, eg {"%s@%s:%s" % (user, host, port) : ssh_obj}17:25
devanandayuriyz: and a try..except..retry hook somewhere to handle failures due to connections timing out or being closed underneath the ssh_obj17:26
*** romcheg1 has quit IRC17:26
devanandaand some garbage collection of old ssh_obj wouldn't be unreasonable17:26
devanandaNobodyCam: https://bugs.launchpad.net/ironic/+bug/126041417:28
devanandalucasagomes: what d oyou think of the comments re: PEP257 here https://review.openstack.org/#/c/60954/3/ironic/api/controllers/v1/chassis.py ?17:29
lucasagomesdevananda, I will comment there. But although it's not mandatory, I think it's valid to add the blank line17:31
lucasagomessince it's help the Emac users to use their function17:31
*** martyntaylor has quit IRC17:31
lucasagomesand some pep257 checkers look at it17:32
lucasagomes$ pep257 ironic/*17:32
lucasagomes/home/lucasagomes/Projects/ironic/ironic/tests/drivers/test_ipminative.py:19:0: Multiline docstring should end with 1 blank line.17:32
*** blamar has quit IRC17:33
devanandalucasagomes: hmm. i don't think this is widely applied to openstack17:34
lucasagomesdevananda, I see17:35
lucasagomeswell it's optional so I think we shouldn't -1 for that17:36
lucasagomesI still use on my patches tho: https://review.openstack.org/#/c/60890/3/ironic/common/utils.py17:36
* lucasagomes apart from allowing emac users to run their commands, and pep257 checkers point to it, I also think it looks better17:37
devanandathere seem to be several things in http://www.python.org/dev/peps/pep-0257/#multi-line-docstrings that we as a community don't adhere to17:37
lucasagomesdevananda, true, we don't have any checks on hacking for that17:37
lucasagomesso it's not enforced17:38
devanandai'm looking eg. at nova.compute.manager17:38
devanandaand there is no apparent consistency17:38
devanandasingle-line doc strings with no following blank line. multiline doc strings with no blank line after the first line.17:38
devanandaetc17:38
lucasagomesyea17:38
devanandasome have a blank line before closing """, some do not17:39
devanandait's actually pretty ugly ...17:39
lucasagomesdevananda, do you think that we - as a project - should create our convention around it?17:39
lucasagomesand then, new patches should follow that17:39
devanandaonly if we can enforce it programatically17:39
lucasagomes+ we can fix the old code with time (like opening a bug for it)17:40
lucasagomesdevananda, we could do in the gate, but I'm afraid we should fix it all before17:40
devanandanah. if we add a pep check for 257 (if we can?) then we'll need to fix it all at once or gate will break17:40
lucasagomesI mean fix the old code17:40
devanandaright17:40
lucasagomesdevananda, yea17:40
devanandaand relying on humans to review for this is tedious and burdensome AND guaranteed to miss a few17:40
*** vetalll has joined #openstack-ironic17:40
devanandait's either in our flake8 cehck or we don't -1 for it17:41
lucasagomesyea17:41
lucasagomesflake8 don't test pep257 unfortunately17:41
*** hemnafk is now known as hemna17:41
lucasagomesmaybe some cosmetic things only17:41
NobodyCamdevananda: -1'd17:42
openstackgerritDevananda van der Veen proposed a change to openstack/ironic: Fix auto-generated REST API formatting  https://review.openstack.org/6095417:42
lucasagomesdevananda, we can start by opening a bug around it, and reviewing it as the new patches comes17:42
lucasagomesand once we are confident we might want to add it to the gate (idk how it would work tho)17:43
lucasagomespeople already -1 patches cause of pep257 in ironic, so no surprise here17:43
lucasagomesconfident = ported the old code17:43
devanandaand rebasing17:44
openstackgerritDevananda van der Veen proposed a change to openstack/ironic: Fix auto-generated REST API formatting  https://review.openstack.org/6095417:44
devanandalucasagomes: i'm checkign to see if we can add pep257 to flake17:45
lucasagomesdevananda, right :)17:46
*** yuriyz has quit IRC17:46
openstackgerritDevananda van der Veen proposed a change to openstack/ironic: Developer doc index page points to correct API docs  https://review.openstack.org/6096017:49
