| *** absubram has quit IRC | 00:00 | |
| *** baoli has joined #openstack-ironic | 00:01 | |
| *** suro-patz has quit IRC | 00:03 | |
| *** rama_y has quit IRC | 00:03 | |
| *** garthb has quit IRC | 00:11 | |
| *** jcook has quit IRC | 00:12 | |
| *** jcook has joined #openstack-ironic | 00:15 | |
| *** fragatin_ has joined #openstack-ironic | 00:36 | |
| *** fragatina has quit IRC | 00:38 | |
| *** thrash|g0ne has quit IRC | 00:38 | |
| *** xek has quit IRC | 00:39 | |
| *** adreznec has quit IRC | 00:39 | |
| *** adreznec has joined #openstack-ironic | 00:47 | |
| *** alexpilotti has quit IRC | 00:50 | |
| openstackgerrit | Merged openstack/ironic: Document how to run the tempest tests https://review.openstack.org/307695 | 00:57 |
|---|---|---|
| *** mtanino has quit IRC | 00:58 | |
| *** jcook has quit IRC | 00:59 | |
| *** jcook has joined #openstack-ironic | 01:01 | |
| *** namnh has joined #openstack-ironic | 01:09 | |
| *** namnh has quit IRC | 01:10 | |
| *** namnh has joined #openstack-ironic | 01:11 | |
| *** Haomeng has joined #openstack-ironic | 01:11 | |
| *** namnh has left #openstack-ironic | 01:11 | |
| *** phuongnh has joined #openstack-ironic | 01:16 | |
| *** vmud213 has joined #openstack-ironic | 01:32 | |
| *** hoangcx has joined #openstack-ironic | 01:32 | |
| *** thrash has joined #openstack-ironic | 01:35 | |
| *** thrash has quit IRC | 01:35 | |
| *** thrash has joined #openstack-ironic | 01:35 | |
| *** adreznec has quit IRC | 01:39 | |
| *** adreznec has joined #openstack-ironic | 01:41 | |
| *** absubram has joined #openstack-ironic | 01:46 | |
| *** absubram_ has joined #openstack-ironic | 01:47 | |
| zer0c00l | Anyone seen this http://www.markshuttleworth.com/archives/1493 | 01:48 |
| zer0c00l | Troll | 01:50 |
| *** alexpilotti has joined #openstack-ironic | 01:50 | |
| *** absubram has quit IRC | 01:51 | |
| *** absubram_ is now known as absubram | 01:51 | |
| Haomeng | zer0c00l: thanks for sharing, NSA-as-a-Service? | 01:53 |
| zer0c00l | Haomeng: Mark argues that providing ironic in a public cloud environment could open up physical hardware firmware level attacks | 01:54 |
| *** alexpilotti has quit IRC | 01:54 | |
| zer0c00l | UEFI secureboot is supposed to solve firmware tampering | 01:55 |
| Haomeng | zer0c00l: yes, it is security gap | 01:55 |
| Haomeng | zer0c00l: we support uefi:) | 01:56 |
| *** vmud213 has quit IRC | 02:02 | |
| openstackgerrit | Merged openstack/ironic-python-agent: Make "make coreos" output moveable without sudo https://review.openstack.org/305988 | 02:14 |
| *** ijw has quit IRC | 02:20 | |
| *** ijw has joined #openstack-ironic | 02:20 | |
| *** ijw has quit IRC | 02:21 | |
| *** Goneri has joined #openstack-ironic | 02:22 | |
| *** fragatin_ has quit IRC | 02:26 | |
| *** sivaramakrishna has joined #openstack-ironic | 02:31 | |
| *** baoli has quit IRC | 02:32 | |
| *** vmud213 has joined #openstack-ironic | 02:34 | |
| openstackgerrit | Zhenguo Niu proposed openstack/ironic: Add nodes tagging support - objs https://review.openstack.org/233357 | 02:39 |
| openstackgerrit | Zhenguo Niu proposed openstack/ironic: Add nodes tagging support - API https://review.openstack.org/250478 | 02:39 |
| openstackgerrit | Zhenguo Niu proposed openstack/ironic: Joined 'tags' column while getting node https://review.openstack.org/253065 | 02:39 |
| openstackgerrit | Zhenguo Niu proposed openstack/ironic: Add DB layer support for node tags filter https://review.openstack.org/272624 | 02:39 |
| *** baoli has joined #openstack-ironic | 02:44 | |
| *** Goneri has quit IRC | 02:49 | |
| *** baoli has quit IRC | 02:50 | |
| *** vmud213 has quit IRC | 03:02 | |
| *** sacharya_ has joined #openstack-ironic | 03:04 | |
| *** sacharya has quit IRC | 03:06 | |
| *** absubram has quit IRC | 03:13 | |
| *** links has joined #openstack-ironic | 03:30 | |
| *** alexpilotti has joined #openstack-ironic | 03:40 | |
| *** amotoki has joined #openstack-ironic | 03:42 | |
| *** alexpilotti has quit IRC | 03:44 | |
| *** vmud213 has joined #openstack-ironic | 03:55 | |
| *** chihhsin has joined #openstack-ironic | 03:55 | |
| *** chihhsin has quit IRC | 03:55 | |
| *** keedya has quit IRC | 03:56 | |
| *** chihhsin has joined #openstack-ironic | 03:56 | |
| *** harlowja_at_home has joined #openstack-ironic | 03:57 | |
| *** hoangcx_ has joined #openstack-ironic | 03:59 | |
| *** vishwanathj is now known as vishwanathj_zzz | 03:59 | |
| *** hoangcx_ has left #openstack-ironic | 03:59 | |
| *** hoangcx has quit IRC | 04:00 | |
| *** amotoki has quit IRC | 04:01 | |
| *** rbudden has joined #openstack-ironic | 04:05 | |
| *** bapalm has quit IRC | 04:10 | |
| *** amotoki has joined #openstack-ironic | 04:11 | |
| *** bapalm has joined #openstack-ironic | 04:12 | |
| *** xavierr has joined #openstack-ironic | 04:27 | |
| *** sacharya_ has quit IRC | 04:29 | |
| *** rbudden has quit IRC | 04:30 | |
| *** xavierr has quit IRC | 04:31 | |
| *** yuikotakada has joined #openstack-ironic | 04:31 | |
| *** sivaramakrishna has quit IRC | 04:32 | |
| *** trusted has joined #openstack-ironic | 04:32 | |
| *** trusted has quit IRC | 04:34 | |
| *** sivaramakrishna has joined #openstack-ironic | 04:34 | |
| *** jaybeale has joined #openstack-ironic | 04:37 | |
| *** harlowja_at_home has quit IRC | 04:38 | |
| *** ChrisAusten has quit IRC | 04:40 | |
| *** ChrisAusten has joined #openstack-ironic | 04:40 | |
| *** BadCub_ has joined #openstack-ironic | 04:40 | |
| *** lekha_ has joined #openstack-ironic | 04:40 | |
| *** zhenguo__ has joined #openstack-ironic | 04:40 | |
| *** jaybeale has quit IRC | 04:42 | |
| *** BadCub has quit IRC | 04:42 | |
| *** BadCub_ is now known as BadCub | 04:42 | |
| *** zhenguo_ has quit IRC | 04:43 | |
| *** lekha has quit IRC | 04:43 | |
| *** zhenguo__ is now known as zhenguo_ | 04:43 | |
| *** lekha_ is now known as lekha | 04:43 | |
| *** yuikotakada has quit IRC | 04:45 | |
| *** hoangcx has joined #openstack-ironic | 04:48 | |
| *** rcernin has quit IRC | 04:48 | |
| *** vinaym213 has joined #openstack-ironic | 04:59 | |
| *** vmud213 has quit IRC | 05:02 | |
| *** ozialien has quit IRC | 05:07 | |
| *** ayoung has quit IRC | 05:10 | |
| lazy_prince | Hi.. I have strange case where on the target node the IPA loads first and then the networking starts. Since IPA starts first and finds no network, its fails to deploy.. | 05:10 |
| *** ChrisAusten has quit IRC | 05:10 | |
| vinaym213 | lazy_prince:Do u see any error on the conductor like "NodeNotFound" or something like that? | 05:11 |
| lazy_prince | vinaym213: I just see the node in wait-call state and then it times out and the instance gets scheduled to other ironic node. | 05:12 |
| vinaym213 | Were u able to ping the node from conductor, before it times out waiting for the lookup from IPA? | 05:14 |
| vinaym213 | s/before it/before conductor | 05:15 |
| *** vinaym213 has quit IRC | 05:15 | |
| *** vmud213 has joined #openstack-ironic | 05:16 | |
| lazy_prince | vinaym213: yes, I was.. | 05:17 |
| *** rcernin has joined #openstack-ironic | 05:17 | |
| vmud213 | lazy_prince:Is the deploy image built with DIB and debuggable? I'd be interested to see if IPA service is properly started | 05:18 |
| lazy_prince | vmud213: its built with DIB but without devuser.. however, the console logs are available and that shows IPA starting first even before DHCP is tried for the network interfaces.. | 05:22 |
| *** e0ne has joined #openstack-ironic | 05:23 | |
| vmud213 | This appears to be an issue.. Because currently to my knowledge there are two bugs reported on the related issue. | 05:24 |
| vmud213 | Is it a fedora image? | 05:24 |
| * vmud213 looking for the patches similar to the issue | 05:24 | |
| vmud213 | https://review.openstack.org/#/c/303503 | 05:25 |
| lazy_prince | vmud213: its based on ubuntu.. | 05:26 |
| vmud213 | https://review.openstack.org/#/c/300548/ | 05:26 |
| vmud213 | lazy_prince: Can u go through those patches that i have just pasted | 05:27 |
| vmud213 | But u'r issue seems to be little different | 05:27 |
| * lazy_prince is looking | 05:27 | |
| *** alexpilotti has joined #openstack-ironic | 05:27 | |
| vmud213 | The first one is for deploy and second one is for inspection | 05:27 |
| *** ianw_ has quit IRC | 05:28 | |
| vmud213 | so i believe we may also have to wait for all interfaces to be up even in the case of deploy | 05:28 |
| *** ianw_ has joined #openstack-ironic | 05:28 | |
| *** sacharya has joined #openstack-ironic | 05:29 | |
| *** ianw_ has quit IRC | 05:30 | |
| *** ianw has joined #openstack-ironic | 05:31 | |
| *** alexpilotti has quit IRC | 05:32 | |
| *** ianw has quit IRC | 05:34 | |
| *** sacharya has quit IRC | 05:34 | |
| *** ianw has joined #openstack-ironic | 05:34 | |
| *** suro-patz has joined #openstack-ironic | 05:35 | |
| *** fragatina has joined #openstack-ironic | 05:40 | |
| *** fragatin_ has joined #openstack-ironic | 05:41 | |
| lazy_prince | vmud213: just to be clear, I am not using inspection but its deploy.. and here is the console log http://paste.openstack.org/show/494913/ | 05:42 |
| *** fragatina has quit IRC | 05:44 | |
| *** ChubYann has quit IRC | 05:45 | |
| *** suro-patz1 has joined #openstack-ironic | 05:48 | |
| *** suro-patz has quit IRC | 05:51 | |
| *** aarefiev22 has quit IRC | 05:51 | |
| *** itamarl has joined #openstack-ironic | 05:57 | |
| * vmud213 in meeting.Will look into the logs soon and see if i get any thing | 05:57 | |
| vmud213 | lazy_prince: I saw the log | 05:59 |
| vmud213 | I think it's ok even if the interface is not up by the time IPA service is started. | 06:00 |
| vmud213 | lazy_prince:IPA keeps trying for node lookup leaving some interval between each try. | 06:01 |
| lazy_prince | vmud213: aha.. okay.. then I will see if there is anything else thats causing problem.. | 06:02 |
| vmud213 | And will succeed once the interface is up.So interfaces should comeup atlease before the node moves to deploy_failed state | 06:03 |
| *** vinaym213 has joined #openstack-ironic | 06:06 | |
| *** moshele has joined #openstack-ironic | 06:09 | |
| *** vmud213 has quit IRC | 06:10 | |
| *** vinay50muddu has joined #openstack-ironic | 06:18 | |
| *** vmud213 has joined #openstack-ironic | 06:20 | |
| *** vinaym213 has quit IRC | 06:21 | |
| *** yuikotakada has joined #openstack-ironic | 06:23 | |
| *** jaybeale has joined #openstack-ironic | 06:27 | |
| *** jaybeale has quit IRC | 06:31 | |
| *** itamarl_ has joined #openstack-ironic | 06:33 | |
| *** itamarl_ has quit IRC | 06:33 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic: Imported Translations from Zanata https://review.openstack.org/307575 | 06:34 |
| *** aarefiev has quit IRC | 06:35 | |
