| *** hw_wutianwei has joined #openstack-ironic | 00:02 | |
| *** bfournie has joined #openstack-ironic | 00:53 | |
| *** phuongnh has joined #openstack-ironic | 00:59 | |
| *** vinsh has quit IRC | 01:02 | |
| *** jkilpatr has quit IRC | 01:04 | |
| *** [1]cdearborn has quit IRC | 01:05 | |
| *** hoangcx has joined #openstack-ironic | 01:06 | |
| *** rloo has quit IRC | 01:06 | |
| *** tuanluong has joined #openstack-ironic | 01:19 | |
| *** maratoid_ has quit IRC | 01:24 | |
| *** jerrygb_ has joined #openstack-ironic | 01:30 | |
| *** vinsh has joined #openstack-ironic | 01:30 | |
| *** jerrygb has quit IRC | 01:31 | |
| *** jerrygb has joined #openstack-ironic | 01:45 | |
| *** jerrygb_ has quit IRC | 01:45 | |
| *** zhangyufei has joined #openstack-ironic | 01:51 | |
| *** jerrygb has quit IRC | 01:59 | |
| *** jerrygb has joined #openstack-ironic | 01:59 | |
| *** shobhar has quit IRC | 02:00 | |
| *** zhangyufei has quit IRC | 02:01 | |
| *** Sukhdev has quit IRC | 02:15 | |
| *** jerrygb has quit IRC | 02:17 | |
| *** jralph_ has quit IRC | 02:17 | |
| *** jerrygb has joined #openstack-ironic | 02:18 | |
| *** shobhar has joined #openstack-ironic | 02:21 | |
| *** jerrygb has quit IRC | 02:22 | |
| *** jralph has joined #openstack-ironic | 02:23 | |
| *** chihhsin has quit IRC | 02:29 | |
| *** vinsh has quit IRC | 02:51 | |
| *** awiddersheim has quit IRC | 02:51 | |
| *** awiddersheim has joined #openstack-ironic | 02:53 | |
| *** rbudden has joined #openstack-ironic | 02:58 | |
| *** jerrygb has joined #openstack-ironic | 02:59 | |
| *** vinsh has joined #openstack-ironic | 03:04 | |
| *** sabeen has joined #openstack-ironic | 03:11 | |
| *** shobhar has quit IRC | 03:15 | |
| *** baoli has quit IRC | 03:17 | |
| *** rbudden has quit IRC | 03:18 | |
| *** vikrant has joined #openstack-ironic | 03:25 | |
| *** g3ek has quit IRC | 03:27 | |
| *** g3ek has joined #openstack-ironic | 03:29 | |
| *** shobhar has joined #openstack-ironic | 03:46 | |
| *** shobhar__ has joined #openstack-ironic | 03:49 | |
| *** shobhar has quit IRC | 03:49 | |
| *** jerrygb has quit IRC | 03:59 | |
| *** jerrygb has joined #openstack-ironic | 04:01 | |
| *** Nisha has joined #openstack-ironic | 04:01 | |
| *** links has joined #openstack-ironic | 04:01 | |
| *** Nisha is now known as Guest10859 | 04:01 | |
| *** sabeen has quit IRC | 04:01 | |
| *** sabeen has joined #openstack-ironic | 04:05 | |
| *** yogi has joined #openstack-ironic | 04:18 | |
| *** syed_ has joined #openstack-ironic | 04:22 | |
| *** absubram has joined #openstack-ironic | 04:23 | |
| *** absubram_ has joined #openstack-ironic | 04:24 | |
| *** Guest10859 is now known as Nisha1 | 04:26 | |
| *** vinsh has quit IRC | 04:27 | |
| *** jerrygb_ has joined #openstack-ironic | 04:28 | |
| *** absubram has quit IRC | 04:28 | |
| *** absubram_ is now known as absubram | 04:28 | |
| *** jerrygb has quit IRC | 04:29 | |
| *** shobhar__ has quit IRC | 04:34 | |
| *** jerrygb has joined #openstack-ironic | 04:44 | |
| *** jerrygb_ has quit IRC | 04:45 | |
| *** absubram has quit IRC | 04:46 | |
| *** jerrygb_ has joined #openstack-ironic | 04:59 | |
| *** jerrygb has quit IRC | 04:59 | |
| *** rama_y has quit IRC | 04:59 | |
| *** rama_y has joined #openstack-ironic | 05:00 | |
| *** sabeen has quit IRC | 05:11 | |
| *** jerrygb_ has quit IRC | 05:15 | |
| *** jerrygb has joined #openstack-ironic | 05:15 | |
| *** jerrygb has quit IRC | 05:33 | |
| *** jerrygb has joined #openstack-ironic | 05:33 | |
| *** jerrygb has quit IRC | 05:33 | |
| *** openstackgerrit has quit IRC | 05:48 | |
| *** openstackgerrit has joined #openstack-ironic | 05:48 | |