devanandarloo: hi! do you still have objections to https://review.openstack.org/#/c/59795/9 ? I see you and lucas talked in the comments...17:51
devanandaalso, lucasagomes, you've commented on ^ several times but not +/-'d lately ?17:52
rloodevananda: in meeting, give me 30 min.17:52
devanandarloo: np17:52
lucasagomesdevananda, I commented on rloo review, and left on my todo list to review17:54
devanandalucasagomes: ack. just checking :)17:54
lucasagomesI will do a review soon :)17:54
NobodyCamlucasagomes: i'm looking at https://review.openstack.org/#/c/54593/12/ironic/conductor/rpcapi.py17:57
NobodyCamare we trying to get away from node_id? and only use UUID?17:57
lucasagomesNobodyCam, in the API yes17:58
*** derekh has quit IRC17:58
lucasagomesinternally we still can use the numeric id17:58
lucasagomesfor e.g that's a rpc call, so if used internally I don't see any problem in passing the numeric id to that function17:59
NobodyCamthe tests pass uuid do you see any reason to test passing a int?18:01
lucasagomesNobodyCam, not really, because it would be redundant. We have a function that add a filter to the query18:04
lucasagomesto make the search by uuid or id18:04
lucasagomesI think that function shoult test it and see if that works18:04
lucasagomesadding tests by uuid and by id on all other tests sounds redundant in my opnion18:05
*** mdurnosvistov has joined #openstack-ironic18:06
lucasagomesNobodyCam, https://github.com/openstack/ironic/blob/master/ironic/db/sqlalchemy/api.py#L6518:09
lucasagomesthat function should test it18:09
lucasagomesI mean, tests for that function should test it18:09
*** vkozhukalov has joined #openstack-ironic18:12
NobodyCamok ..18:13
*** matty_dubs|lunch is now known as matty_dubs18:14
rloodevananda: wrt https://review.openstack.org/#/c/59795/9. the only thing is https://review.openstack.org/#/c/59795/9/ironic/db/api.py,unified, my first comment. not a big deal, except that it isn't a list, nor 'lists' :-)18:19
devanandarloo: ahh. grammar in my comment. thanks :)18:19
rloodevananda: is there a way to delete comments after they are added?18:20
devanandarloo: not afaik18:21
openstackgerritDevananda van der Veen proposed a change to openstack/ironic: Add prepare, clean_up, take_over methods to deploy  https://review.openstack.org/5889418:22
lucasagomesas I came to the office today, I gotta run now :) devananda, NobodyCam, rloo, everyone, have a good night :D18:22
rloodevananda: too bad.18:22
rloociao lucasagomes!18:22
NobodyCamhave a safe trip home lucasagomes18:23
NobodyCamneed to do walkies.... brb18:23
*** lucasagomes has quit IRC18:24
openstackgerritA change was merged to openstack/ironic: API: Add sample() method on Port and PortCollection  https://review.openstack.org/6175218:24
openstackgerritA change was merged to openstack/ironic: API: Validate and normalize address  https://review.openstack.org/6089018:24
openstackgerritA change was merged to openstack/ironic: Imported Translations from Transifex  https://review.openstack.org/5962718:24
*** ndipanov has quit IRC18:26
devanandag'night, lucas!18:26
openstackgerritdekehn proposed a change to openstack/ironic: Method to generate PXE options for Neutron ports  https://review.openstack.org/6090418:27
*** mdenny has joined #openstack-ironic18:38
NobodyCamand back18:45
NobodyCamdevananda: what can we do about this : https://review.openstack.org/#/c/57550/18:52
*** datajerk has joined #openstack-ironic18:54
*** blamar has joined #openstack-ironic18:59
*** romcheg has joined #openstack-ironic19:12
openstackgerritDevananda van der Veen proposed a change to openstack/ironic: Add prepare, clean_up, take_over methods to deploy  https://review.openstack.org/5889419:22
devanandaNobodyCam: someone needs to rebase it19:22
* devananda does it19:23
NobodyCam:-p19:23
* devananda kicks the interwebs19:23
openstackgerritDevananda van der Veen proposed a change to openstack/ironic: PEP8-ify imports in test_ipmitool  https://review.openstack.org/5755019:26
openstackgerritDevananda van der Veen proposed a change to openstack/ironic: Add prepare, clean_up, take_over methods to deploy  https://review.openstack.org/5889419:29
* NobodyCam makes a bagle19:29
* devananda steps away again, also foor food19:29
