| *** anniec has quit IRC | 00:00 | |
| NobodyCam | brb | 00:01 |
|---|---|---|
| *** epim has quit IRC | 00:02 | |
| *** anniec has joined #openstack-ironic | 00:07 | |
| *** epim has joined #openstack-ironic | 00:09 | |
| *** harlowja has quit IRC | 00:40 | |
| *** harlowja has joined #openstack-ironic | 00:42 | |
| *** epim has quit IRC | 00:53 | |
| *** jcooley_ has joined #openstack-ironic | 00:56 | |
| *** epim has joined #openstack-ironic | 00:57 | |
| *** anniec has quit IRC | 00:59 | |
| *** grimtrigger has joined #openstack-ironic | 01:02 | |
| grimtrigger | is this the place to discuss diskimage-builder? if not, where? | 01:18 |
| *** nosnos has joined #openstack-ironic | 01:42 | |
| pleia2 | grimtrigger: I'd say #tripleo is the best place | 01:44 |
| Haomeng | grimtrigger: diskimage-builder should be one component tripleo project | 01:44 |
| Haomeng | grimtrigger: ironic is consumer for diskimage-builder | 01:44 |
| *** sjing has joined #openstack-ironic | 01:54 | |
| *** epim has quit IRC | 01:54 | |
| *** jcooley_ has quit IRC | 02:03 | |
| *** jcooley_ has joined #openstack-ironic | 02:04 | |
| *** jcooley_ has quit IRC | 02:08 | |
| *** anniec has joined #openstack-ironic | 02:10 | |
| *** rongze has joined #openstack-ironic | 02:12 | |
| sjing | Morning, Ironic | 02:13 |
| *** jcooley_ has joined #openstack-ironic | 02:19 | |
| *** epim has joined #openstack-ironic | 02:33 | |
| grimtrigger | haomeng: thanks muchly | 02:43 |
| Haomeng | grimtrigger: welcome:) | 02:48 |
| *** epim has quit IRC | 02:48 | |
| *** grimtrigger has left #openstack-ironic | 02:49 | |
| *** anniec has quit IRC | 03:07 | |
| *** datajerk has joined #openstack-ironic | 03:27 | |
| *** jcooley_ has quit IRC | 03:29 | |
| *** jcooley_ has joined #openstack-ironic | 03:30 | |
| *** jcooley_ has quit IRC | 03:34 | |
| *** anniec has joined #openstack-ironic | 03:37 | |
| *** anniec has quit IRC | 03:45 | |
| openstackgerrit | Haomeng,Wang proposed a change to openstack/ironic: Expose created_at/updated_at properties in the REST API https://review.openstack.org/53627 | 03:50 |
| *** rloo has quit IRC | 04:02 | |
| *** datajerk has joined #openstack-ironic | 04:27 | |
| *** datajerk has quit IRC | 04:31 | |
| *** rongze has quit IRC | 04:50 | |
| *** prekarat has joined #openstack-ironic | 04:54 | |
| *** rongze has joined #openstack-ironic | 04:54 | |
| *** jcooley_ has joined #openstack-ironic | 04:55 | |
| *** jcooley_ has quit IRC | 05:22 | |
| *** jcooley_ has joined #openstack-ironic | 05:22 | |
| *** jcooley_ has quit IRC | 05:27 | |
| *** datajerk has joined #openstack-ironic | 05:28 | |
| *** urulama has joined #openstack-ironic | 05:31 | |
| *** datajerk has quit IRC | 05:32 | |
| *** datajerk has joined #openstack-ironic | 05:58 | |
| openstackgerrit | Jenkins proposed a change to openstack/ironic: Imported Translations from Transifex https://review.openstack.org/59627 | 06:00 |
| *** datajerk has quit IRC | 06:02 | |
| *** blamar has quit IRC | 06:12 | |
| *** anniec has joined #openstack-ironic | 06:14 | |
| *** anniec_ has joined #openstack-ironic | 06:16 | |
| *** anniec_ has quit IRC | 06:17 | |
| *** anniec has quit IRC | 06:18 | |
| *** vkozhukalov has joined #openstack-ironic | 06:26 | |
| *** datajerk has joined #openstack-ironic | 06:29 | |
| *** nosnos_ has joined #openstack-ironic | 06:30 | |
| *** nosnos has quit IRC | 06:33 | |
| *** datajerk has quit IRC | 06:34 | |
| *** urulama has quit IRC | 06:45 | |
| *** rainya has quit IRC | 06:49 | |
| *** vkozhukalov has quit IRC | 06:51 | |
| *** max_lobur_afk is now known as max_lobur | 06:57 | |
| *** Haomeng has quit IRC | 06:57 | |
| *** Haomeng has joined #openstack-ironic | 06:57 | |
| *** datajerk has joined #openstack-ironic | 07:00 | |
| *** mdurnosvistov has joined #openstack-ironic | 07:04 | |
| *** datajerk has quit IRC | 07:05 | |
| *** urulama has joined #openstack-ironic | 07:09 | |
| *** datajerk has joined #openstack-ironic | 07:31 | |
| *** datajerk has quit IRC | 07:35 | |
| *** ifarkas has joined #openstack-ironic | 07:46 | |
| *** mdurnosvistov has quit IRC | 08:16 | |
| *** mdurnosvistov has joined #openstack-ironic | 08:16 | |
| *** mdurnosvistov has quit IRC | 08:21 | |
| *** davidlenwell_ has joined #openstack-ironic | 08:29 | |
| *** vkozhukalov has joined #openstack-ironic | 08:29 | |
| *** sanek11 has joined #openstack-ironic | 08:30 | |
| *** ndipanov has joined #openstack-ironic | 08:30 | |
| *** ifarkas_ has joined #openstack-ironic | 08:34 | |
| *** romcheg1 has joined #openstack-ironic | 08:34 | |
| *** lifeless has quit IRC | 08:34 | |
| *** romcheg has quit IRC | 08:34 | |
| *** ifarkas has quit IRC | 08:34 | |
| *** agordeev has quit IRC | 08:34 | |
| *** davidlenwell has quit IRC | 08:34 | |