| *** suro-patz1 has quit IRC | 06:36 | |
| *** itamarl has quit IRC | 06:36 | |
| *** yolanda has quit IRC | 06:37 | |
| *** jtomasek has quit IRC | 06:38 | |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Allow to set Neutron port setup delay from config https://review.openstack.org/293876 | 06:46 |
| *** yolanda has joined #openstack-ironic | 06:47 | |
| *** hoangcx_ has joined #openstack-ironic | 06:53 | |
| *** hoangcx has quit IRC | 06:54 | |
| *** aarefiev has joined #openstack-ironic | 07:00 | |
| *** e0ne has quit IRC | 07:01 | |
| *** karimb has joined #openstack-ironic | 07:02 | |
| *** tesseract has joined #openstack-ironic | 07:03 | |
| *** tesseract is now known as Guest14509 | 07:04 | |
| openstackgerrit | Ankit Kumar proposed openstack/ironic: Add support for building ISO for deploy ramdisk https://review.openstack.org/264590 | 07:05 |
| openstackgerrit | Ankit Kumar proposed openstack/ironic: Add support for dib based agent ramdisk in lib/ironic https://review.openstack.org/264579 | 07:05 |
| *** e0ne has joined #openstack-ironic | 07:07 | |
| openstackgerrit | Xavier proposed openstack/ironic: Add Dynamic Allocation feature for the OneView drivers https://review.openstack.org/286192 | 07:12 |
| *** alexpilotti has joined #openstack-ironic | 07:15 | |
| *** alexpilotti has quit IRC | 07:20 | |
| *** sacharya has joined #openstack-ironic | 07:31 | |
| *** e0ne has quit IRC | 07:32 | |
| *** mkovacik__ has joined #openstack-ironic | 07:35 | |
| *** sacharya has quit IRC | 07:35 | |
| *** mkovacik_ has quit IRC | 07:35 | |
| *** garyk1 has joined #openstack-ironic | 07:49 | |
| garyk1 | lifeless: hi, you around? | 07:51 |
| lifeless | garyk1: ish | 07:53 |
| *** ifarkas has joined #openstack-ironic | 07:54 | |
| *** hoangcx_ has quit IRC | 07:56 | |
| *** hoangcx has joined #openstack-ironic | 07:56 | |
| garyk1 | lifeless: have a few questions about irninc integrations with neutron. do you have some time to pick your brain? | 07:57 |
| openstackgerrit | Xavier proposed openstack/ironic: Updating links and removing unnecessary dollar symbol from doc page https://review.openstack.org/308821 | 07:59 |
| *** zzzeek has quit IRC | 08:00 | |
| *** zzzeek has joined #openstack-ironic | 08:00 | |
| lifeless | garyk1: sure | 08:01 |
| *** ohamada has joined #openstack-ironic | 08:02 | |
| *** mbound has joined #openstack-ironic | 08:09 | |
| *** mbound has quit IRC | 08:09 | |
| *** mbound has joined #openstack-ironic | 08:10 | |
| lucasagomes | morning all | 08:11 |
| *** jistr has joined #openstack-ironic | 08:13 | |
| *** athomas has joined #openstack-ironic | 08:14 | |
| *** jaybeale has joined #openstack-ironic | 08:15 | |
| *** MattMan has quit IRC | 08:18 | |
| *** MattMan has joined #openstack-ironic | 08:18 | |
| *** jaybeale has quit IRC | 08:20 | |
| aarefiev | morning! | 08:22 |
| yuikotakada | lucasagomes, aarefiev, morning! | 08:25 |
| aarefiev | yuikotakada: morning | 08:25 |
| openstackgerrit | Ghanshyam Mann proposed openstack/ironic: Fix usage of rest_client expected_success() in tests https://review.openstack.org/308834 | 08:25 |
| *** jtomasek has joined #openstack-ironic | 08:31 | |
| *** shu-mutou has quit IRC | 08:31 | |
| vmud213 | morning all | 08:32 |
| *** derekh has joined #openstack-ironic | 08:34 | |
| *** dtantsur|afk is now known as dtantsur | 08:36 | |
| vdrok | morning vmud213 lucasagomes yuikotakada and aarefiev | 08:36 |
| dtantsur | Morning Ironic | 08:36 |
| * dtantsur has a terrible headache today, so he's not quite responsive | 08:36 | |
| vmud213 | morning vdrok,dtantsur | 08:36 |
| vdrok | morning dtantsur, get better! | 08:37 |
| yuikotakada | vdrok, dtantsur, morning | 08:37 |
| *** mgould|afk is now known as mgould | 08:37 | |
| mgould | morning Ironic! | 08:37 |
| mgould | dtantsur, get well soon! | 08:37 |
| vdrok | morning mgould | 08:37 |
| aarefiev | morning vmud213 vdrok dtantsur mgould | 08:38 |
| aarefiev | dtantsur: I hope you get well soon | 08:38 |
| dtantsur | aarefiev, ouch, I forgot I have a -1 on test refactoring | 08:39 |
| aarefiev | dtantsur: I can take care on it | 08:39 |
| dtantsur | great, thanks! | 08:40 |
| * dtantsur has way too many things to do before the summit :( | 08:40 | |
| vmud213 | morning aarefiev | 08:40 |
| vsaienko | morning vdrok, vmud213, yuikotakada, dtantsur aarefiev, mgould and all ;) | 08:43 |
| aarefiev | morning vsaienko | 08:44 |
| vdrok | morning vsaienko :) | 08:44 |
| *** electrofelix has joined #openstack-ironic | 08:44 | |
| lucasagomes | vdrok, good morning :-) | 08:45 |
| lucasagomes | mgould, vsaienko yuikotakada aarefiev morning! | 08:45 |
| yuikotakada | morning, vsaienko :) | 08:45 |
| aarefiev | morning lucasagomes | 08:46 |
| vmud213 | vsaienko, lucasagomes: good morning | 08:46 |
| vmud213 | lucasagomes: I know it's really busy time for u, but when you get a chance today would you mind reviewing https://review.openstack.org/#/c/302143 ? | 08:48 |
| lucasagomes | vmud213, hi there, will take a look | 08:52 |
| vmud213 | lucasagomes: thanks. | 08:52 |
| mgould | vdrok aarefiev vmud213 lucasagomes yuikotakada morning! | 08:56 |
| yuikotakada | mgould, morning! | 08:57 |
| vmud213 | mgould o/ :) | 08:57 |
| *** amotoki has quit IRC | 09:02 | |
| *** alexpilotti has joined #openstack-ironic | 09:03 | |
| *** kromanenko has joined #openstack-ironic | 09:05 | |
| *** alexpilotti has quit IRC | 09:08 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: DevStack: Parametrize automated_clean https://review.openstack.org/308856 | 09:17 |
| vmud213 | lucasagomes: thanks for your time :). I have seen the comment. Will update in the new patch. | 09:21 |
| lucasagomes | cool :-) | 09:21 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic: Fixed nits in the new inspection doc page https://review.openstack.org/308862 | 09:22 |
| *** karimb has quit IRC | 09:23 | |
| *** sacharya has joined #openstack-ironic | 09:31 | |
| sambetts|afk | Morning all | 09:33 |
| *** sambetts|afk is now known as sambetts | 09:33 | |
| dtantsur | morning sambetts | 09:33 |
| openstackgerrit | Anton Arefiev proposed openstack/ironic-inspector: Test: do not merge https://review.openstack.org/308868 | 09:33 |
| sambetts | hey dtantsur | 09:33 |
| yuikotakada | sambetts, o/ | 09:33 |
| sambetts | hey yuikotakada! | 09:34 |
| *** sacharya has quit IRC | 09:36 | |
| aarefiev | morning sambetts | 09:37 |
| sambetts | devananda, krtaylor: So did we come to concusion that the reason we aren't leaving a verified is our fault or something misconfigured for Ironic in gerrit? | 09:37 |
| sambetts | Hey aarefiev | 09:37 |
| *** amotoki has joined #openstack-ironic | 09:59 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic-specs: Promote agent vendor passthru to core API https://review.openstack.org/306418 | 10:03 |
| dtantsur | the next iteration ^^^ | 10:03 |
| *** e0ne has joined #openstack-ironic | 10:12 | |
| *** athomas has quit IRC | 10:19 | |
| *** Guest14509 has quit IRC | 10:22 | |
| *** oberezovskyi has joined #openstack-ironic | 10:24 | |
| *** Guest14509 has joined #openstack-ironic | 10:24 | |
| *** hoangcx has quit IRC | 10:32 | |
| -openstackstatus- NOTICE: OVH servers are down, we are working to solve it. This will cause that jobs queue is processed slowly, please have patience. | 10:37 | |
| openstackgerrit | Ghanshyam Mann proposed openstack/ironic: Fix usage of rest_client expected_success() in tests https://review.openstack.org/308834 | 10:37 |
| *** baoli has joined #openstack-ironic | 10:41 | |
| *** baoli_ has joined #openstack-ironic | 10:44 | |
| *** alexpilotti has joined #openstack-ironic | 10:46 | |
| *** baoli has quit IRC | 10:47 | |
| vdrok | ;( | 10:48 |
| *** e0ne has quit IRC | 10:49 | |
| *** athomas has joined #openstack-ironic | 10:49 | |
| dtantsur | ouch... | 10:50 |
| *** sivaramakrishna has quit IRC | 10:53 | |
| *** mbound has quit IRC | 10:56 | |
| *** mbound has joined #openstack-ironic | 10:57 | |
| *** mkovacik_ has joined #openstack-ironic | 11:02 | |
| *** lucasagomes is now known as lucas-hungry | 11:05 | |
| *** amotoki has quit IRC | 11:07 | |
| *** __maurosr__ has joined #openstack-ironic | 11:07 | |
| *** mkovacik__ has quit IRC | 11:08 | |
| *** maurosr has quit IRC | 11:08 | |
| *** __maurosr__ is now known as maurosr | 11:08 | |
| jroll | morning y'all | 11:14 |
| sambetts | Hey jroll | 11:14 |
| *** itamarl has joined #openstack-ironic | 11:16 | |
| sambetts | dtantsur: you say that heartbeats are idempotent, but they change their response depending on if the node is locked right? Isn't that the same issue we had with idempotency in inspecotor | 11:17 |
| sambetts | ? | 11:17 |
| dtantsur | morning jroll | 11:18 |
| dtantsur | sambetts, hmm, similar, but not entirely. ironic has internal retries for locking, so lock failure will only be caused by a really long process in the heartbeat handler | 11:19 |
| dtantsur | which is not something we do, afaik | 11:19 |
| dtantsur | so in most cases (unless something is broken), the retry logic will cover the problem | 11:20 |
| dtantsur | also the heartbeat endpoint is async, so from user's standpoint nothing wrong happens even if node is locked | 11:21 |
| sambetts | dtantsur: I've seen node is locked messages in the IPA log when it tries to heartbeat | 11:21 |
| dtantsur | which raises an interesting question: usually we lock before returning to a caller, but maybe in case of heartbeats, we can always return | 11:21 |
| jroll | heartbeats also write to the db | 11:21 |
| jroll | I wouldn't call that idempotent :) | 11:21 |
| dtantsur | I can change it to avoid confusion, but wdyt about not not returning lock errors to a caller? | 11:22 |
| dtantsur | we could not do it with passthru, but here I guess we can | 11:22 |
| sambetts | not not? so we do return lock messages? | 11:22 |
| dtantsur | currently we do; but we can change it with the new API | 11:22 |
| jroll | dtantsur: well, if the node is locked the heartbeat won't be recorded, so I think we do want to return that error | 11:23 |
| jroll | to let the client know that the heartbeat was not successful | 11:23 |
| jroll | side note: this is silly and needs to be landed: https://review.openstack.org/#/c/308834/2 | 11:23 |
| dtantsur | jroll, "the client" is a ramdisk here; logging it on ironic side has much wider visibility | 11:23 |
| jroll | we do log it on ironic side no? | 11:23 |
| dtantsur | I hope so :) | 11:24 |
| dtantsur | ok, I'm just throwing ideas in the air, nevermind :) I'll change to POST | 11:24 |
| jroll | I believe we do log it | 11:24 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic-specs: Promote agent vendor passthru to core API https://review.openstack.org/306418 | 11:25 |
| dtantsur | sambetts, jroll ^^ | 11:25 |