| *** shuquan has joined #openstack-ironic | 05:50 | |
| *** shuquan has quit IRC | 06:05 | |
| *** shuquan has joined #openstack-ironic | 06:07 | |
| *** vinsh has joined #openstack-ironic | 06:27 | |
| *** vinsh has quit IRC | 06:33 | |
| *** mgoddard has quit IRC | 06:40 | |
| *** yolanda has joined #openstack-ironic | 06:51 | |
| *** yolanda_ has joined #openstack-ironic | 06:51 | |
| *** yolanda_ has quit IRC | 06:52 | |
| *** rbartal has joined #openstack-ironic | 06:55 | |
| *** fxpester has joined #openstack-ironic | 06:58 | |
| *** jtomasek has joined #openstack-ironic | 07:01 | |
| *** syed_ has quit IRC | 07:05 | |
| *** Nisha1 has quit IRC | 07:07 | |
| *** Nisha1 has joined #openstack-ironic | 07:07 | |
| *** pester has joined #openstack-ironic | 07:09 | |
| *** fxpester has quit IRC | 07:12 | |
| *** shuquan has quit IRC | 07:14 | |
| *** tesseract has joined #openstack-ironic | 07:14 | |
| *** tesseract is now known as Guest25209 | 07:14 | |
| *** milan has joined #openstack-ironic | 07:28 | |
| *** jaosorior has joined #openstack-ironic | 07:28 | |
| milan | morning Ironic! :) | 07:29 |
|---|---|---|
| *** shuquan has joined #openstack-ironic | 07:30 | |
| *** jerrygb has joined #openstack-ironic | 07:34 | |
| *** rcernin has joined #openstack-ironic | 07:34 | |
| *** jerrygb has quit IRC | 07:39 | |
| *** jtomasek_ has joined #openstack-ironic | 07:41 | |
| *** jtomasek has quit IRC | 07:41 | |
| *** moshele has joined #openstack-ironic | 07:42 | |
| *** milan has quit IRC | 07:43 | |
| *** openstackgerrit has quit IRC | 07:48 | |
| *** shuquan has quit IRC | 07:48 | |
| *** openstackgerrit has joined #openstack-ironic | 07:48 | |
| *** g3ek has quit IRC | 07:52 | |
| *** fragatina has joined #openstack-ironic | 07:53 | |
| *** fragatina has quit IRC | 07:53 | |
| *** fragatina has joined #openstack-ironic | 07:53 | |
| *** ccamacho has joined #openstack-ironic | 07:54 | |
| *** shuquan has joined #openstack-ironic | 07:58 | |
| *** zzzeek has quit IRC | 08:00 | |
| *** zzzeek has joined #openstack-ironic | 08:00 | |
| *** g3ek has joined #openstack-ironic | 08:00 | |
| *** penick has joined #openstack-ironic | 08:01 | |
| *** penick has quit IRC | 08:04 | |
| *** milan has joined #openstack-ironic | 08:05 | |
| *** Nisha1 has quit IRC | 08:17 | |
| *** vsaienk0 has quit IRC | 08:18 | |
| milan | aarefiev, morning! I'm staring into the discovery logs, wondering what's going on... do you already have some theory? | 08:21 |
| *** shuquan has quit IRC | 08:21 | |
| *** athomas has joined #openstack-ironic | 08:22 | |
| *** shuquan has joined #openstack-ironic | 08:23 | |
| *** gcb has quit IRC | 08:27 | |
| *** gcb has joined #openstack-ironic | 08:28 | |
| *** vsaienk0 has joined #openstack-ironic | 08:29 | |
| *** ralonsoh has joined #openstack-ironic | 08:37 | |
| *** Nisha1 has joined #openstack-ironic | 08:37 | |
| *** ralonsoh has quit IRC | 08:38 | |
| *** ralonsoh has joined #openstack-ironic | 08:38 | |
| *** rama_y has quit IRC | 08:41 | |
| *** rama_y has joined #openstack-ironic | 08:41 | |
| *** ohamada has joined #openstack-ironic | 08:51 | |
| *** jpich has joined #openstack-ironic | 08:54 | |
| *** aparnav has joined #openstack-ironic | 08:56 | |
| *** shuquan has quit IRC | 08:58 | |
| *** jtomasek_ is now known as jtomasek | 09:02 | |
| *** _milan_ has joined #openstack-ironic | 09:04 | |
| *** milan has quit IRC | 09:04 | |
| *** amotoki has quit IRC | 09:04 | |
| aarefiev | morning! | 09:05 |
| *** mjura_ has joined #openstack-ironic | 09:06 | |
| aarefiev | milan: hey, yeah I have | 09:06 |
| *** shuquan has joined #openstack-ironic | 09:06 | |