| *** rongze_ has joined #openstack-ironic | 08:35 | |
| *** lifeless has joined #openstack-ironic | 08:36 | |
| *** rongze has quit IRC | 08:37 | |
| *** sjing has quit IRC | 08:44 | |
| *** jistr has joined #openstack-ironic | 08:48 | |
| *** Haomeng has quit IRC | 08:57 | |
| *** Haomeng has joined #openstack-ironic | 08:58 | |
| *** nosnos_ has quit IRC | 09:29 | |
| *** nosnos has joined #openstack-ironic | 09:29 | |
| *** sanek11 is now known as agordeev | 09:30 | |
| *** mdurnosvistov has joined #openstack-ironic | 09:40 | |
| *** lucasagomes has joined #openstack-ironic | 09:54 | |
| *** nosnos_ has joined #openstack-ironic | 10:44 | |
| *** nosnos has quit IRC | 10:48 | |
| *** max_lobur is now known as max_lobur_afk | 10:50 | |
| *** yfujioka has joined #openstack-ironic | 10:58 | |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Refactor /node/<uuid>/state https://review.openstack.org/60299 | 11:02 |
| *** yfujioka has quit IRC | 11:12 | |
| *** rongze_ has quit IRC | 11:17 | |
| *** rongze has joined #openstack-ironic | 11:28 | |
| *** tatyana has joined #openstack-ironic | 11:33 | |
| *** urulama has quit IRC | 11:42 | |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Add PatchType class https://review.openstack.org/60194 | 11:45 |
| *** urulama has joined #openstack-ironic | 11:46 | |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Rework patch validation on Ports https://review.openstack.org/56984 | 11:49 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Reworks Node validations https://review.openstack.org/57495 | 11:51 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Refactor /node/<uuid>/state https://review.openstack.org/60299 | 11:52 |
| *** urulama has quit IRC | 12:00 | |
| *** martyntaylor has joined #openstack-ironic | 12:01 | |
| *** urulama has joined #openstack-ironic | 12:06 | |
| *** urulama has quit IRC | 12:19 | |
| *** martyntaylor has quit IRC | 12:20 | |
| *** urulama has joined #openstack-ironic | 12:25 | |
| *** lucasagomes is now known as lucas-hungry | 12:30 | |
| *** martyntaylor has joined #openstack-ironic | 12:35 | |
| *** urulama has quit IRC | 12:39 | |
| *** vkozhukalov has quit IRC | 12:41 | |
| *** vkozhukalov has joined #openstack-ironic | 12:45 | |
| *** urulama has joined #openstack-ironic | 12:45 | |
| *** martyntaylor1 has joined #openstack-ironic | 12:50 | |
| *** martyntaylor has quit IRC | 12:51 | |
| *** urulama has quit IRC | 12:59 | |
| *** urulama has joined #openstack-ironic | 13:05 | |
| *** nosnos_ has quit IRC | 13:11 | |
| *** urulama has quit IRC | 13:15 | |
| *** urulama has joined #openstack-ironic | 13:16 | |
| *** urulama has quit IRC | 13:24 | |
| *** urulama has joined #openstack-ironic | 13:24 | |
| *** jdob has joined #openstack-ironic | 13:28 | |
| *** urulama has quit IRC | 13:34 | |
| *** urulama has joined #openstack-ironic | 13:35 | |
| *** lucas-hungry is now known as lucasagomes | 13:35 | |
| *** urulama_ has joined #openstack-ironic | 13:44 | |
| *** urulama has quit IRC | 13:45 | |
| *** urulama_ has quit IRC | 13:49 | |
| *** urulama has joined #openstack-ironic | 13:54 | |
| *** urulama has quit IRC | 14:04 | |
| *** urulama has joined #openstack-ironic | 14:04 | |
| *** urulama has quit IRC | 14:14 | |
| *** urulama has joined #openstack-ironic | 14:14 | |
| *** urulama has quit IRC | 14:18 | |
| *** urulama has joined #openstack-ironic | 14:23 | |
| *** matty_dubs|gone is now known as matty_dubs | 14:28 | |
| *** urulama has quit IRC | 14:33 | |
| *** urulama_ has joined #openstack-ironic | 14:33 | |
| *** russellb is now known as rustlebee | 14:38 | |
| *** urulama_ has quit IRC | 14:43 | |
| *** urulama has joined #openstack-ironic | 14:43 | |
| openstackgerrit | Alexandra Fomina proposed a change to openstack/python-ironicclient: Sync exceptions from Oslo https://review.openstack.org/60528 | 14:47 |
| *** yomeyo has joined #openstack-ironic | 14:49 | |
| *** urulama has quit IRC | 14:53 | |
| *** urulama_ has joined #openstack-ironic | 14:53 | |
| *** urulama_ has quit IRC | 14:59 | |
| *** rongze has quit IRC | 15:00 | |
| *** urulama has joined #openstack-ironic | 15:03 | |
| *** ndipanov is now known as nick_long | 15:04 | |
| *** yuriyz has quit IRC | 15:08 | |
| *** yomeyo has quit IRC | 15:08 | |
| *** vkozhukalov has quit IRC | 15:12 | |
| *** urulama has quit IRC | 15:13 | |
| *** urulama_ has joined #openstack-ironic | 15:13 | |
| *** datajerk has joined #openstack-ironic | 15:15 | |
| *** rongze has joined #openstack-ironic | 15:16 | |
| *** urulama_ has quit IRC | 15:23 | |
| *** urulama has joined #openstack-ironic | 15:23 | |
| openstackgerrit | Alexandra Fomina proposed a change to openstack/python-ironicclient: Use improved exceptions functionality from Oslo blueprint common-client-library-2 https://review.openstack.org/60528 | 15:32 |