| jroll | dtantsur: I'll review that if you can hit 308834, there's also a mitaka backport, that's blocking a tempest patch | 11:26 |
| jroll | :) | 11:26 |
| dtantsur | okie | 11:27 |
| sambetts | wow... why would you make .status and ['status'] different | 11:30 |
| sambetts | -_- | 11:30 |
| dtantsur | crazy API design, if you ask me | 11:30 |
| *** e0ne has joined #openstack-ironic | 11:30 | |
| dtantsur | +A | 11:30 |
| *** sacharya has joined #openstack-ironic | 11:32 | |
| jroll | as monty would say, it's totally bonghits | 11:32 |
| *** Goneri has joined #openstack-ironic | 11:32 | |
| sambetts | reading the agent api spec again, I really wish we could reuse the lookup endpoint for the claims/search thing :'( | 11:33 |
| dtantsur | yep :( except for authentication | 11:34 |
| jroll | well, I'm totally happy to block this until we do client cert auth :) | 11:34 |
| dtantsur | jroll, I'm not. it will block the driver composition | 11:34 |
| jroll | heh | 11:34 |
| sambetts | I wish there was someway to implement a one time pass thing for the agent to get auth info :/ | 11:34 |
| jroll | I'm half-joking | 11:34 |
| dtantsur | jroll, I don't believe it's possible to make a reliable auth here | 11:34 |
| jroll | s/reliable/secure/ | 11:34 |
| dtantsur | yep | 11:34 |
| jroll | except for client certs | 11:34 |
| dtantsur | no? | 11:35 |
| jroll | which means the distributed ramdisks may not be used | 11:35 |
| dtantsur | jroll, except for non-PXE environment using only virtual media | 11:35 |
| jroll | right | 11:35 |
| openstackgerrit | Kyrylo Romanenko proposed openstack/python-ironicclient: Use name randomizer from tempest https://review.openstack.org/307904 | 11:35 |
| dtantsur | anything you put on the ramdisk in the PXE environment can be fetched by an intruder | 11:35 |
| sambetts | i mean there isn't anything stopping a user using the this endpoint today right? | 11:35 |
| sambetts | unless you protect it by locking it down to a specifc IP range only or somehting | 11:36 |
| dtantsur | sambetts, right. one of the things why we need tenant separation | 11:36 |
| jroll | dtantsur: bleh, true | 11:36 |
| sambetts | even with tenant separation it doesn't get protected inherently, you still need some logic to check if its come from a specifc IP, and even then IP spoofing | 11:37 |
| dtantsur | in theory you can have different sets of endpoints for the provisioning network NIC and for user API NIC | 11:37 |
| *** sacharya has quit IRC | 11:37 | |
| sambetts | tbh, I think the ramdisk shouldn't talk to the Ironic API service | 11:37 |
| sambetts | seems like a layor break to me | 11:37 |
| sambetts | layer* | 11:37 |
| *** Goneri has quit IRC | 11:38 | |
| sambetts | it should be Ramdisk <--> conductor <--> API | 11:38 |
| dtantsur | this might happen one day | 11:38 |
| dtantsur | but the Ramdisk <--> conductor should be somehow protected as well | 11:39 |
| sambetts | the conductor can be plugged into the tenant network/ provisioning network, and you don't have to expose the Ironic API to that network, and you don't have to route the provisioning network out into the API space | 11:39 |
| dtantsur | true | 11:40 |
| dtantsur | when talking about inspector HA here in Brno we considered doing something similar for inspector | 11:40 |
| dtantsur | but it has its own weirdness iirc | 11:40 |
| jroll | how would IPA talk to the conductor? | 11:41 |
| sambetts | I mean its a pretty standard web app model, DB <--> application <--> web server | 11:41 |
| dtantsur | jroll, oslo.msg? | 11:41 |
| * dtantsur hides | 11:41 | |
| * jroll kicks dtantsur | 11:41 | |
| sambetts | jroll: it would have its own API implementing just heartbeat and lookup | 11:41 |
| * dtantsur flies away :( | 11:41 | |
| jroll | hrm | 11:41 |
| jroll | I mean, if we were going to do that, I'd just make a separate ironic-agent-api service that only has those endpoints | 11:42 |
| jroll | I don't want to put a web server in the conductor :) | 11:42 |
| dtantsur | makes some sense to me.. we could do the same for inspector too | 11:42 |
| sambetts | and you if made it a separate thing, e.g. conductor-agent, then you have one per namespace, then you could have IPA <-http-> agent <-rabit-> conductor <-rabit-> API <--> me | 11:42 |
| TheJulia | good mornin | 11:42 |
| dtantsur | morning TheJulia | 11:43 |
| jroll | morning TheJulia | 11:43 |
| sambetts | much much easier to fit tenant isolation into that model than the existing one | 11:43 |
| sambetts | and you get the benfit of not having unauthed APIs on your public API server | 11:44 |
| dtantsur | jroll, sambetts, do we want to put this on the agent API spec? | 11:44 |
| dtantsur | like, create a separate service? it's not as complex as it sounds | 11:44 |
| dtantsur | and if we agree on a core API, it will be hard to deprecate it... | 11:44 |
| jroll | hmm | 11:45 |
| sambetts | yeah pretty simple really, the conductor RPC api would already have to exist, so we'd just remove it from the Ironic API, and make a new thing that calls it | 11:46 |
| dtantsur | "remove it from Ironic API" <- remove what? removing a passthru is more or less easy, removing a core API endpoint is next to impossible | 11:47 |
| dtantsur | hence I ask if I should put this suggestion on my spec right now | 11:47 |
| sambetts | I guess that path doesn't exist just yet, but remove the existing path, and replace it with this one | 11:48 |
| jroll | there's lots of dragons here | 11:48 |
| sambetts | what I was saying is that the work needed to implement it would basicaly be the same, except instead of adding things to the Ironic API we'd add it to a new thing | 11:49 |
| jroll | if we ask ops to stand up a new service, we'll need to give them time to do so (probably at least until ocata) | 11:49 |
| TheJulia | jroll: tons... I think | 11:49 |
| jroll | not to mention the tons of bikeshedding on how the second api service works | 11:49 |
| dtantsur | jroll, we are not removing the old passthru yet | 11:49 |
| jroll | and all the docs to write on how these should be segregated | 11:49 |
| *** lucas-hungry is now known as lucasagomes | 11:49 | |
| TheJulia | jroll: and how to provide the separately provisioned tenant provisioning network that would ultimately kind of be required | 11:49 |
| dtantsur | yeah, maybe not worth it... but then we're committing to supporting the unauthenticated endpoints ~ forever | 11:49 |
| jroll | not saying we shouldn't do it, but I want to be clear it's a ton of work | 11:49 |
| sambetts | dtantsur: not completely, the new service might eventually support auth too, but it does mean it can be a different type of auth to the auth used by the public API | 11:50 |
| jroll | and it feels like one of those things that we'll never get done because people would prefer to argue about it rather than getting it done | 11:50 |
| dtantsur | oh that's so true... | 11:50 |
| sambetts | thats why we should just do it :-P | 11:51 |
| dtantsur | ok folks, let me put this all to "alternatives", and see how people react :) | 11:51 |
| * jroll promises to drink more coffee and think about this more | 11:52 | |
| sambetts | even if we have to add lookup and heartbeat to the main API initially while we work on this new service and then deprecated them in favor of that new endpoint | 11:52 |
| sambetts | I mean the logic for things wouldn't change just the URL | 11:52 |
| sambetts | as long as we make them the same | 11:53 |
| sambetts | but /me would prefer not to do that | 11:53 |
| sambetts | and /me really wants to add namespace support to the conductor and this would make it a hell of a lot easier | 11:54 |
| TheJulia | sambetts: ++ although, it makes things far far more complex networking wise at the same time | 11:54 |
| sambetts | complex is my middle name ;) hahaha | 11:55 |
| TheJulia | maybe a child conductor of sorts that we spawn off from the main conductor, keep an eye on it, and expect it to do all the work for the node | 11:55 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic-specs: Promote agent vendor passthru to core API https://review.openstack.org/306418 | 11:56 |
| dtantsur | updated alternatives ^^^ | 11:56 |
| dtantsur | I've put there that the same effect can probably be achieved by carefully configuring a WSGI container for ironic API | 11:56 |
| sambetts | TheJulia: thats essentially what I expect a conductor agent to be | 11:57 |
| *** amotoki has joined #openstack-ironic | 11:58 | |
| sambetts | TheJulia: its the only way I can see for dealing with tenant networks that might have conflicting IP ranges | 11:58 |
| TheJulia | yup | 11:59 |
| * sambetts has a crazy idea of creating an Ironic driver that uses soft power control for everything via a custom in instance agent, for use with a BMCless hardware, (using manual power control or potentially WOL if its supported, to initally kick the machine) but that requires the conductor can talk inside a tenant network at least a little bit | 12:01 | |
| TheJulia | sambetts: well, if people want to netboot a host forever.... | 12:01 |
| dtantsur | sambetts, you're not the first one with this idea. iirc someone from mirantis already proposed it | 12:02 |
| TheJulia | but that might be a separate maintenance service | 12:02 |
| *** phuongnh has quit IRC | 12:02 | |
| sambetts | TheJulia: also that, although that can be solved with the silly proxy thing I built initially :-P | 12:02 |
| TheJulia | sambetts: with isolation for the tftp namespace? | 12:03 |
| sambetts | dtantsur: damn ... every idea I come up with has always been taken haha | 12:03 |
| TheJulia | s/namespace/filesystem visible space/ | 12:03 |
| sambetts | TheJulia: yeah, making only certain parts of the tftp server visiable to a specific node is an interesting problem | 12:04 |
| * dtantsur hunts for food, brb | 12:04 | |
| dtantsur | sambetts, with iPXE (HTTP) it's much easier | 12:04 |
| sambetts | dtantsur: oooooh yeah, that makes it much easier | 12:04 |
| TheJulia | sambetts: I see it as much more an operational requirement | 12:04 |
| dtantsur | and with dynamic iPXE configuration which lucasagomes will hopefully finish one day, even more easy (maybe) | 12:05 |
| * lucasagomes reads | 12:05 | |
| *** dtantsur is now known as dtantsur|brb | 12:05 | |
| *** amotoki_ has joined #openstack-ironic | 12:09 | |
| *** amotoki has quit IRC | 12:11 | |
| *** amotoki_ has quit IRC | 12:11 | |
| *** alexpilotti has quit IRC | 12:14 | |
| *** krotscheck_dcm is now known as krotscheck | 12:15 | |
| *** amotoki has joined #openstack-ironic | 12:21 | |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Allow to set Neutron port setup delay from config https://review.openstack.org/293876 | 12:27 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Update ironic.config.sample https://review.openstack.org/308951 | 12:27 |