| aarefiev | our infra became pretty fast | 09:06 |
| _milan_ | aarefiev, hey | 09:07 |
| _milan_ | yup | 09:07 |
| _milan_ | so seems the ramdisk is faster than the sync :-/ | 09:07 |
| *** ChubYann has quit IRC | 09:07 | |
| _milan_ | aarefiev, ^ correct? | 09:07 |
| aarefiev | yep | 09:07 |
| * _milan_ got disconnected a bit here | 09:07 | |
| aarefiev | https://review.openstack.org/#/c/392943/ | 09:07 |
| patchbot | patch 392943 - ironic-inspector - Allow to configure cache clean up period | 09:07 |
| _milan_ | aarefiev, ack | 09:08 |
| aarefiev | propose new option | 09:08 |
| * _milan_ would +2 it already | 09:08 | |
| _milan_ | but wanted to test in devstack | 09:08 |
| _milan_ | which is failing for me | 09:08 |
| _milan_ | *local devstack :-/ | 09:08 |
| * _milan_ 's OS fault | 09:08 | |
| aarefiev | we need project config patch is merged first | 09:08 |
| _milan_ | aarefiev, patch link handy? | 09:09 |
| aarefiev | I think about 15 sec period | 09:09 |
| _milan_ | aarefiev, I'd make it less | 09:09 |
| aarefiev | patch is on way | 09:09 |
| _milan_ | the diff is 30s | 09:09 |
| *** chihhsin has joined #openstack-ironic | 09:09 | |
| aarefiev | there was default 60 sec | 09:09 |
| aarefiev | and it works | 09:09 |
| aarefiev | worked | 09:09 |
| aarefiev | heh | 09:09 |
| aarefiev | maybe 10 sec | 09:09 |
| _milan_ | http://logs.openstack.org/43/392943/4/check/gate-tempest-dsvm-ironic-inspector-discovery/3705288/logs/screen-ironic-inspector.txt.gz#_2016-11-02_21_17_18_442 | 09:10 |
| _milan_ | not even 30 in my case | 09:10 |
| _milan_ | not even 20 :D | 09:10 |
| _milan_ | hell | 09:10 |
| _milan_ | 16 if I'm counting OK? | 09:10 |
| _milan_ | I'd vote for a sync of 5s :D | 09:10 |
| *** gcb has quit IRC | 09:11 | |
| _milan_ | aarefiev, ^ | 09:11 |
| aarefiev | I'm fine with 5 sec too | 09:11 |
| _milan_ | ack | 09:11 |
| aarefiev | doesn't matter here, it's hacky test | 09:12 |
| _milan_ | yeah | 09:12 |
| *** gcb has joined #openstack-ironic | 09:13 | |
| * _milan_ worried about our logic though --- if it's missing in ironic but present in inspector we blow it up :-/ | 09:13 | |
| _milan_ | not 100% sure it's correct | 09:13 |
| _milan_ | aarefiev, went? | 09:13 |
| _milan_ | *wdyt | 09:13 |
| aarefiev | it was in ironic and we deleted it | 09:14 |
| aarefiev | it found it first and then it's missing | 09:14 |
| aarefiev | many it | 09:14 |
| aarefiev | inspector | 09:14 |
| _milan_ | aarefiev, so test deletes the node from Ironic, correct? | 09:15 |
| aarefiev | yeah | 09:15 |
| _milan_ | but the sync doesn't propagate the change to inspector in time, correct? | 09:15 |
| aarefiev | it is before yesterday :) | 09:16 |
| _milan_ | yeah :) so inspector still tracks stale record that is missing in ironic already when the introspection data arrives, correct? | 09:16 |
| *** portdirect_away is now known as portdirect | 09:19 | |
| aarefiev | so inspector receives introspected data, it's looking for node, found it | 09:19 |
| aarefiev | then sync delete node from cache | 09:19 |
| * _milan_ checks logs | 09:19 | |
| aarefiev | and inspector has split brain | 09:20 |
| _milan_ | aarefiev, the sync didn't even run in my case http://logs.openstack.org/43/392943/4/check/gate-tempest-dsvm-ironic-inspector-discovery/3705288/logs/screen-ironic-inspector.txt.gz#_2016-11-02_21_17_34_332 | 09:21 |
| _milan_ | sync in my case is 15 sec later from ^ | 09:22 |
| *** lucas-afk is now known as lucasagomes | 09:22 | |
| lucasagomes | morning all | 09:22 |
| _milan_ | morning lucasagomes! :) | 09:22 |
| aarefiev | morning lucasagomes | 09:22 |
| aarefiev | _milan_: yeah, so whats here | 09:23 |