| *** urulama has quit IRC | 15:33 | |
| *** urulama has joined #openstack-ironic | 15:33 | |
| *** urulama has quit IRC | 15:37 | |
| *** vkozhukalov has joined #openstack-ironic | 15:40 | |
| *** urulama has joined #openstack-ironic | 15:42 | |
| openstackgerrit | A change was merged to openstack/python-ironicclient: Enable rebooting in the client lib and cli https://review.openstack.org/60353 | 15:49 |
| *** urulama_ has joined #openstack-ironic | 15:52 | |
| *** urulama has quit IRC | 15:53 | |
| lucasagomes | it's so quiet here today | 15:53 |
| *** tatyana has quit IRC | 15:54 | |
| romcheg1 | Morning lucasagomes :) | 15:56 |
| lucasagomes | romcheg1, hey buddy how are u? | 15:56 |
| lucasagomes | morning/afternoon :) | 15:56 |
| romcheg1 | lucasagomes: Everything is great by my wisdom tooth and unfinished revolution in Ukraine :) | 15:57 |
| openstackgerrit | Alexandra Fomina proposed a change to openstack/python-ironicclient: Use improved exceptions functionality from Oslo https://review.openstack.org/60528 | 15:57 |
| *** romcheg1 is now known as romcheg | 15:57 | |
| romcheg | *but | 15:57 |
| lucasagomes | romcheg, haha I see | 15:57 |
| lucasagomes | over here things r same, calm and cold :) | 15:58 |
| romcheg | lucasagomes: I've seen some pro-Ukrainian protests in Dublin :) | 15:59 |
| *** urulama_ has quit IRC | 16:02 | |
| *** urulama has joined #openstack-ironic | 16:02 | |
| *** bigjools has quit IRC | 16:05 | |
| lucasagomes | romcheg, great! many people showed up? I'm sure the citizens of the world are supporting u guys | 16:06 |
| openstackgerrit | A change was merged to openstack/python-ironicclient: Move from inheritance HTTPClient in Ironic client https://review.openstack.org/60187 | 16:06 |
| *** bigjools has joined #openstack-ironic | 16:07 | |
| *** romcheg has quit IRC | 16:08 | |
| *** jcooley_ has joined #openstack-ironic | 16:10 | |
| *** urulama_ has joined #openstack-ironic | 16:12 | |
| *** urulama has quit IRC | 16:13 | |
| *** r-mibu has quit IRC | 16:22 | |
| *** viktors has left #openstack-ironic | 16:22 | |
| *** urulama has joined #openstack-ironic | 16:22 | |
| *** urulama_ has quit IRC | 16:23 | |
| *** r-mibu has joined #openstack-ironic | 16:23 | |
| *** urulama has quit IRC | 16:28 | |
| *** urulama has joined #openstack-ironic | 16:31 | |
| *** mdurnosvistov has quit IRC | 16:35 | |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/python-ironicclient: Rename nodes/<uuid>/state to nodes/<uuid>/states https://review.openstack.org/60558 | 16:40 |
| *** urulama_ has joined #openstack-ironic | 16:41 | |
| *** urulama has quit IRC | 16:42 | |
| *** vkozhukalov has quit IRC | 16:43 | |
| *** urulama_ has quit IRC | 16:51 | |
| *** urulama has joined #openstack-ironic | 16:51 | |
| *** urulama has quit IRC | 16:53 | |
| *** urulama has joined #openstack-ironic | 17:01 | |
| *** jcooley_ has quit IRC | 17:10 | |
| *** urulama has quit IRC | 17:10 | |
| *** urulama has joined #openstack-ironic | 17:11 | |
| devananda | morning, all | 17:12 |
| NobodyCam | good mornig Ironic | 17:12 |
| NobodyCam | :) | 17:12 |
| NobodyCam | hi devananda | 17:12 |
| NobodyCam | :) | 17:12 |
| * NobodyCam is enjoying 1st cup of coffee | 17:12 | |
| lucasagomes | morning devananda NobodyCam | 17:17 |
| NobodyCam | moning lucasagomes | 17:19 |
| NobodyCam | hey lucasagomes is not string.replace deperacated? | 17:20 |
| NobodyCam | see bottom of http://docs.python.org/2/library/string.html | 17:20 |
| lucasagomes | NobodyCam, not that I know | 17:21 |
| *** urulama_ has joined #openstack-ironic | 17:21 | |
| *** urulama has quit IRC | 17:21 | |
| lucasagomes | http://docs.python.org/2/library/stdtypes.html#str.replace | 17:21 |
| lucasagomes | http://docs.python.org/3.3/library/stdtypes.html#str.replace | 17:21 |
| NobodyCam | the link i posted lists it as deprecated | 17:22 |
| lucasagomes | but that's the string module | 17:22 |
| lucasagomes | ohh | 17:22 |
| lucasagomes | yes I can see now | 17:22 |
| lucasagomes | I mean, it's deperacate if you use the string module, but the built-in for string types | 17:23 |
| lucasagomes | doesn't seems to be | 17:23 |
| NobodyCam | ahh ok | 17:23 |
| lucasagomes | :) | 17:25 |
| NobodyCam | I'll refactor that patch | 17:25 |
| *** datajerk has quit IRC | 17:25 | |
| *** urulama_ has quit IRC | 17:26 | |
| lucasagomes | right | 17:29 |
| *** urulama has joined #openstack-ironic | 17:31 | |
| *** Maha has joined #openstack-ironic | 17:31 | |
| devananda | lucasagomes: why is 60299 removing the states/provision end point? | 17:32 |
| lucasagomes | devananda, it was just returning NotImplemented, I thought about reintroducing it for our deploy trigger | 17:33 |
| lucasagomes | raising* | 17:33 |