| openstackgerrit | Anton Arefiev proposed openstack/ironic-inspector: Refactor test_process https://review.openstack.org/304185 | 12:32 |
| *** dtantsur|brb is now known as dtantsur | 12:33 | |
| openstackgerrit | Mathieu Mitchell proposed openstack/ironic: Allow configuring shred's final overwrite with zeros https://review.openstack.org/304101 | 12:35 |
| *** moshele has quit IRC | 12:38 | |
| *** alexpilotti has joined #openstack-ironic | 12:39 | |
| *** e0ne has quit IRC | 12:39 | |
| *** e0ne has joined #openstack-ironic | 12:42 | |
| *** alexpilotti has quit IRC | 12:43 | |
| *** jaypipes has joined #openstack-ironic | 12:44 | |
| sambetts | dtantsur: laid out some comments in that patch pointing out the advantages/differences of a separate endpoint in relation to the existing spec | 12:46 |
| -openstackstatus- NOTICE: OVH provider is enabled again, please wait for the job queue to be processed | 12:48 | |
| openstackgerrit | Michael Krotscheck proposed openstack/ironic-webclient: Added provision state manipulation https://review.openstack.org/307928 | 12:50 |
| *** amotoki has quit IRC | 12:52 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: API: Check for reserved words when naming a node https://review.openstack.org/308965 | 12:53 |
| *** keedya has joined #openstack-ironic | 12:54 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: API: Check for reserved words when naming a node https://review.openstack.org/308965 | 12:56 |
| *** rbudden has joined #openstack-ironic | 12:58 | |
| *** links has quit IRC | 12:59 | |
| *** e0ne has quit IRC | 13:01 | |
| *** rloo has joined #openstack-ironic | 13:07 | |
| TheJulia | good morning rloo | 13:07 |
| rloo | GOOD morning TheJulia | 13:08 |
| dtantsur | morning rloo | 13:08 |
| sambetts | hey rloo o/ | 13:08 |
| rloo | hi dtantsur, sambetts | 13:08 |
| *** mtanino has joined #openstack-ironic | 13:09 | |
| lucasagomes | TheJulia, rloo sambetts jroll morning | 13:10 |
| jroll | morning lucasagomes, rloo :) | 13:11 |
| rloo | hiya lucasagomes, jroll | 13:11 |
| TheJulia | good morning lucasagomes | 13:12 |
| *** alexpilotti has joined #openstack-ironic | 13:13 | |
| *** thiagop has joined #openstack-ironic | 13:13 | |
| thiagop | Good morning Ironic | 13:14 |
| *** alexpilotti has quit IRC | 13:14 | |
| *** amotoki has joined #openstack-ironic | 13:16 | |
| rloo | hi thiagop | 13:17 |
| *** cdearborn has joined #openstack-ironic | 13:17 | |
| thiagop | rloo: o/ | 13:18 |
| *** mtanino has quit IRC | 13:20 | |
| openstackgerrit | Jim Rollenhagen proposed openstack/ironic: Add some docs about firmware security https://review.openstack.org/308978 | 13:20 |
| jroll | I'd love some eyes on this | 13:20 |
| jroll | TheJulia: devananda: ^ you will certainly have opinions | 13:21 |
| jroll | JayF: ^ | 13:21 |
| TheJulia | jroll: will look this morning | 13:21 |
| jroll | it's incomplete but /shrug | 13:21 |
| *** karimb has joined #openstack-ironic | 13:21 | |
| *** Goneri has joined #openstack-ironic | 13:23 | |
| TheJulia | early feedback just means faster iterration | 13:24 |
| jroll | yep :) | 13:25 |
| sambetts | jroll: is that in response to NSAaaS? | 13:25 |
| jroll | sambetts: timing is somewhat motivated by that, but we need it anyway | 13:26 |
| sambetts | its definatly a very interesting problem to work, and very hardware vendor specific | 13:27 |
| jroll | yep, which is why the best we can give is "work with your vendor" | 13:28 |
| jroll | you can't just buy something off the shelf and put it in a public ironic cloud (yet) | 13:28 |
| *** absubram has joined #openstack-ironic | 13:29 | |
| *** trown|outtypewww is now known as trown | 13:29 | |
| sambetts | jroll: do we provide OOB clean steps yet? | 13:29 |
| vdrok | morning sambetts jroll TheJulia rloo and thiagop | 13:29 |
| jroll | sambetts: yep, just plug them into the driver | 13:29 |
| TheJulia | but not everything in some cases :\ | 13:30 |
| jroll | sambetts: a clean step in an ironic driver can run any python code | 13:30 |
| *** vmud213 has quit IRC | 13:30 | |
| jroll | TheJulia: ? | 13:30 |
| jroll | hi vdrok \o | 13:30 |
| TheJulia | oob on the ilo driver can't see disks, so disk and disk backplane firmware can't be updated without the system being in a running state | 13:30 |
| *** alexpilotti has joined #openstack-ironic | 13:31 | |
| thiagop | Thanks TheJulia, looking into your comments now | 13:31 |
| jroll | TheJulia: well, that's an ilo problem, not an ironic problem :P | 13:31 |
| TheJulia | indeed, was just pointing out a case where oob won't alway solve the problem, which is likely to be the case with multiple vendors | 13:31 |
| jroll | right, can do things inband as well though | 13:32 |
| sambetts | I'm thinking combined OOB and in band cleaning, combined, e.g. I need if I need to flip some bits in the bios in OOB, to allow IPA to then do a cleaning step such as wiping and reseting the firmware, and then I need to flip them back into a secure state as like the last cleaning step | 13:32 |
| jroll | sambetts: :) | 13:32 |
| TheJulia | jroll: only so much in some cases, thinking network cards in particular | 13:32 |
| jroll | sambetts: we actually do that all inband, we have different bios settings for cleaning and tenants | 13:33 |
| krotscheck | betherly: There's still a few reviews on the ironic-webclient that need your approval, could you take a look? https://review.openstack.org/#/q/status:open+-label:Workflow-1+-(label:Code-Review%252B2%252Cself+OR+label:Code-Review%252B1%252Cself+OR+label:Code-Review-1%252Cself+OR+label:Code-Review-1%252Cself)+label:Verified+project:openstack/ironic-webclient | 13:33 |
| jroll | TheJulia: yeah, it all depends | 13:33 |
| *** sacharya has joined #openstack-ironic | 13:33 | |
| sambetts | jroll: do you have IPA flip the bits then? doesn't that mean that a tenant could also flip bits? | 13:33 |
| betherly | krotscheck: sure | 13:34 |
| jroll | sambetts: yep, it requires a password though, and there's protection against brute force | 13:34 |
| sambetts | jroll: oh right... I guess you have to get the password to IPA somehow :/ | 13:35 |
| *** garyk1 has quit IRC | 13:37 | |
| *** sacharya has quit IRC | 13:37 | |
| TheJulia | jroll: a few nits/thoughts posted inline, nothing really major though. | 13:38 |
| jroll | TheJulia: thanks | 13:39 |
| TheJulia | jroll: any chance of what you were thinking regarding the TODO item? | 13:39 |
| jroll | TheJulia: at a minimum, links to cleaning / clean steps docs | 13:39 |
| TheJulia | ++ | 13:39 |
| jroll | maybe also links to various articles about how people can own firmware | 13:40 |
| lucasagomes | sambetts, heh gotta love the NSAaaS thingy | 13:40 |
| jroll | for network security, at least blurbs about securing your bmc net | 13:40 |
| TheJulia | I was kind of thinking that might be goodd | 13:40 |
| TheJulia | jroll: also stressing separating bmc from production network | 13:40 |
| lucasagomes | acutally I wanted to make a t-shirt with mark shuttleworth face with a donald trump wig and a hitler mustache | 13:40 |
| lucasagomes | but no time for it before the summit I guess :-/ | 13:40 |
| jroll | TheJulia: yep | 13:40 |
| jroll | LOL | 13:40 |
| TheJulia | a list of best practices :) | 13:40 |
| TheJulia | lol | 13:41 |
| *** ametts has joined #openstack-ironic | 13:43 | |
| TheJulia | jroll: when the todo is filled out, I'll get my other half to take a look at it since she works in the infosec field | 13:43 |
| jroll | TheJulia: awesome, thanks | 13:43 |
| *** [1]cdearborn has joined #openstack-ironic | 13:46 | |
| *** aarefiev22 has joined #openstack-ironic | 13:53 | |
| *** mjturek1 has joined #openstack-ironic | 13:53 | |
| mjturek1 | wondering if someone could clarify something for me. When using the deprecated bash ramdisk is the cleaning step a noop or does it perform some kind of cleaning? | 13:56 |
| jroll | mjturek1: no cleaning in the bash ramdisk | 13:56 |
| jroll | it's a noop | 13:56 |
| mjturek1 | jroll: cool, thank you | 13:57 |
| jroll | np | 13:57 |
| *** cdearborn has quit IRC | 13:59 | |
| *** garthb has joined #openstack-ironic | 14:03 | |
| keedya | Good Morning All 0/ | 14:06 |
| jroll | hey keedya :) | 14:06 |
| keedya | hi Jroll | 14:06 |
| *** mtanino has joined #openstack-ironic | 14:08 | |
| sambetts | jroll: thoughts on -> https://review.openstack.org/#/c/308348 | 14:09 |
| rloo | TheJulia: about https://review.openstack.org/#/c/302819/. What does 'fallback' mean? | 14:11 |
| rloo | TheJulia: I mean, I don't know what will be done when it falls back. Falls back to what? | 14:12 |
| openstackgerrit | Andre keedy proposed openstack/ironic-inspector: Replace node data via uuid/data/unprocessed API https://review.openstack.org/308379 | 14:12 |
| rloo | TheJulia: or does it mean, it is up to the agent to decide/code the fallback? | 14:12 |
| *** e0ne has joined #openstack-ironic | 14:12 | |
| jroll | sambetts: I'll have to look later, stepping away for a bit | 14:12 |
| sambetts | jroll: sure :) | 14:13 |
| sambetts | thanks | 14:13 |
| TheJulia | rloo: it allows the configuration for an operator to choose to permit fallback from a failure of secure erase to shred. Under defaults, IPA fails presently and the node goes to cleanfail | 14:13 |
| rloo | TheJulia: ok, but when you permit 'fallback', what happens? what does the fallback do? | 14:14 |
| TheJulia | rloo: when you permit it, it falls back to shred | 14:14 |
| TheJulia | but by default, that is disabled | 14:14 |
| TheJulia | If the node just doesn't support secure erase, it will default to shred as well | 14:14 |
| rloo | TheJulia: OK, I think I parsed the sentence incorrectly. I thought the failure was the shredding. | 14:15 |
| *** daemontool has joined #openstack-ironic | 14:15 | |
| TheJulia | Ahhh | 14:16 |
| keedya | I need to update release notes and documentations for my review :https://review.openstack.org/308379. Can u point me to the indtructions? | 14:17 |
| rloo | TheJulia: I don't think the term 'fallback' is needed. Am trying to think of how to word it. | 14:17 |
| TheJulia | makes sense to me to use it since secure erase failing is a hard stop and move to cleanfail under defaults | 14:18 |
| TheJulia | but, there may be a better word :) | 14:18 |
| rloo | TheJulia: Am thinking, will comment in the patch. | 14:18 |
| TheJulia | ok | 14:19 |
| sambetts | keedya: http://docs.openstack.org/developer/ironic/dev/faq.html#id4 how to create a release note | 14:19 |
| rloo | TheJulia: another question, is there only one clean step in IPA where we try to do a secure erase? And/or if there is more than one clean step, this option would apply to all? | 14:19 |