| aarefiev | inspector found node http://logs.openstack.org/43/392943/4/check/gate-tempest-dsvm-ironic-inspector-discovery/3705288/logs/screen-ironic-inspector.txt.gz#_2016-11-02_21_17_34_318 | 09:23 |
| aarefiev | before error | 09:23 |
| _milan_ | aarefiev, so in my case it seems 1) node deleted in Ironic 2) data arrives to inspector 3) inspector finds the node in cache 4) inspector fails to match cache with ironic and blows up the introspection | 09:23 |
| _milan_ | * discovery | 09:24 |
| aarefiev | so, yeah, same races | 09:24 |
| _milan_ | yup race | 09:24 |
| *** xek has quit IRC | 09:24 | |
| * _milan_ worried that it is a bug actually | 09:25 | |
| _milan_ | aarefiev, do you think we should blow up in case of the inconsistency if we're discovering? | 09:25 |
| _milan_ | shouldn't inspector push the data to ironic? | 09:25 |
| *** roeyc has joined #openstack-ironic | 09:26 | |
| aarefiev | _milan_: it's how our test works, it was expected | 09:26 |
| aarefiev | that it will fail one day | 09:26 |
| * _milan_ reminds himself of the IT joke: there are just 2 big challenges: naming, cache consistency and off by 1 errors ;) | 09:26 | |
| aarefiev | yeah :) | 09:27 |
| * _milan_ checks processing.py | 09:27 | |
| aarefiev | if we will add some notifications to ironic, we can notify inspector about deleted node and don't rely on sync | 09:29 |
| _milan_ | aarefiev, so in my case, since https://github.com/openstack/ironic-inspector/blob/master/ironic_inspector/process.py#L96 isn't manifested in the log, the plugin isn't triggered at all, yeah, not sure what to do besides bumping up the sync frequency... not a bug | 09:29 |
| _milan_ | aarefiev, notification +1 | 09:29 |
| _milan_ | but that's racy anyway as there might be other issues :D | 09:30 |
| aarefiev | yeah, we are running inspection on node and then delete it and assume it's new node | 09:30 |
| *** dtantsur|afk is now known as dtantsur | 09:30 | |
| dtantsur | Morning Ironic | 09:30 |
| aarefiev | morning dtantsur | 09:30 |
| _milan_ | morning dtantsur! :) | 09:30 |
| *** derekh has joined #openstack-ironic | 09:31 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic: API: lookup() ignore malformed MAC addresses https://review.openstack.org/393188 | 09:33 |
| dtantsur | lucasagomes, cherry-picked ^^^ | 09:33 |
| *** jerrygb has joined #openstack-ironic | 09:35 | |
| lucasagomes | dtantsur, thanks | 09:36 |
| *** jerrygb has quit IRC | 09:41 | |
| *** jaosorior is now known as jaosorior_lunch | 09:41 | |
| pas-ha | morning Ironic | 09:43 |
| aarefiev | morning pas-ha | 09:44 |
| *** jpich has quit IRC | 09:45 | |
| aarefiev | _milan_ : https://review.openstack.org/#/c/393193/1 | 09:45 |
| patchbot | patch 393193 - openstack-infra/project-config - Decrease cache sync time in inspector discovery job | 09:45 |
| aarefiev | if dtantsur agree ^ | 09:45 |
| dtantsur | aarefiev, I can't find the variable you're referencing in this patch | 09:46 |
| aarefiev | dtantsur: sorry, it's new https://review.openstack.org/#/c/392943/4 | 09:47 |
| patchbot | patch 392943 - ironic-inspector - Allow to configure cache clean up period | 09:47 |
| * _milan_ checks | 09:47 | |
| dtantsur | aarefiev, ok, please use depends-on to indicate relationship | 09:47 |
| dtantsur | oh, I see, there's chicken-and-egg problem here... | 09:48 |
| aarefiev | yeah | 09:48 |
| _milan_ | dtantsur, exactly :D | 09:48 |
| aarefiev | we need infra patch first | 09:48 |
| *** jpich has joined #openstack-ironic | 09:49 | |
| aarefiev | what happen with infra | 09:50 |
| * _milan_ +1 | 09:50 | |
| aarefiev | why it's so fast | 09:50 |
| _milan_ | yeah, good question :D | 09:51 |
| pas-ha | kvm? | 09:51 |