| devananda | lucasagomes: we'll need it, ofc | 17:34 |
| devananda | lucasagomes: also as I was reading this, i realised we will need several more states/xxx controllers, eg. for console, volume, etc | 17:34 |
| *** matty_dubs is now known as matty_dubs|lunch | 17:35 | |
| lucasagomes | devananda, oh yea, you think it I should reintroduce it on that patch the way it was? (just raising NotImplementedError) | 17:35 |
| lucasagomes | devananda, yes, so we can just use the _custom_method map dictonary for the additional controllers | 17:35 |
| devananda | right | 17:36 |
| devananda | just pointing out we'll have more | 17:36 |
| openstackgerrit | Chris Krelle proposed a change to openstack/ironic: Correct error with unicode mac address https://review.openstack.org/60300 | 17:36 |
| devananda | and yes, i would prefer this patch preserve the endpoint, even if it's stll just raising NotImplemented (which is different from NotFound) | 17:37 |
| NobodyCam | brb... post breakfast walkies | 17:37 |
| devananda | lucasagomes: i also posted some comments on the commit message just now | 17:37 |
| lucasagomes | devananda, right, thanks for the review I will update the patch | 17:38 |
| *** urulama has quit IRC | 17:41 | |
| *** urulama_ has joined #openstack-ironic | 17:41 | |
| *** urulama_ has quit IRC | 17:50 | |
| *** urulama has joined #openstack-ironic | 17:51 | |
| *** mdurnosvistov has joined #openstack-ironic | 17:55 | |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Refactor /node/<uuid>/state https://review.openstack.org/60299 | 17:58 |
| *** urulama has quit IRC | 18:00 | |
| *** urulama has joined #openstack-ironic | 18:01 | |
| *** urulama has quit IRC | 18:05 | |
| *** urulama has joined #openstack-ironic | 18:10 | |
| *** urulama_ has joined #openstack-ironic | 18:20 | |
| *** urulama has quit IRC | 18:20 | |
| *** matty_dubs|lunch is now known as matty_dubs | 18:23 | |
| *** ifarkas_ has quit IRC | 18:23 | |
| NobodyCam | lucasagomes: still around? | 18:26 |
| lucasagomes | NobodyCam, yup | 18:26 |
| NobodyCam | off the wall question. | 18:26 |
| NobodyCam | have you put any thought in to ipv6 addresses in ironic | 18:26 |
| *** martyntaylor1 has left #openstack-ironic | 18:26 | |
| lucasagomes | NobodyCam, uuu not really | 18:28 |
| NobodyCam | hehehe nor I .. | 18:28 |
| NobodyCam | just putting it out there | 18:28 |
| lucasagomes | does iscsi supports it? | 18:28 |
| NobodyCam | I would think. though I have not looked | 18:29 |
| lucasagomes | yea at first glance it does (googled here) | 18:29 |
| *** urulama_ has quit IRC | 18:29 | |
| lucasagomes | but def worth some mapping on potential problems with that | 18:30 |
| lucasagomes | wondering if neutron already does ipv6 | 18:30 |
| *** urulama has joined #openstack-ironic | 18:30 | |
| lucasagomes | NobodyCam, https://blueprints.launchpad.net/neutron/+spec/ipv6-feature-parity | 18:30 |
| *** nick_long has quit IRC | 18:30 | |
| NobodyCam | :) | 18:32 |
| devananda | NobodyCam, lucasagomes: not sure if I pointed it out yesterday -- we have a page on http://status.openstack.org/reviews/ for ironic and ironicclient now. | 18:33 |
| devananda | useful for prioritizing what we review, when our time is too short to review /everything/ | 18:33 |
| devananda | like, as we approach a release, that will help us spot the reviews that pertain to bugs tagged HIGH or CRITICAL, or blueprints targeted to the upcoming release, etc, etc | 18:34 |
| *** urulama has quit IRC | 18:35 | |
| NobodyCam | ahh cool!!! | 18:35 |
| lucasagomes | o wow! I didn't know such thing existed! | 18:35 |
| devananda | :) | 18:35 |
| openstackgerrit | A change was merged to openstack/ironic: Correct error with unicode mac address https://review.openstack.org/60300 | 18:37 |
| *** urulama has joined #openstack-ironic | 18:37 | |
| *** jistr has quit IRC | 18:45 | |
| *** urulama has quit IRC | 18:48 | |
| *** hemna has joined #openstack-ironic | 18:49 | |
| *** urulama has joined #openstack-ironic | 18:50 | |
| *** vkozhukalov has joined #openstack-ironic | 18:50 | |
| devananda | hmm | 18:53 |
| NobodyCam | humm? | 18:53 |
| devananda | https://review.openstack.org/#/c/55549/12/ironic/api/controllers/v1/node.py | 18:53 |
| devananda | lucasagomes' comment got me thinking about something we talked about the other day | 18:53 |
| devananda | whether or not there should be any sort of admin-CLI-like tool for doing direct db manipulation | 18:54 |
| openstackgerrit | A change was merged to openstack/ironic: Expose created_at/updated_at properties in the REST API https://review.openstack.org/53627 | 18:54 |
| devananda | like, say, breaking a lock | 18:54 |
| devananda | as opposed to exposing that in the API,w hich as lucasagomes pointed out, is a bit awkward if we allow POST but not GET for some resource | 18:54 |