| TheJulia | afaik one clean-step, but JayF and I are likely going to discuss alternative cleaning methods, I believe as he suggested separate hardware managers, for different types of disks that we don't really know what to do with now so things default to shred. | 14:21 |
| TheJulia | but, there is another IPA conversation I may be mixing up in my head since we've been talking about multipath IO as well | 14:21 |
| rloo | TheJulia. Ok, I was thinking of mentioning the actual clean step but better not. | 14:21 |
| sambetts | keedya: the docs to do with the API are in ironic-inspector/doc/source/http-api.rst | 14:21 |
| keedya | Thanks sambetts | 14:22 |
| openstackgerrit | Merged openstack/ironic: Updated from global requirements https://review.openstack.org/306942 | 14:23 |
| *** Goneri has quit IRC | 14:24 | |
| *** ayoung has joined #openstack-ironic | 14:31 | |
| *** jayc_ has joined #openstack-ironic | 14:31 | |
| *** alexpilo_ has joined #openstack-ironic | 14:32 | |
| *** alexpilotti has quit IRC | 14:33 | |
| *** Goneri has joined #openstack-ironic | 14:41 | |
| openstackgerrit | Andre keedy proposed openstack/ironic-inspector: Replace node data via uuid/data/unprocessed API https://review.openstack.org/308379 | 14:41 |
| openstackgerrit | Andre keedy proposed openstack/ironic-inspector: Replace node data via uuid/data/unprocessed API https://review.openstack.org/308379 | 14:45 |
| *** Goneri has quit IRC | 14:46 | |
| rloo | vdrok: about https://review.openstack.org/#/c/300983/, is it standard to use/look at only the last one? | 14:51 |
| vdrok | rloo: I think there is no standard, just the last one be the last before saving to DB :) | 14:52 |
| rloo | vdrok: so if someone invokes a command and tries to set the same argument more than once, your patch will ONLY look at the last one. | 14:53 |
| vdrok | rloo: the standard allows to specify as many values for the same path as you need, in our case I think we just need to make sure that the correct value is saved | 14:53 |
| vdrok | yes | 14:53 |
| rloo | vdrok: so if eg with name, the first name is incorrect, that won't cause it to fail with your patch. whereas it would fail now. | 14:54 |
| vdrok | rloo: yes, that is true | 14:55 |
| vdrok | you think all the values should be checked? | 14:56 |
| vdrok | my logic was to check only the one that will be actually stored | 14:56 |
| rloo | vdrok: i am puzzled by what our code is doing now. based on your description, it sounds like it checks the first name for correctness, but uses the last name? | 14:56 |
| rloo | vdrok: i'd think it would check all the names and use the last name. | 14:57 |
| vdrok | rloo: right now, yes. patches are applied consequently | 14:57 |
| rloo | vdrok: so the code now has a check for name1, apply name1, check name2, apply name2? | 14:58 |
| vdrok | rloo: that makes sense too, I can change that | 14:58 |
| vdrok | rloo: but still, check for name1, apply name1, check name2, apply name2, store name2 | 14:58 |
| vdrok | so only last one is stored | 14:59 |
| openstackgerrit | Andre keedy proposed openstack/ironic-inspector: Replace node data via uuid/data/unprocessed API https://review.openstack.org/308379 | 14:59 |
| vdrok | but I see your point | 14:59 |
| rloo | vdrok: well, that makes sense to me. I don't know what others think. But I think *just* using the last seems odd unless we decide that's what we want to do. | 14:59 |
| vdrok | jroll: lucasagomes how do you think - if multiple values for one patch are provided in single patch, should we check all the values, or only the last one | 15:00 |
| rloo | vdrok: whatever we do, will be similar behaviour for other fields, eg uuid | 15:00 |
| vdrok | yep | 15:01 |
| rloo | vdrok: you're going to hate me, but i was wondering what/whether there is some sort of consistency with the rest of openstack projects | 15:01 |
| *** Goneri has joined #openstack-ironic | 15:01 | |
| jroll | vdrok: as in, [{op:replace, key:name,value:foo}, {op:replace, key:name, value:bar}] ? | 15:01 |
| lucasagomes | vdrok, lemme read the scrollback. I'm out of context | 15:01 |
| vdrok | jroll: yes | 15:01 |
| jroll | I don't see why we'd care about the first if we don't write it | 15:01 |
| lucasagomes | its two updates for the same attribute in one patch? Yeah probably checking the last one is grand | 15:02 |
| lucasagomes | gotta see how json.patch handles that tho | 15:02 |
| * lucasagomes tries it out | 15:02 | |
| vdrok | lucasagomes: just applies consequently | 15:02 |
| rloo | jroll, lucasagomes, vdrok : so I'm not against *only* using the last one. I just want to make sure that's the behaviour we've all agreed on. | 15:03 |
| vdrok | rloo: I doubt that, lots of projects don't have patch at all :) | 15:03 |
| sambetts | shouldn't we just blow up with a Bad Requeset?? | 15:03 |
| vdrok | as for consistency | 15:03 |
| sambetts | Request | 15:03 |
| vdrok | sambetts: why, it is a valid patch | 15:03 |
| rloo | sambetts: yes, that's an alternative. | 15:03 |
| jroll | so, intuitively, I would think that each one is applied in order, and the result is the last | 15:04 |
| rloo | vdrok: the problem is that we haven't stated what ironic's behaviour is in that case. which is what we're discussing now :) | 15:04 |
| jroll | which seems to be what this is doing | 15:04 |
| rloo | jroll: right, that's what i thought intuitively too. which means that if each is applied and one of them is incorrect, we need to let the user know? | 15:04 |
| rloo | jroll: which is different from ignoring all except the last. | 15:05 |
| jroll | rloo: yeah, maybe, but does anyone (the user, ironic) care, as long as the result is correct? | 15:05 |
| rloo | jroll: i dunno. I wonder why the user even did what they did. | 15:05 |
| lucasagomes | sambetts, we need to check if the specification says anything about it | 15:06 |
| jroll | sure | 15:06 |
| jroll | yeah, I'd defer to jsonpatch spec first | 15:06 |
| lucasagomes | so far, by testing it locally with the python library it accepts 2 patches for the same endpoint | 15:06 |
| sambetts | yeah, just looking at the PATCH RFC | 15:06 |
| vdrok | sambetts: lucasagomes https://tools.ietf.org/html/rfc6902 | 15:06 |
| lucasagomes | and it applies the last one, so I think only checking the last one makes sense to me | 15:06 |
| vdrok | nothing about forbidding multiple operations on the same patch | 15:06 |
| vdrok | s/patch/path | 15:06 |
| lucasagomes | vdrok, yeah | 15:06 |
| jroll | well, the open question is, if the first is an invalid value, should it error or only apply the last | 15:07 |
| sambetts | "Each operation in the sequence is applied to | 15:07 |
| sambetts | the target document; the resulting document becomes the target of the | 15:07 |
| sambetts | next operation. " | 15:07 |
| sambetts | Evaluation continues until all operations are | 15:07 |
| sambetts | successfully applied or until an error condition is encountered. | 15:07 |
| jroll | aha | 15:07 |
| jroll | so yes, we should error if any in the chain are invalid | 15:07 |
| lucasagomes | touche :-) | 15:07 |
| rloo | thx sambetts ! | 15:07 |
| lucasagomes | ok so we need to check the whole chain | 15:07 |
| rloo | where should we document that for the user? | 15:08 |
| rloo | I mean, besides fixing the code :) | 15:08 |
| lucasagomes | rloo, perhaps just pointing them to the spec? | 15:08 |
| rloo | lucasagomes, vdrok: in the commit (and bug) it makes sense to mention the spec. | 15:09 |
| rloo | lucasagomes: I am thinking in our user documentation. not that we have much wrt the client. | 15:09 |
| rloo | lucasagomes, vdrok: on second thought, forget about external doc. for now anyway. maybe just add the link to spec, as a docstring in the code. | 15:10 |
| *** kbyrne has quit IRC | 15:10 | |
| vdrok | rloo: OK, will add that, thanks! | 15:10 |
| rloo | thx vdrok! | 15:10 |
| lucasagomes | rloo, right, in any case we should somewhere in our docs also mention that ironic does partial updates of resources based on that spec | 15:11 |
| lucasagomes | but can be outisde vdrok's patch | 15:11 |
| *** Goneri has quit IRC | 15:11 | |
| sambetts | Does the client ensure that the order of operations is maintained? I've never thought about it before | 15:11 |
| rloo | lucasagomes: yup. (especially since i can't think of where in our doc to put it). | 15:11 |
| rloo | sambetts: that's a good question. i would hope so but never checked. | 15:12 |
| lucasagomes | sambetts, would be good to check, I bet it does | 15:13 |
| sambetts | rloo: it looks like it does, args_array_to_patch, handles the args in order | 15:13 |
| rloo | sambetts: yah, we did something right! :D | 15:13 |
| lucasagomes | honestly the client is odd... because patches are complex they can add stuff and remove other stuff as part of the same patch | 15:14 |
| lucasagomes | currently it's not supported by the CLI | 15:14 |
| lucasagomes | we only do 1 operation per time | 15:14 |
| lucasagomes | add or remove or replace etc | 15:14 |
| lucasagomes | if something more complex is needed I think the library or API will need to be used | 15:14 |
| jroll | oh that's interesting, and I think it would (kind of) break here | 15:14 |
| jroll | add /foo, remove /foo, would only apply remove /foo and error | 15:15 |
| *** e0ne has quit IRC | 15:15 | |
| lucasagomes | jroll, right, if foo is mandatory yes | 15:15 |
| lucasagomes | but what's going to break here? | 15:16 |
| jroll | well, no | 15:16 |
| sambetts | I think what jroll means is the client will only read the last one of the args list | 15:16 |
| jroll | it would error because /foo isn't there to be removed | 15:16 |
| jroll | not the client | 15:16 |
| jroll | the api | 15:16 |
| jroll | er wow, ignore me | 15:16 |
| lucasagomes | well it will be there after the add | 15:16 |
| jroll | I was still thinking of the "only apply last" case | 15:16 |
| openstackgerrit | Merged openstack/ironic-webclient: Added capitalize filter https://review.openstack.org/300212 | 15:17 |
| jroll | I clearly need coffee | 15:17 |
| *** e0ne has joined #openstack-ironic | 15:17 | |
| lucasagomes | heh it's fine (/me needs coffee too) | 15:18 |
| openstackgerrit | Merged openstack/ironic-webclient: Renamed IronicNodeTransition to IronicNodeProvisionTransition https://review.openstack.org/300225 | 15:18 |
| *** raildo is now known as raildo-afk | 15:19 | |
| lucasagomes | JayF, quick q, I noticed that after cleaning we do a hard power off in the node (oob). If we do things like software raid and stuff that can be problematic | 15:19 |
| lucasagomes | JayF, you guys do a lot of operations as part of cleaning right? Have you ever bumped in some data loss problem due that? | 15:19 |