openstackgerritVitalii Lebedynskyi proposed a change to openstack/ironic: Do not use node argument in drivers  https://review.openstack.org/6185920:01
vetalllHi all20:03
vetalllRecently i joined to your community. Now i have completed first step of my task.20:05
NobodyCamvetalll: task?20:08
NobodyCamwelcome20:08
*** rloo has quit IRC20:09
*** rloo has joined #openstack-ironic20:09
romchegNobodyCam: I think vetalll's talking about this bug https://bugs.launchpad.net/ironic/+bug/125934620:10
vetalllIn the day before yesterday i take bug that was reported by NobodyCam20:10
romchegMorning, btw :)20:10
NobodyCammornig romcheg and vetalll :)20:10
romchegNobodyCam: good evening :) vetalll is my intern20:10
NobodyCamahh :)20:11
NobodyCamwelcome20:11
NobodyCamvetalll: I have not looked at the entire patch20:13
NobodyCambut the first file i looked at is still passing the node?20:14
NobodyCamhttps://review.openstack.org/#/c/61859/1/ironic/conductor/utils.py20:15
NobodyCamit is marked as wip20:15
romchegNobodyCam: I think it's better not change everything with one patch20:16
romchegNobodyCam: This patch removes usage of passed node in drivers20:17
NobodyCamahh would help if I actually read the commit message :-p20:17
NobodyCamromcheg: ++20:17
vetalllYes. i cannot complete this bug without touching of big peace of code.20:18
NobodyCamI totaly under stand. :)20:18
NobodyCamI did not read the commit msg20:19
lifelessdevananda: does ironic support tags on nodes today?20:19
devanandalifeless: to some definition of support, yes20:28
devanandalifeless: we have an 'extra' field which supports arbitrary key/value pairs. however, Ironic has no capacity to search on that field today20:29
rloohey devananda. question about https://review.openstack.org/#/c/58894/13/ironic/drivers/modules/pxe.py,unified20:33
rloodevananda: _create_token_file() is in deploy(), and _destroy_token_file() is in clean_up()20:33
rloodevananda: seems a bit 'unsymmetrical'.20:33
devanandavetalll: good start to the bug! i've left feedback on the patch20:34
devanandarloo: correct. perhaps taht needs a comment, but it is intentional20:35
devanandarloo: the token file is needed only briefly, in the time between deploy() finishing and _continue_deploy() starting20:35
devanandarloo: so you'll see taht _continue_deploy also removes the token file20:35
rloodevananda, yeah i saw that. why not put _destroy_token_file in tear_down?20:36
devanandarloo: I added taht to clean_up as well just to be on the safe side -- it shouldn't error if the file is already gone -- just in case a deploy is interrupted and tear_down is never called20:36
devanandawhich is an assumption on my part at this time, since we haven't finished writing the undeploy code paths in conductor.manager yet20:37
rloodevananda. hmm. ok, i know it won't err, i just like the symmetry of stuff in prepare -> cleanup, and deploy -> teardown. that's all.20:37
devanandaso I think this is the right way, and even if it's not, calling _destroy_token_file an extra time won't hurt20:37
rlooand calling _create_token_file in prepare doesn't make as much sense.20:38
devanandarloo: right. callign it in prepare is actually incorrect behavior20:38
devanandathe token file should only exist on the node performing the deploy20:39
rloodevananda: ok, thx.20:39
* devananda really wants to land https://review.openstack.org/#/c/60954/ soon20:40
rloodevananda, what's holding back 60954? or are you just impatient?20:41
devanandarloo: just impatient :)20:43
rloodevananda: ha ha ha20:43
*** romcheg has quit IRC20:45
*** romcheg has joined #openstack-ironic20:45
*** kui has joined #openstack-ironic20:47
NobodyCamdevananda: on 60954 your using uuid as param in several places (ie. chassis and node) yet in port you stick with the port_uuid nomenclature20:49
devanandaNobodyCam: it's a mess20:49
NobodyCambut you want to land it20:49
NobodyCamlol20:49
*** jcooley_ has quit IRC20:49
devanandaNobodyCam: but see my comment to lucas in a prior patch. I dont want to mix clean-up-param-names with -fix the doc formatting20:49
devanandaNobodyCam: i'm about to just rip this patch into pieces if it doesn't land soon -- and just have one patch that does nothing but adds whitespace to fix the WSME bug20:50