| devananda | actually, lucasagomes, is that so very different from GET /v1/nodes/xxx/states && POST /v1/ndoes/xxx/states/power ? | 18:55 |
| devananda | the more i think about exposing POST /v1/nodes/xxx/reservation, the more I think we should not enable this. "if we build it, they will use it" and this really shouldn't be (ab)used | 18:56 |
| devananda | it's an implementation detail of our current lock mechanism. not a user-settable property of the node | 18:57 |
| lucasagomes | devananda, I think it's OK to POST but not GET from a specific resource if that resource offers a Location that the user can use to track the request | 18:59 |
| lucasagomes | if it's an async call ofc | 18:59 |
| lucasagomes | devananda, I see, but the way it's on the patch | 19:00 |
| lucasagomes | is awkward as well because 'reservation' is not part of the document | 19:00 |
| lucasagomes | the idea of patch is to patch the document | 19:00 |
| devananda | right | 19:00 |
| lucasagomes | so user GET it and then PATCH parts of it | 19:00 |
| lucasagomes | reservation not being there looks like an easter egg | 19:00 |
| lucasagomes | some hidden feature | 19:01 |
| devananda | ah, yea. it's not even in the GET /v1/nodes/xxx | 19:01 |
| *** Maha has quit IRC | 19:01 | |
| lucasagomes | yes | 19:01 |
| lucasagomes | I think I gave two suggestion, one is to include it on the document | 19:01 |
| lucasagomes | so that we can allow peple patch just like it's being done on that review | 19:02 |
| devananda | so that is a valid objection to the way this is implemented. i'm also raising another concern -- is the API even the right way to expose this functionality? | 19:02 |
| devananda | if we let folks break locks this easily (just a PATCH request) then they will do exactly that | 19:02 |
| lucasagomes | devananda, I don't like it, I know it's needed to break the lock | 19:02 |
| lucasagomes | but I don't like it to be that easy u know | 19:02 |
| lucasagomes | for me the lock could be break if the node is for e.g in an failed/error state | 19:03 |
| lucasagomes | and still has a lock | 19:03 |
| devananda | the only time this functionality should be necessary, INSHO, is if an ironic-conductor process dies suddenly, while holding a lock | 19:04 |
| lucasagomes | devananda, https://review.openstack.org/#/c/55549/6/ironic/api/controllers/v1/node.py | 19:04 |
| devananda | so, I'm not sure I agree there | 19:05 |
| devananda | let's say someone calls driver.power.reboot() | 19:05 |
| devananda | a TaskManager lock is held while that runs, and waits for the node to come back online | 19:05 |
| devananda | and then the power supply on the ironic-conductor host explodes | 19:05 |
| devananda | and the node actually powers on successfully | 19:06 |
| lucasagomes | yea right, but we have infos on the db that shows that the conductor helding that lock died | 19:06 |
| lucasagomes | so then I think we can break it | 19:06 |
| devananda | the DB will show "target_power_state": power on. "current_power_state": power off. "reservation": some-dead-host | 19:06 |
| devananda | no error | 19:06 |
| *** urulama has quit IRC | 19:07 | |
| devananda | with the heartbeat / consistent hashing code taht i've been working on, another conductor will be able to take over for the dead one | 19:07 |
| devananda | but it won't break the lock | 19:07 |
| devananda | and as we discussed at the summit, automatic breaking of locks will be dangerous once we start talkign to BIOS and firmware | 19:08 |
| devananda | perhaps we add a check in the API that the lock being broken is held by a dead conductor? | 19:08 |
| devananda | that feels kludgy :( | 19:09 |
| lucasagomes | yea, hmm so with the consistent hashing in place, does we need it? since there's a takeover without breaking the lock | 19:09 |
| devananda | the takeover is just goign to be another conductor calling some not-yet-written takeover method | 19:11 |
| lucasagomes | and after the takeover will it release the node? once the target_* states set to null? | 19:11 |
| devananda | takeover_node() won't be able to get a lock on the node without breaking the lock first (if a lock was held) | 19:12 |
| devananda | setting target_* is not the same as breaking the lock | 19:12 |
| lucasagomes | I thought the takeover_node would force acquiring the node someway | 19:13 |
| lucasagomes | independent of the lock | 19:13 |
| devananda | at ConductorManager level, it'll do things like call the node.driver.deploy.prepare() | 19:14 |
| devananda | we need the PXE driver to update neutron at that point | 19:14 |
| devananda | neither of those require an exclusive lock right now (and i dont think they should) | 19:15 |
| lucasagomes | +1 prepare just do some setups... then deploy() needs the lock | 19:15 |