| lucasagomes | IMHO we should mimic what deploy does, sync() + oob power off OR in-band power off | 15:19 |
| jroll | lucasagomes: we do syncs and whatnot in our clean steps that do that sort of thing | 15:20 |
| lucasagomes | jroll, right on | 15:22 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Allow to set Neutron port setup delay from config https://review.openstack.org/293876 | 15:22 |
| jroll | lucasagomes: it really depends on the clean step if it should be unband or oob | 15:22 |
| jroll | s/unband/inband | 15:22 |
| lucasagomes | jroll, right. But at the end, the last one when we tear it down | 15:23 |
| lucasagomes | we just hard power off the node | 15:23 |
| jroll | lucasagomes: right, the last step may require an oob reboot | 15:23 |
| lucasagomes | jroll, because in realitly even sync is not 100% guaranteed that the data was actually flushed | 15:23 |
| jroll | e.g. firmware update | 15:23 |
| openstackgerrit | Merged openstack/ironic: Devstack: Change to use 'ovs-vsctl get port tag' https://review.openstack.org/307543 | 15:23 |
| jroll | right | 15:23 |
| lucasagomes | jroll, https://github.com/openstack/ironic/blob/master/ironic/drivers/modules/deploy_utils.py#L1053 | 15:25 |
| *** Goneri has joined #openstack-ironic | 15:26 | |
| jroll | lucasagomes: yes, I know | 15:27 |
| jlvillal | lucasagomes: Thanks for fixing the gate :) | 15:27 |
| jroll | I mean, maybe clean steps can indicate if they require hard reboot or not | 15:27 |
| lucasagomes | jlvillal, no problem :-) glad it's fixed now | 15:28 |
| jroll | yes ++ thank you | 15:28 |
| lucasagomes | jroll, yeah +1 to have such control on the clean steps | 15:29 |
| jroll | 'reboot_requested': 'hard' | 15:29 |
| *** e0ne has quit IRC | 15:29 | |
| jroll | or something | 15:29 |
| *** rama_y has joined #openstack-ironic | 15:29 | |
| devananda | g'morning! | 15:31 |
| sambetts | Hey devananda !\ | 15:31 |
| *** itamarl has quit IRC | 15:32 | |
| rloo | morning devananda | 15:32 |
| lucasagomes | jroll, yeah, sounds good! Will take a look into it | 15:33 |
| lucasagomes | jroll, actually one more thing, re software raid | 15:33 |
| *** sacharya has joined #openstack-ironic | 15:33 | |
| lucasagomes | jroll, we need to run it as part of the automated clean, but we could have the case where we need to have hardware RAID and software RAID depending on the node | 15:34 |
| lucasagomes | so we can't use a config file to enable/disable raid | 15:34 |
| lucasagomes | it should be per node | 15:34 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [database] https://review.openstack.org/301815 | 15:34 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [inspector] https://review.openstack.org/301955 | 15:34 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [irmc] https://review.openstack.org/302476 | 15:34 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [seamicro] https://review.openstack.org/303113 | 15:34 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [neutron] https://review.openstack.org/304838 | 15:34 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [ssh] https://review.openstack.org/303649 | 15:34 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [iboot] https://review.openstack.org/301856 | 15:34 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [cimc] https://review.openstack.org/301376 | 15:34 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [swift] https://review.openstack.org/303653 | 15:34 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [cisco_ucs] https://review.openstack.org/301382 | 15:34 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [dhcp] https://review.openstack.org/301830 | 15:34 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [ipmi] https://review.openstack.org/301967 | 15:34 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [console] https://review.openstack.org/295946 | 15:34 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [virtualbox] https://review.openstack.org/303655 | 15:34 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [oneview] https://review.openstack.org/303103 | 15:34 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [glance] https://review.openstack.org/304829 | 15:34 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [snmp] https://review.openstack.org/303644 | 15:34 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [ilo] https://review.openstack.org/301947 | 15:34 |
| * lucasagomes wow | 15:34 | |
| * TheJulia blinks | 15:34 | |
| sambetts | woah... | 15:34 |
| jroll | lol | 15:34 |
| jroll | rama_y: \o/ | 15:35 |
| jroll | lucasagomes: well, if it's out of tree I'd recommend putting something in node.extra | 15:35 |
| *** e0ne has joined #openstack-ironic | 15:36 | |
| lucasagomes | jroll, and in tree driver_info ? :-) | 15:36 |
| lucasagomes | cause software raid is something we want to have in the product AFACT | 15:36 |
| jroll | lucasagomes: yeah idk | 15:36 |
| jroll | I guess? | 15:36 |
| lucasagomes | + the driver composition will make it easy to swap raid interfaces, so my goal is to make it a 1st class citizen thing | 15:37 |
| lucasagomes | jroll, ack yeah | 15:37 |
| *** sacharya has quit IRC | 15:38 | |
| devananda | software RAID is not somthing I want, fwiw | 15:41 |
| rloo | lucasagomes: would this help? https://review.openstack.org/#/c/238858/ | 15:41 |
| devananda | I've had that design discussion several times in the past. it's too OS specific | 15:41 |
| jroll | I do agree there's a number of image/hardware related dragons/dragonslayers there | 15:42 |
| lucasagomes | rloo, not sure, cause raid is only enabled for manual clean AFAICT | 15:42 |
| sambetts | shouldn't software raid be built into the image being deployed? (whole disk image) I don't know how it would work for a partition image :/ | 15:42 |
| lucasagomes | which is fine if you only want hardware raid | 15:42 |
| jroll | it was actually incredibly hard to do in onmetal v2 | 15:42 |
| lucasagomes | devananda, yeah :-/ we do have a lot of requests to support it tho | 15:43 |
| devananda | lucasagomes: sure. and we can keep saying "no" :) | 15:43 |
| lucasagomes | heh right that's a way too | 15:43 |
| lucasagomes | tho, it's not a invalid request | 15:43 |
| devananda | Nova removed file injection because it's architecturally bad for Nova to be mucking with what's inside the user's image | 15:44 |
| rloo | "no" is great. makes things easier ;) | 15:44 |
| devananda | I feel pretty much the same way about this | 15:44 |
| jroll | it can be done without touching the image, if the image is built correctly | 15:44 |
| jroll | :) | 15:44 |
| devananda | yep | 15:44 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [DEFAULT] https://review.openstack.org/309070 | 15:45 |
| jroll | we do already have all the things to enable a deployer/vendor to do it | 15:45 |
| *** jayc_ has quit IRC | 15:45 | |
| lucasagomes | IPA should build it and write the image to the device | 15:45 |
| jroll | lucasagomes: yep, however there's things you need to do to the image to make it boot from the raid | 15:45 |
| jroll | ask natorious or JayF about it | 15:45 |
| jroll | they'll probably never touch an initramfs again | 15:46 |
| lucasagomes | right, like passing the right root device to it | 15:46 |
| jroll | it isn't that easy | 15:46 |
| jroll | it's actually OS-specific | 15:46 |
| jroll | you need to make it so that one of the boot stages can assemble the raid, forget which one | 15:46 |
| lucasagomes | right I have to read more about it, I iwll make sure I sync with JayF in austin about it | 15:46 |
| jroll | which isn't built in to most distros | 15:46 |
| jroll | I *think* one of the centoses works out of the box | 15:47 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [DEFAULT] https://review.openstack.org/309070 | 15:47 |
| *** sacharya has joined #openstack-ironic | 15:48 | |
| sambetts | jroll: could we provide a multi disk, whole disk image to allow someone to prebake an image with software raid?? | 15:48 |
| *** openstackgerrit has quit IRC | 15:48 | |
| lucasagomes | jroll, thanks! Yeah i def need to read more about it. Even if it's to say "no" for the feature I still need to get my arguments right there | 15:48 |
| jroll | sambetts: no clue | 15:48 |
| jroll | idk enough about building images to know if we could ship a DIB element that does this, for instance | 15:49 |
| *** openstackgerrit has joined #openstack-ironic | 15:49 | |
| jroll | I just know people smarter than me still have bruises from banging their head on their desk on this :) | 15:49 |
| thiagop | lol | 15:50 |
| *** sacharya has quit IRC | 15:52 | |
| vsaienko | morning jroll | 15:53 |
| *** trown is now known as trown|lunch | 15:53 | |
| jroll | heya vsaienko :) | 15:53 |
| vsaienko | jroll could you please have look on https://review.openstack.org/#/c/304473/ once have a time | 15:53 |
| vsaienko | it is a patch that adds job to test ironic cleaning at the gates | 15:53 |
| jroll | vsaienko: shred is way too slow to run in the gate | 15:54 |
| vsaienko | we can decrease disk size to 1Gb I think | 15:55 |
| jroll | vsaienko: I'll think about it, I'm about to jump in to meetings | 15:55 |
| *** jayc_ has joined #openstack-ironic | 15:59 | |
| *** baoli_ has quit IRC | 16:00 | |
| *** baoli has joined #openstack-ironic | 16:00 | |
| natorious | lucasagomes: to assemble the raid in initramfs via legacy boot, you have to make sure the proper modules are loaded in init and that udev created the right mapping for /dev/disk/by-label/root etc | 16:00 |
| natorious | so you can use a generic mdadm.conf that doesn't define the array | 16:01 |
| natorious | different distros might have different versions of mdadm and so the hooks might need a little adjustments but it should be pretty standard with hardware/software raid on legacy boot | 16:02 |
| natorious | lucasagomes: if your doing uefi then you'd need to do it all thru grub and can load the raid inside grub depending on your raid type to assemble and find your boot and root partitions on the raid | 16:03 |
| natorious | both seemed to work in our cases though you might have different results. Lmk if you need any help :) | 16:04 |
| openstackgerrit | Gonéri Le Bouder proposed openstack/ironic: iscsi: wipe the disk before deployment https://review.openstack.org/303605 | 16:05 |
| lucasagomes | natorious, oh thanks for that! I will, def ping you soon-ish about it | 16:06 |
| *** fhubik has quit IRC | 16:07 | |
| natorious | lucasgomes: all of our kickstart files are currently public and show how we're doing it too - ie https://github.com/rackerlabs/base-image-blueprints/blob/master/Fedora_23_Teeth.cfg | 16:07 |
| *** fhubik has joined #openstack-ironic | 16:08 | |