devanandamaybe I should do that ....20:50
NobodyCamahh. just checking YOU are planning a clean up patch too.20:51
devanandawell. it needs to happen20:51
devanandai'm fine if someone else does it :)20:51
NobodyCamhow about a comment in the commit msg that states you are NOT cleaning up the param names and that SHOULD be done in another patch20:53
rlooi thought lucas opened a bz about the param names but i could be wrong.20:54
NobodyCamrloo: I think you are correct but I now have way to many windows open to look20:55
NobodyCamlol20:55
devanandahttps://bugs.launchpad.net/ironic/+bug/125221320:55
rloohttps://bugs.launchpad.net/ironic/+bug/125961920:55
devanandathat one20:55
openstackgerritA change was merged to openstack/ironic: PEP8-ify imports in test_ipmitool  https://review.openstack.org/5755020:55
devanandawoops. yes, rloo, you got the right link20:56
* devananda has too many bugs open20:56
rloodevananda: https://review.openstack.org/#/c/58894/. I don't see a unit test for take_over. I don't really care though.20:59
devanandarloo: https://review.openstack.org/#/c/58894/13/ironic/tests/drivers/test_fake.py20:59
devanandaooh21:00
devanandayou mean in pxe driver21:00
rloothat's the test?! devananda, that doesn't even touch pxe.py. i'll give it a pass but ... :-)21:00
devanandadamn, yea, i missed that21:00
devanandalemme add that real quick. it /should/ be there21:00
rloook, i'll wait... :-)21:00
NobodyCambrb21:02
devanandarloo: hummmm. so, pxe._update_neutron() doesn't actually do anything -yet-21:03
devanandarloo: so a unit test isn't going to do anything either :)21:04
rloodevananda: yes, I know. that's why i suggested but didn't really care.21:04
devanandarloo: i can test that take_over calls it, but that's it21:04
openstackgerritDevananda van der Veen proposed a change to openstack/ironic: Add prepare, clean_up, take_over methods to deploy  https://review.openstack.org/5889421:05
devanandadone21:05
rloodevananda. it would only have been a placeholder, to be filled in whenever someone added something to _update_neutron().21:05
*** rwsu has quit IRC21:05
openstackgerritDevananda van der Veen proposed a change to openstack/ironic: Add prepare, clean_up, take_over methods to deploy  https://review.openstack.org/5889421:06
*** epim has joined #openstack-ironic21:09
romchegOk folks, it's quite late here now21:14
romchegGotta go now21:14
romchegHave a good evening.21:14
NobodyCamnight romcheg21:15
NobodyCamand vetalll too21:16
devanandag'night!21:20
devanandai need to step away for a bit again... should be back later this afternoon21:20
vetalllgood night you to21:20
NobodyCam:)21:20
*** vetalll has quit IRC21:23
devanandalifeless: re: nova blueprints, i havent been following baremetal-preserve-ephemeral as closely as I should (and will) be21:35
devanandalifeless: that being said, are you guys going to port that to ironic?21:35
lifelessdevananda: we'll need to yes21:36
lifelessor rather21:36
lifelesswe'll need the feature21:36
devanandalifeless: great. i'd say add Ironic to that BP, but LP doesn't support that, so please track it however you feel appropriate (new BP in ironic, or bug(s))21:37
lifelessthere should be a bug with a task on ironic already21:37
devanandalifeless: https://bugs.launchpad.net/ironic/+bug/1174154?21:38
*** romcheg has left #openstack-ironic21:38
lifelessdevananda: yes21:41
devanandalifeless: how do you want to track all the relevant nova patches within ironic? port each one with a partial-bug tag?21:49
lifelessfor that blueprint ?21:50
*** jdob has quit IRC21:50
devanandafor that feature, yes21:51
devanandagotta run out for an errand... bbiafh21:53
lifelessI'd wait for it to be a little more settled in nova then make a matching patch chain for it yeah22:03
*** linggao has quit IRC22:07
*** mdurnosvistov has quit IRC22:15
*** vkozhukalov has quit IRC22:15
*** matty_dubs is now known as matty_dubs|gone22:32
NobodyCambrb22:37
openstackgerritHaomeng,Wang proposed a change to openstack/ironic: Command call should log the stdout and stderr  https://review.openstack.org/6166423:47
« Wednesday, 2013-12-11 Index Friday, 2013-12-13 »

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