| devananda | normal path: prepare, @lock deploy | 19:16 |
| devananda | takeover path: prepare, notify_neutron | 19:16 |
| devananda | so i dont think takeover requires a lock | 19:17 |
| lucasagomes | prepare, notify_neutron, break the lock, check the target_states to know what the states that node want's to transition to | 19:17 |
| lucasagomes | and do the work | 19:17 |
| lucasagomes | or check the target_states first | 19:18 |
| devananda | well, takeover might not have to break a lock | 19:18 |
| devananda | yea | 19:18 |
| devananda | if the node was already running fine and no one had a lock, then takeover is just recreating the deploy environment (tftp config, etc) and updting neutron | 19:18 |
| lucasagomes | yea | 19:18 |
| devananda | so, unwinding a bit - where'd we get to w.r.t. the API for breaking a lock? :) | 19:19 |
| lucasagomes | heh I think that if we are exposing that reservation field on the document, the patch seems ok | 19:25 |
| lucasagomes | idk lemme try to find an example of it | 19:25 |
| lucasagomes | devananda, but you don't think we need some checks anyway? | 19:27 |
| devananda | checks? | 19:27 |
| lucasagomes | node.reservation not in get_list_of_active_conductors() | 19:27 |
| lucasagomes | break | 19:27 |
| devananda | ah. yea, something like that as a sanity check seems reasonable | 19:27 |
| lucasagomes | devananda, is the reservation something we want to expose in a node-show? | 19:28 |
| lucasagomes | or it's just internal and people shouldn't care about it | 19:28 |
| devananda | unsure | 19:28 |
| lucasagomes | I think that if we expose it I'm fine PATCH nodes/<uuid> | 19:29 |
| devananda | it's like nova exposing the host name where your VM is running | 19:29 |
| devananda | which it does not | 19:29 |
| lucasagomes | if not I think the subresource looks better PUT {'lock': 'break'} nodes/<uuid>/reservation | 19:30 |
| lucasagomes | both doing the sanity check to see if that conductor is really dead | 19:30 |
| devananda | hm. correcting myself | 19:33 |
| devananda | nova *does* expose the hypervisor_hostname | 19:33 |
| devananda | where your VM is running | 19:33 |
| devananda | | OS-EXT-SRV-ATTR:hypervisor_hostname | 44458af3-3700-4dbc-87a1-d9868aa864ac | | 19:33 |
| devananda | granted, it's an extension, but it's enabled by default, sooo.... | 19:34 |
| lucasagomes | hehe so PATCH ? :) | 19:36 |
| lucasagomes | maybe before any decision | 19:36 |
| lucasagomes | let's add to the meeting on monday | 19:36 |
| devananda | ++ | 19:36 |
| lucasagomes | so we can grab more ideas | 19:37 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/python-ironicclient: driver-list command to show the list of conductors https://review.openstack.org/60602 | 19:37 |
| lucasagomes | devananda, will add on the FFT topic there | 19:38 |
| devananda | thanks! | 19:38 |
| lucasagomes | done :D np | 19:40 |
| lucasagomes | I'm done for today :) | 19:44 |
| lucasagomes | have a good night and a good weekend everyone | 19:44 |
| lucasagomes | devananda, NobodyCam g'night | 19:44 |
| devananda | lucasagomes: thanks! you too! | 19:44 |
| *** lucasagomes has quit IRC | 19:45 | |
| devananda | NobodyCam: how do you feel about the ironic walkthrough using python-ironicclient locally instead of from inside the undercloud vm? | 19:47 |
| *** epim has joined #openstack-ironic | 19:47 | |
| NobodyCam | :) just need to install it localy | 19:48 |
| NobodyCam | but ya should all be good | 19:48 |
| devananda | install or use its venv? | 19:48 |
| NobodyCam | could use venv. | 19:49 |
| *** mwagner_lap has joined #openstack-ironic | 19:52 | |
| NobodyCam | devananda: note second to last line: http://paste.openstack.org/show/AslZxmNfru9KCaD7xGSc/ | 19:56 |
| *** urulama has joined #openstack-ironic | 19:56 | |
| devananda | fyi, i'm adding a few things to ironic wiki | 19:57 |
| NobodyCam | awesome :) | 19:57 |
| devananda | current walkthrough worked fine | 19:57 |
| NobodyCam | which walk thru? we have three | 19:58 |
| NobodyCam | :-p | 19:58 |
| devananda | i'm going to add some script bits for installing the client and using it from the host | 19:59 |
| devananda | heh. the dib walkthrough | 19:59 |
| NobodyCam | :) sweet | 19:59 |
| devananda | NobodyCam: that paste seems to have unusually many Ports with teh same UUID | 19:59 |
| NobodyCam | its like four different log of the same thing really ... at different stages | 20:00 |
| NobodyCam | ports, P: , port | 20:01 |
| NobodyCam | :-p | 20:01 |
| *** urulama has quit IRC | 20:01 | |
| NobodyCam | that has lots of extra unneeded logging just for me | 20:01 |
| NobodyCam | devananda: should I patch https://github.com/openstack/tripleo-incubator/blob/master/scripts/setup-clienttools | 20:05 |