| natorious | any with a Teeth suffix is our madness | 16:08 |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic: Fix API node name updates https://review.openstack.org/300983 | 16:08 |
| natorious | *lucasagomes | 16:08 |
| lucasagomes | natorious, hah cool, I will take a look | 16:09 |
| natorious | one of these days I'll be allowed to do upstream work again :) | 16:09 |
| *** vishwanathj_zzz is now known as vishwanathj | 16:10 | |
| JayF | I mean, to be fair | 16:12 |
| JayF | 100% the hardest part of RAID in OMv2 was the images | 16:12 |
| *** karimb has quit IRC | 16:16 | |
| *** sacharya has joined #openstack-ironic | 16:19 | |
| sambetts | lucasagomes: https://review.openstack.org/#/c/308348 hopfully deals with the cause of the nova races | 16:19 |
| *** jistr has quit IRC | 16:21 | |
| JayF | jcook: ^ I bet you'd find that (https://review.openstack.org/#/c/308348) interesting, as it might change or make some of the cases better in your nova patch | 16:22 |
| lucasagomes | sambetts, cool I will take a look soon | 16:23 |
| lucasagomes | I will call it a day | 16:24 |
| lucasagomes | gotta catch the train back from the office | 16:24 |
| lucasagomes | have a great evening all | 16:24 |
| lucasagomes | sambetts, added to my list | 16:24 |
| *** ijw has joined #openstack-ironic | 16:24 | |
| sambetts | thanks lucasagomes | 16:25 |
| lucasagomes | see y'all | 16:25 |
| JayF | o/ | 16:25 |
| *** lucasagomes is now known as lucas-dinner | 16:26 | |
| * sambetts has the day off tomorrow so might not see people again until texas | 16:26 | |
| jlvillal | sambetts: I rarely see you :P | 16:27 |
| sambetts | hey jlvillal, you also might be interested in https://review.openstack.org/#/c/308348/2 | 16:28 |
| JayF | jlvillal: make any further progress on grenade shenanigans? | 16:28 |
| jlvillal | JayF: Working on it this morning :) | 16:29 |
| jlvillal | sambetts: Thanks. mrda and I will check it out | 16:29 |
| sambetts | jlvillal: awesome :D | 16:29 |
| *** ifarkas has quit IRC | 16:29 | |
| jlvillal | sambetts: But probably after the summit.... | 16:30 |
| sambetts | sure :-P | 16:30 |
| *** mbound has quit IRC | 16:35 | |
| *** baoli_ has joined #openstack-ironic | 16:38 | |
| *** [1]cdearborn has quit IRC | 16:39 | |
| *** baoli has quit IRC | 16:40 | |
| *** Guest14509 has quit IRC | 16:40 | |
| jlvillal | sambetts: Regarding that patch. I assume you will be pushing a new version, based on the -1 | 16:41 |
| dtantsur | morning devananda, jlvillal | 16:41 |
| jlvillal | Hi dtantsur :) | 16:41 |
| openstackgerrit | Thiago Paiva Brito proposed openstack/ironic: Add Dynamic Allocation feature for the OneView drivers https://review.openstack.org/286192 | 16:42 |
| sambetts | jlvillal: I'm not sure how to respond to the -1, they don't really indicate a way to update it | 16:44 |
| jlvillal | sambetts: Maybe you can ping Sylvain on IRC? | 16:45 |
| jlvillal | sambetts: I'm just assuming since they are a core in Nova, probably need to resolve the issue... | 16:45 |
| openstackgerrit | Merged openstack/ironic: Imported Translations from Zanata https://review.openstack.org/307575 | 16:45 |
| openstackgerrit | Merged openstack/ironic-webclient: Added IronicNodePowerTransition resource https://review.openstack.org/300226 | 16:46 |
| openstackgerrit | Merged openstack/ironic-webclient: Available node transitions now available in node_list https://review.openstack.org/300227 | 16:47 |
| openstackgerrit | Merged openstack/ironic-webclient: Basic node action hooks. https://review.openstack.org/281900 | 16:47 |
| sambetts | jlvillal: yeah I'm hoping for more feedback than just one person, because I've had it with a neutron patch where 1 core -1d but others where happy with it :/ | 16:47 |
| *** baoli_ has quit IRC | 16:48 | |
| jlvillal | Okay :) | 16:48 |
| *** baoli has joined #openstack-ironic | 16:48 | |
| *** ayoung has quit IRC | 16:48 | |
| sambetts | definatly needs a conversation though, that bug is just bumming me out because its the only thing thats blemishing my Ironic CIs perfect track record | 16:49 |
| jroll | sambetts: bauzas is definitely one of the experts on this thing | 16:49 |
| jroll | scheduling stuff in general | 16:49 |
| *** alexpilotti has joined #openstack-ironic | 16:49 | |
| sambetts | oh cool | 16:50 |
| sambetts | :) | 16:50 |
| *** ijw has quit IRC | 16:52 | |
| sambetts | you guys will have to point him out to me at the summit so I can chase him on it :-P | 16:53 |
| * sambetts is heading out now | 16:53 | |
| *** sambetts is now known as sambetts|afk | 16:54 | |
| *** Goneri has quit IRC | 16:55 | |
| *** baoli has quit IRC | 16:55 | |
| *** alexpilotti has quit IRC | 16:55 | |
| thiagop | night sambetts|afk | 16:57 |
| thiagop | guys, going earlier today. Looking forward to see you on summit | 16:58 |
| thiagop | til' there | 16:59 |
| TheJulia | thiagop: have good flights, and see you there | 16:59 |
| thiagop | you too TheJulia | 16:59 |
| thiagop | safe travels for everyone | 16:59 |
| *** thiagop has quit IRC | 16:59 | |
| *** zer0c00l_ has joined #openstack-ironic | 17:01 | |
| *** yonglihe has quit IRC | 17:01 | |
| *** absubram_ has joined #openstack-ironic | 17:01 | |
| *** zer0c00l has quit IRC | 17:01 | |
| *** trown|lunch is now known as trown | 17:02 | |
| *** jayc_ has quit IRC | 17:02 | |
| *** yarkot has quit IRC | 17:02 | |
| *** sinh has joined #openstack-ironic | 17:03 | |
| *** serverascode has quit IRC | 17:03 | |
| *** yonglihe has joined #openstack-ironic | 17:03 | |
| *** igordcar1 has joined #openstack-ironic | 17:03 | |
| *** keekz has quit IRC | 17:04 | |
| *** sinh_ has quit IRC | 17:04 | |
| *** igordcard has quit IRC | 17:04 | |
| *** igordcar1 has quit IRC | 17:04 | |
| *** absubram has quit IRC | 17:04 | |
| *** absubram_ is now known as absubram | 17:04 | |
| *** igordcard has joined #openstack-ironic | 17:04 | |
| *** garthb has quit IRC | 17:04 | |
| *** derekh has quit IRC | 17:05 | |
| *** serverascode has joined #openstack-ironic | 17:05 | |
| *** keekz has joined #openstack-ironic | 17:05 | |
| *** yarkot has joined #openstack-ironic | 17:05 | |
| *** stevemar has quit IRC | 17:06 | |
| *** gabriel-bezerra has quit IRC | 17:07 | |
| *** stevemar has joined #openstack-ironic | 17:08 | |
| *** gabriel-bezerra has joined #openstack-ironic | 17:08 | |
| *** ekarlso has quit IRC | 17:08 | |
| *** suro-patz has joined #openstack-ironic | 17:08 | |
| *** [1]cdearborn has joined #openstack-ironic | 17:09 | |
| *** yuikotak_ has joined #openstack-ironic | 17:09 | |
| openstackgerrit | Mathieu Mitchell proposed openstack/ironic: Allow configuring shred's final overwrite with zeros https://review.openstack.org/304101 | 17:10 |
| mat128 | JayF, NobodyCam, TheJulia, sambetts|afk: build is finally green :) https://review.openstack.org/#/c/304578/ | 17:11 |
| *** mkovacik__ has joined #openstack-ironic | 17:11 | |
| JayF | mat128: nit: s/ramdon iterations/agent.shred_random_overwrite_interations/ | 17:12 |
| JayF | mat128: if you don't mind and agree, then I'll put my vote back on 304101 | 17:12 |
| JayF | I will anyway if you meh at me too | 17:12 |
| JayF | lol | 17:12 |
| NobodyCam | mat128: w00t :-) | 17:13 |
| JayF | 304578 +2 | 17:13 |
| mat128 | JayF: Is that typo mine or yours? | 17:13 |
| JayF | mat128: I think mine | 17:14 |
| mat128 | yeah, found it | 17:14 |
| NobodyCam | And morning from the roads of phoenix AZ | 17:14 |
| JayF | mat128: just a peeve, I like the fully qualified option names so I don't wtf at the config file later | 17:14 |
| mat128 | about to push a new changeset | 17:14 |
| mat128 | yeah | 17:14 |
| *** vdrok_ has joined #openstack-ironic | 17:15 | |
| openstackgerrit | Mathieu Mitchell proposed openstack/ironic: Allow configuring shred's final overwrite with zeros https://review.openstack.org/304101 | 17:15 |
| mat128 | ^updated :) | 17:15 |
| *** jayc_ has joined #openstack-ironic | 17:15 | |
| *** pleia2_ has joined #openstack-ironic | 17:15 | |
| openstackgerrit | John L. Villalovos proposed openstack/ironic: WIP: Grenade testing patch https://review.openstack.org/309115 | 17:16 |
| JayF | I would love if some folks could find time to re-review https://review.openstack.org/#/c/263842/, it's been back and forth a few times, and I think it's in a landable state now. Test failures are leftover from gate shenanigans and recheck is already in progress. | 17:18 |
| *** jmccrory- has joined #openstack-ironic | 17:18 | |
| *** electrofelix1 has joined #openstack-ironic | 17:18 | |
| *** electrofelix has quit IRC | 17:18 | |
| *** aarefiev22 has quit IRC | 17:19 | |
| *** maurosr has quit IRC | 17:19 | |
| *** mkovacik_ has quit IRC | 17:19 | |
| *** kromanenko has quit IRC | 17:19 | |
| *** yuikotakada has quit IRC | 17:19 | |
| *** amit213 has quit IRC | 17:19 | |
| *** jmccrory has quit IRC | 17:19 | |
| *** pleia2 has quit IRC | 17:19 | |
| *** davideagnello has quit IRC | 17:19 | |
| *** sergek has quit IRC | 17:19 | |
| *** jasondotstar has quit IRC | 17:19 | |
| *** vdrok has quit IRC | 17:19 | |
| *** agordeev has quit IRC | 17:19 | |
| *** cppforlife_ has quit IRC | 17:19 | |
| *** electrofelix1 is now known as help | 17:19 | |
| openstackgerrit | Julia Kreger proposed openstack/ironic: Active Node Creation via adopt state https://review.openstack.org/275766 | 17:19 |
| *** jmccrory- is now known as jmccrory | 17:19 | |
| mat128 | JayF: I will get Martin to check your comment. He's new to reviews but I'm sure he will find a way to address your test failure | 17:19 |
| *** help is now known as electrofelix | 17:19 | |
| *** maurosr has joined #openstack-ironic | 17:19 | |
| JayF | mat128: well, I mean, I'm also not fully convinced putting something on the driver is a good idea either :) | 17:20 |
| JayF | mat128: it was just extra convienient that I got to have the easier "it dun work" argument | 17:20 |
| JayF | lol | 17:20 |
| *** vdrok_ is now known as vdrok | 17:20 | |
| *** pleia2_ is now known as pleia2 | 17:20 | |
| *** ekarlso has joined #openstack-ironic | 17:21 | |
| *** alexpilotti has joined #openstack-ironic | 17:22 | |
| *** ijw has joined #openstack-ironic | 17:22 | |
| *** amit213 has joined #openstack-ironic | 17:22 | |
| *** daemontool has quit IRC | 17:22 | |
| *** ChubYann has joined #openstack-ironic | 17:22 | |
| *** alexpilotti has quit IRC | 17:23 | |
| *** cppforlife_ has joined #openstack-ironic | 17:23 | |
| *** jasondotstar has joined #openstack-ironic | 17:24 | |
| *** ayoung has joined #openstack-ironic | 17:26 | |
| *** davideagnello has joined #openstack-ironic | 17:26 | |
| mat128 | JayF: Trying to reproduce locally with a driver member. I must say I prefer this to a global simply because it seems much cleaner. | 17:26 |
| *** ijw has quit IRC | 17:27 | |
| mat128 | The member could be called something more specific than "deprecation_seen", but it should work | 17:27 |
| JayF | Yeah, I mean if it's fine to put random attributes on the driver I agree it's cleaner | 17:27 |
| *** alex_xu has quit IRC | 17:27 | |
| JayF | just didn't know how "ok" that would be | 17:27 |
| JayF | and when I tested it I used my attr names :) | 17:27 |
| mat128 | :) | 17:27 |
| *** ijw has joined #openstack-ironic | 17:27 | |
| mat128 | Good thing this is Python, you won't have to add this attribute to DriverBase | 17:27 |
| *** sergek has joined #openstack-ironic | 17:27 | |
| *** penick has joined #openstack-ironic | 17:28 | |
| *** agordeev has joined #openstack-ironic | 17:29 | |
| dtantsur | see you tomorrow! | 17:30 |
| *** dtantsur is now known as dtantsur|afk | 17:30 | |
| *** alex_xu has joined #openstack-ironic | 17:31 | |
| * jroll -> lunch | 17:33 | |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [DEFAULT] https://review.openstack.org/309070 | 17:34 |
| TheJulia | rloo: w/r/t 302819, why not 'allow operators', given it is in the context of their fleet management and how IPA behaves with their fleet when cleaning? | 17:35 |
| *** ohamada has quit IRC | 17:35 | |
| rloo | TheJulia: cuz what distinguishes that for operators, vs all the other configs we add/change. | 17:35 |
| TheJulia | good point | 17:36 |
| rloo | TheJulia: just really a consistency issue. | 17:36 |
| rloo | TheJulia: everything we do is for operators :D | 17:36 |
| TheJulia | :) | 17:38 |
| *** mgould has quit IRC | 17:39 | |
| *** garthb has joined #openstack-ironic | 17:39 | |
| openstackgerrit | Julia Kreger proposed openstack/ironic: Add config option for ATA erase fallback in agent https://review.openstack.org/302819 | 17:46 |
| *** ayoung has quit IRC | 17:47 | |
| *** ayoung has joined #openstack-ironic | 17:47 | |
| *** Sukhdev has joined #openstack-ironic | 17:47 | |
| *** ijw has quit IRC | 17:48 | |
| *** Goneri has joined #openstack-ironic | 17:49 | |
| openstackgerrit | Julia Kreger proposed openstack/ironic: Add config option for ATA erase fallback in agent https://review.openstack.org/302819 | 17:56 |
| *** electrofelix has quit IRC | 17:57 | |
| *** baoli has joined #openstack-ironic | 17:58 | |
| *** ChrisAusten has joined #openstack-ironic | 18:02 | |
| *** jaybeale has joined #openstack-ironic | 18:06 | |
| *** jaybeale_ has joined #openstack-ironic | 18:08 | |
| *** jaybeale has quit IRC | 18:08 | |
| *** ijw has joined #openstack-ironic | 18:09 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic-python-agent: Updated from global requirements https://review.openstack.org/306943 | 18:12 |
| *** [1]cdearborn has quit IRC | 18:18 | |
| *** baoli has quit IRC | 18:19 | |
| * jlvillal wonders if anyone here is coming to Portland for PyCon | 18:20 | |
| jlvillal | May 28th - June 5th :) | 18:20 |
| *** ayoung has quit IRC | 18:21 | |
| *** ijw has quit IRC | 18:24 | |
| *** cdearborn has joined #openstack-ironic | 18:26 | |
| *** fragatin_ has quit IRC | 18:27 | |
| *** fragatina has joined #openstack-ironic | 18:27 | |
| *** fhubik has quit IRC | 18:30 | |
| *** ijw has joined #openstack-ironic | 18:30 | |
| *** Sukhdev has quit IRC | 18:30 | |
| *** ijw has quit IRC | 18:35 | |
| *** fragatina has quit IRC | 18:40 | |
| *** aarefiev22 has joined #openstack-ironic | 18:46 | |
| *** athomas has quit IRC | 18:49 | |
| openstackgerrit | Julia Kreger proposed openstack/ironic: Add config option for ATA erase fallback in agent https://review.openstack.org/302819 | 18:50 |
| *** cdearborn has quit IRC | 18:53 | |
| *** Sukhdev has joined #openstack-ironic | 18:55 | |
| *** rama_y has quit IRC | 18:55 | |
| *** cdearborn has joined #openstack-ironic | 18:57 | |
| *** ChrisAusten has quit IRC | 19:00 | |
| *** lennyb has quit IRC | 19:01 | |
| *** baoli has joined #openstack-ironic | 19:02 | |
| *** amotoki has quit IRC | 19:05 | |
| *** ijw has joined #openstack-ironic | 19:08 | |
| *** baoli has quit IRC | 19:08 | |
| *** ijw has quit IRC | 19:13 | |
| *** ijw has joined #openstack-ironic | 19:13 | |
| *** rcernin has quit IRC | 19:19 | |
| *** mbound has joined #openstack-ironic | 19:23 | |
| *** fragatina has joined #openstack-ironic | 19:28 | |
| *** jaybeale_ has quit IRC | 19:30 | |
| *** jaybeale has joined #openstack-ironic | 19:31 | |
| *** absubram has quit IRC | 19:32 | |
| *** ppiela has quit IRC | 19:39 | |
| *** ppiela has joined #openstack-ironic | 19:39 | |
| *** baoli has joined #openstack-ironic | 19:43 | |
| *** ppiela_ has joined #openstack-ironic | 19:44 | |
| *** ppiela has quit IRC | 19:46 | |
| *** jaybeale has quit IRC | 19:50 | |
| *** jaybeale has joined #openstack-ironic | 19:51 | |
| *** Goneri has quit IRC | 19:53 | |
| *** ayoung has joined #openstack-ironic | 19:55 | |
| *** jaybeale has quit IRC | 19:55 | |
| *** jayc_ has quit IRC | 20:05 | |
| *** Sukhdev has quit IRC | 20:09 | |
| *** baoli_ has joined #openstack-ironic | 20:09 | |
| *** baoli___ has joined #openstack-ironic | 20:12 | |
| *** baoli has quit IRC | 20:12 | |
| *** rcernin has joined #openstack-ironic | 20:12 | |
| *** baoli_ has quit IRC | 20:14 | |
| *** ijw has quit IRC | 20:15 | |
| *** jayc_ has joined #openstack-ironic | 20:19 | |
| *** baoli___ has quit IRC | 20:20 | |
| *** baoli has joined #openstack-ironic | 20:21 | |
| *** thrash is now known as thrash|g0ne | 20:21 | |
| *** e0ne has quit IRC | 20:24 | |
| *** e0ne has joined #openstack-ironic | 20:24 | |
| *** ijw has joined #openstack-ironic | 20:32 | |
| *** jjohnson2 has joined #openstack-ironic | 20:34 | |
| jjohnson2 | I finally am going to submit a talk for a conference, hurray | 20:35 |
| *** rloo has quit IRC | 20:35 | |
| *** rama_y has joined #openstack-ironic | 20:35 | |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Remove two DEPRECATED config options from [agent] https://review.openstack.org/307456 | 20:36 |
| *** piet has joined #openstack-ironic | 20:38 | |
| *** alexpilo_ has quit IRC | 20:38 | |
| *** alexpilotti has joined #openstack-ironic | 20:38 | |
| *** trown is now known as trown|outtypewww | 20:39 | |
| *** jaybeale has joined #openstack-ironic | 20:43 | |
| *** alexpilotti has quit IRC | 20:44 | |
| *** ijw has quit IRC | 20:46 | |
| *** ijw has joined #openstack-ironic | 20:48 | |
| *** krotscheck has quit IRC | 20:48 | |
| *** krotscheck has joined #openstack-ironic | 20:48 | |
| *** sacharya has quit IRC | 20:51 | |
| *** rloo has joined #openstack-ironic | 20:52 | |
| *** ijw has quit IRC | 20:53 | |
| *** sacharya has joined #openstack-ironic | 20:53 | |
| *** suro-patz has quit IRC | 21:01 | |
| *** suro-patz has joined #openstack-ironic | 21:01 | |
| *** suro-patz has quit IRC | 21:05 | |
| *** ijw has joined #openstack-ironic | 21:06 | |
| *** e0ne has quit IRC | 21:07 | |
| *** e0ne has joined #openstack-ironic | 21:08 | |
| *** jjohnson2 has quit IRC | 21:08 | |
| *** Sukhdev has joined #openstack-ironic | 21:15 | |
| *** lennyb has joined #openstack-ironic | 21:22 | |
| *** ijw has quit IRC | 21:23 | |
| *** suro-patz has joined #openstack-ironic | 21:24 | |
| openstackgerrit | Kyrylo Romanenko proposed openstack/python-ironicclient: Use name randomizer from tempest https://review.openstack.org/307904 | 21:26 |
| *** jtomasek has quit IRC | 21:27 | |
| *** ijw has joined #openstack-ironic | 21:28 | |
| *** suro-patz has quit IRC | 21:32 | |
| *** suro-patz has joined #openstack-ironic | 21:32 | |
| *** ijw has quit IRC | 21:33 | |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [api] https://review.openstack.org/309186 | 21:33 |
| *** suro-patz has quit IRC | 21:34 | |
| *** praneshp has joined #openstack-ironic | 21:37 | |
| *** piet has quit IRC | 21:38 | |
| *** rcernin has quit IRC | 21:39 | |
| *** ijw has joined #openstack-ironic | 21:44 | |
| *** ijw has quit IRC | 21:50 | |
| *** ijw has joined #openstack-ironic | 21:51 | |
| openstackgerrit | Merged openstack/ironic: Update ironic.config.sample https://review.openstack.org/308951 | 21:53 |
| *** suro-patz has joined #openstack-ironic | 21:54 | |
| *** [1]cdearborn has joined #openstack-ironic | 21:58 | |
| *** Sukhdev has quit IRC | 22:03 | |
| *** Sukhdev has joined #openstack-ironic | 22:04 | |
| *** cdearborn has quit IRC | 22:09 | |
| *** alexpilotti has joined #openstack-ironic | 22:09 | |
| *** Sukhdev has quit IRC | 22:09 | |
| *** mbound has quit IRC | 22:14 | |
| *** Sukhdev has joined #openstack-ironic | 22:15 | |
| *** baoli has quit IRC | 22:20 | |
| *** ijw has quit IRC | 22:21 | |
| *** suro-patz has quit IRC | 22:25 | |
| devananda | JayF: the etherpad for 'making ops less worse' looks a bit light | 22:26 |
| devananda | JayF: is the goal just to gather feedback? | 22:26 |
| *** Haomeng has quit IRC | 22:27 | |
| JayF | devananda: mostly to gather feedback, I intend on getting a few specific examples into the etherpad | 22:28 |
| *** alexpilotti has quit IRC | 22:28 | |
| JayF | devananda: and if feedback runs dry, we can chat about how to solve those cases | 22:28 |
| devananda | k | 22:30 |
| *** krotscheck is now known as krotscheck_dcm | 22:32 | |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [deploy] https://review.openstack.org/309206 | 22:34 |
| *** ametts has quit IRC | 22:38 | |
| *** rloo has quit IRC | 22:40 | |
| mrda | Morning Ironic | 22:42 |
| *** alexpilotti has joined #openstack-ironic | 22:45 | |
| cinerama | hi mrda how goes it! | 22:45 |
| mrda | hey cinerama | 22:46 |
| * devananda finishes reviewing summit etherpads | 22:51 | |
| *** moshele has joined #openstack-ironic | 22:59 | |
| *** Sukhdev has quit IRC | 22:59 | |
| *** rbudden has quit IRC | 23:02 | |
| *** Sukhdev has joined #openstack-ironic | 23:04 | |
| *** moshele has quit IRC | 23:06 | |
| openstackgerrit | Mario Villaplana proposed openstack/ironic: Add notifications to Ironic https://review.openstack.org/298461 | 23:14 |
| *** rloo has joined #openstack-ironic | 23:17 | |
| openstackgerrit | Mario Villaplana proposed openstack/ironic: Add notifications to Ironic https://review.openstack.org/298461 | 23:21 |
| *** moshele has joined #openstack-ironic | 23:22 | |
| *** alexpilotti has quit IRC | 23:22 | |
| *** rloo has quit IRC | 23:25 | |
| *** jjohnson2 has joined #openstack-ironic | 23:34 | |
| *** penick has quit IRC | 23:39 | |
| *** Sukhdev has quit IRC | 23:42 | |
| *** e0ne has quit IRC | 23:47 | |
| *** [1]cdearborn has quit IRC | 23:48 | |
| *** Goneri has joined #openstack-ironic | 23:49 | |
| *** jayc_ has quit IRC | 23:51 | |
| *** ozialien has joined #openstack-ironic | 23:52 | |
| *** jjohnson2 has quit IRC | 23:56 | |
| *** penick has joined #openstack-ironic | 23:58 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!