| NobodyCam | to include ironic client at this point | 20:05 |
| devananda | hmm | 20:05 |
| devananda | i need to push ironic up to pypi for that to work | 20:05 |
| devananda | we need to do that anyway | 20:06 |
| *** vkozhukalov has quit IRC | 20:06 | |
| NobodyCam | ya we have talked about just that several times | 20:06 |
| devananda | yep | 20:06 |
| devananda | lemme finish this wiki edit then i'll do that | 20:06 |
| NobodyCam | :) after that I'll patch setup-clienttools | 20:09 |
| *** rongze has quit IRC | 20:15 | |
| *** urulama has joined #openstack-ironic | 20:18 | |
| *** harlowja has quit IRC | 20:23 | |
| *** harlowja has joined #openstack-ironic | 20:23 | |
| devananda | https://wiki.openstack.org/wiki/Ironic#Install_the_Ironic_client | 20:24 |
| devananda | ok, the Test It section needs to go :) | 20:25 |
| *** urulama has quit IRC | 20:28 | |
| NobodyCam | :) | 20:30 |
| *** urulama has joined #openstack-ironic | 20:45 | |
| *** urulama has quit IRC | 20:58 | |
| *** urulama has joined #openstack-ironic | 21:11 | |
| devananda | NobodyCam: https://review.openstack.org/#/c/60614/2 in case you're curious how python-*client get released | 21:22 |
| *** urulama has quit IRC | 21:23 | |
| *** urulama_ has joined #openstack-ironic | 21:23 | |
| NobodyCam | i am | 21:25 |
| *** anniec has joined #openstack-ironic | 21:32 | |
| *** urulama_ has quit IRC | 21:33 | |
| *** urulama has joined #openstack-ironic | 21:33 | |
| *** urulama has quit IRC | 21:43 | |
| *** urulama has joined #openstack-ironic | 21:43 | |
| devananda | i'll push our first tag in ~20 minutes (want to give puppet enough time to get the job pipelines ready) | 21:44 |
| *** urulama has quit IRC | 21:48 | |
| NobodyCam | devananda: if I push a new driver up can you click the wip | 21:49 |
| devananda | sure | 21:52 |
| *** urulama has joined #openstack-ironic | 21:53 | |
| NobodyCam | devananda: please click wip button | 21:54 |
| NobodyCam | :) | 21:54 |
| devananda | awesome, tyvm | 21:56 |
| NobodyCam | no no thank you | 21:57 |
| NobodyCam | :-p | 21:57 |
| * NobodyCam wander away for some food!!!! | 21:57 | |
| NobodyCam | wanders even | 21:57 |
| NobodyCam | humm don't think that is my error: E: nova.conf.sample is not up to date, please run tools/config/generate_sample.sh | 21:59 |
| devananda | huh? | 22:00 |
| devananda | where do you see that | 22:00 |
| *** jdob has quit IRC | 22:01 | |
| NobodyCam | thats on zuul for pep8 check on the patch i just pushed | 22:01 |
| *** matty_dubs is now known as matty_dubs|gone | 22:01 | |
| NobodyCam | https://jenkins02.openstack.org/job/gate-nova-pep8/9714/ | 22:01 |
| *** urulama has quit IRC | 22:02 | |
| *** urulama has joined #openstack-ironic | 22:03 | |
| *** jcooley_ has joined #openstack-ironic | 22:03 | |
| *** urulama has quit IRC | 22:12 | |
| *** urulama has joined #openstack-ironic | 22:13 | |
| *** harlowja has quit IRC | 22:21 | |
| *** urulama_ has joined #openstack-ironic | 22:23 | |
| *** urulama has quit IRC | 22:23 | |
| *** jcooley_ has quit IRC | 22:24 | |
| *** epim has quit IRC | 22:24 | |
| *** Haomeng has quit IRC | 22:26 | |
| *** Haomeng has joined #openstack-ironic | 22:26 | |
| NobodyCam | updated my repo to pull patch #4 | 22:27 |
| *** epim has joined #openstack-ironic | 22:27 | |
| *** urulama_ has quit IRC | 22:33 | |
| *** urulama has joined #openstack-ironic | 22:33 | |
| devananda | neat! so nova is forcing that -- we should copy that jenins job | 22:34 |
| NobodyCam | :-P | 22:35 |
| devananda | NobodyCam: actually the error is | 22:36 |
| devananda | CRITICAL nova [-] Unable to import module nova.virt.ironic.driver | 22:36 |
| devananda | No module named ironicclient | 22:36 |
| NobodyCam | where are you getting that | 22:36 |
| devananda | https://jenkins02.openstack.org/job/gate-nova-pep8/9714/consoleFull | 22:37 |
| *** urulama has quit IRC | 22:38 | |
| NobodyCam | also looks like my local falake8 is not testing hte same as that | 22:39 |
| NobodyCam | flake8 even | 22:39 |
| NobodyCam | odd that that failes in the pep8 check and not the py27 or 26 checks | 22:40 |
| *** urulama has joined #openstack-ironic | 22:42 | |
| *** urulama has quit IRC | 22:52 | |
| *** urulama_ has joined #openstack-ironic | 22:52 | |
| NobodyCam | devananda: still here | 22:53 |
| NobodyCam | ? | 22:53 |
| *** urulama has joined #openstack-ironic | 23:02 | |
| *** urulama_ has quit IRC | 23:02 | |
| devananda | NobodyCam: ya | 23:06 |
| NobodyCam | one more wip click | 23:07 |
| devananda | also | 23:07 |
| devananda | https://pypi.python.org/pypi?:action=display&name=python-ironicclient&version=0.0.1 | 23:07 |
| NobodyCam | correcting the pep8 conplant | 23:07 |
| devananda | k | 23:07 |
| NobodyCam | hitting review now | 23:07 |
| NobodyCam | w00t!!!!! | 23:08 |
| * NobodyCam feels like steve martin when he found his name in hte phone book | 23:08 | |
| *** urulama has quit IRC | 23:12 | |
| *** urulama has joined #openstack-ironic | 23:12 | |
| NobodyCam | http://www.youtube.com/watch?v=-7aIf1YnbbU | 23:13 |
| devananda | LOL | 23:14 |
| devananda | things are happening ... | 23:14 |
| NobodyCam | hehehe | 23:14 |
| NobodyCam | gah | 23:15 |
| devananda | so actually, that nova driver will probably fail | 23:15 |
| devananda | until you can add python-ironiccilent to requirements.txt | 23:15 |
| devananda | so that jenkins will install it | 23:15 |
| NobodyCam | ya | 23:15 |
| devananda | which can't happen until it's in global-requirements AND the pypi mirrors get updated (later tonight) | 23:15 |
| *** urulama has quit IRC | 23:17 | |
| NobodyCam | devanda one more wip click | 23:19 |
| NobodyCam | its TODO(name) not TODO (name) | 23:20 |
| NobodyCam | :-p | 23:20 |
| *** urulama has joined #openstack-ironic | 23:22 | |
| *** anniec has quit IRC | 23:25 | |
| *** urulama_ has joined #openstack-ironic | 23:31 | |
| *** urulama has quit IRC | 23:32 | |
| devananda | heh | 23:33 |
| NobodyCam | :) but yep still need to update requirments.txt | 23:34 |
| NobodyCam | actually we import ironic too | 23:34 |
| devananda | eh? | 23:34 |
| devananda | should only be importing the client libs | 23:34 |
| devananda | ahhhh | 23:35 |
| devananda | 29# FIXME(deva): remove dependency on ironic | 23:35 |
| devananda | 30# refactor states into client | 23:35 |
| devananda | 31from ironic.common import states as ironic_states | 23:35 |
| NobodyCam | :) yep | 23:35 |
| devananda | THAT is wrong | 23:35 |
| devananda | needs to be fixed | 23:35 |
| NobodyCam | :) | 23:35 |
| devananda | it was an early hack :p | 23:35 |
| devananda | actually | 23:35 |
| NobodyCam | ya I will do that next | 23:35 |
| devananda | the problem is that ironic.common.states is needed by both python-ironicclient and by nova.drivers.ironic | 23:36 |
| devananda | both packages need to be sending the state strings in plain text | 23:36 |
| devananda | and i hate duplicating ENUM definitions | 23:36 |
| NobodyCam | they are different packages. what are you thinking some kind of api query to get the mapping???? | 23:40 |
| *** urulama_ has quit IRC | 23:41 | |
| NobodyCam | ironic states-list | 23:41 |
| NobodyCam | :-p | 23:41 |
| *** urulama has joined #openstack-ironic | 23:41 | |
| *** anniec has joined #openstack-ironic | 23:43 | |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Add tools/conf/check_uptodate to tox.ini https://review.openstack.org/60639 | 23:44 |
| devananda | i'm thinking that the client lib should know the mapping | 23:44 |
| devananda | also, ^ should fix our sample.conf is out of date problems | 23:45 |
| devananda | if I did the tox.ini change right | 23:45 |
| NobodyCam | :) | 23:45 |
| devananda | and https://review.openstack.org/60638 will add the client to requirements | 23:45 |
| NobodyCam | :) | 23:46 |
| *** urulama has quit IRC | 23:46 | |
| NobodyCam | devananda: on 60638 should it not be python-ironicclient>=0.0.1 ?/? | 23:49 |
| devananda | no - nothing's less than 0.0.1 | 23:49 |
| NobodyCam | just to keep formating | 23:50 |
| devananda | we dont need to pin a min version yet | 23:50 |
| NobodyCam | ok :) | 23:50 |
| devananda | nah. look up in the file. plenty of things don't pin | 23:50 |
| *** anniec has quit IRC | 23:50 | |
| NobodyCam | :) | 23:50 |
| NobodyCam | ya | 23:50 |
| *** anniec has joined #openstack-ironic | 23:50 | |
| devananda | when nova requires a feature that is in 0.0.x, and not in 0.0.1, we'll need to pin | 23:50 |
| NobodyCam | :) | 23:51 |
| *** urulama has joined #openstack-ironic | 23:51 | |
| *** anniec has quit IRC | 23:51 | |
| *** anniec has joined #openstack-ironic | 23:52 | |
| devananda | sweeet | 23:53 |
| NobodyCam | ??? | 23:53 |
| devananda | openstack mirrors still not up to date, but... | 23:53 |
| devananda | sudo pip install https://pypi.python.org/packages/source/p/python-ironicclient/python-ironicclient-0.0.1.tar.gz | 23:54 |
| devananda | works fine | 23:54 |
| NobodyCam | awesome ness | 23:54 |
| NobodyCam | I'll get OOO's setup client updated this weekend | 23:54 |
| NobodyCam | or at least put up the patch | 23:54 |
| NobodyCam | some land quick others ... lol | 23:55 |
| devananda | after tonight, you should be able to just say "pip install python-ironicclient" | 23:55 |
| devananda | there was a glitch in the job, so the mirrors didn't get updated as quickly as usual | 23:55 |
| NobodyCam | cool thats what setup-clienttools is doing along with some sym links | 23:57 |
| NobodyCam | https://github.com/openstack/tripleo-incubator/blob/master/scripts/setup-clienttools | 23:57 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!