| _milan_ | dtantsur, fyi in my case introspection took 16s :D | 09:51 |
| aarefiev | ha, was that patch merged? | 09:51 |
| * _milan_ didn't follow kvm stuff for infra | 09:51 | |
| pas-ha | recently merged to ironic to use nested kvm where available | 09:52 |
| pas-ha | afaik yes | 09:52 |
| _milan_ | and hi pas-ha! :) | 09:52 |
| aarefiev | I see now :) | 09:52 |
| pas-ha | hi _milan_ | 09:52 |
| *** shuquan has quit IRC | 09:52 | |
| _milan_ | oh ya! nested kvm \o/ :D | 09:52 |
| pas-ha | but 16s is still suspiciously low as to me | 09:52 |
| *** shuquan has joined #openstack-ironic | 09:53 | |
| _milan_ | pas-ha, I'm fine with that, otherwise we'd have to have a sync of 1s for our discovery test to run | 09:57 |
| pas-ha | otoh, just rebuilt devstack, and virtual bm node just booted super-fast :-O | 09:58 |
| *** jerrygb has joined #openstack-ironic | 09:58 | |
| *** shuquan has quit IRC | 09:59 | |
| *** jerrygb has quit IRC | 10:03 | |
| *** shuquan has joined #openstack-ironic | 10:05 | |
| *** [1]cdearborn has joined #openstack-ironic | 10:13 | |
| *** mgoddard has joined #openstack-ironic | 10:14 | |
| TheJulia | Good morning everyone | 10:17 |
| lucasagomes | TheJulia, hi there | 10:18 |
| _milan_ | morning TheJulia! :) | 10:18 |
| _milan_ | TheJulia, what time is it for you atm? | 10:18 |
| TheJulia | 6:18 AM | 10:19 |
| _milan_ | wow | 10:19 |
| _milan_ | 11:19 here and I could sleep ;) | 10:19 |
| TheJulia | I've actually been up for about two hours, remnants of stackplague :( | 10:20 |
| *** hoangcx has quit IRC | 10:20 | |
| dtantsur | morning pas-ha, aarefiev, TheJulia | 10:20 |
| lucasagomes | oh, that sucks :-/ | 10:20 |
| _milan_ | oh, sorry TheJulia | 10:20 |
| dtantsur | yeah... | 10:21 |
| TheJulia | But I actually feel moderately human so far today | 10:21 |
| dtantsur | cool! I felt better yesterday :( | 10:22 |
| _milan_ | TheJulia, get better soon! | 10:22 |
| * _milan_ wonders what the stackpleague might actually be, seems it's not flu, that would be faster | 10:23 | |
| sambetts|afk | Morning all | 10:23 |
| *** sambetts|afk is now known as sambetts | 10:23 | |
| _milan_ | morning sambetts! :) | 10:23 |
| sambetts | hey milan | 10:23 |
| *** _milan_ is now known as milan | 10:23 | |
| TheJulia | _milan_: many people I know who got sick had gotten their yearly flu vaccine in the last couple months. I suspect it is just a run of the mill headcold that is able to take hold due to stress and the airplane travel. | 10:24 |
| *** shuquan has quit IRC | 10:24 | |
| *** roeyc has quit IRC | 10:25 | |
| openstackgerrit | Merged openstack/bifrost: Enable upper requirements on bifrost https://review.openstack.org/391828 | 10:25 |
| * milan wishes there was a vaccine for that | 10:25 | |
| * milan opens http://www.commoncold.org/understand.htm | 10:26 | |
| dtantsur | morning sambetts | 10:27 |
| *** Nisha1 has quit IRC | 10:27 | |
| dtantsur | TheJulia++ that's what I assume too | 10:27 |
| sambetts | o/ dtantsur, TheJulia | 10:27 |
| *** mgould|afk is now known as mgould | 10:29 | |
| TheJulia | tl;dr I think we need to move gatherings to nice tropical beaches where we can all have extended stays | 10:29 |
| *** vinsh has joined #openstack-ironic | 10:29 | |
| milan | TheJulia, +1000 | 10:29 |
| * milan loves pina colada :D | 10:29 | |
| openstackgerrit | Galyna Zholtkevych proposed openstack/ironic: ETAG supporting to enhance API evolution https://review.openstack.org/392213 | 10:29 |
| mgould | TheJulia: AIUI flu vaccines only protect against whichever strains are predicted to be nastiest that winter, but there are lots of others | 10:30 |
| TheJulia | I would <3 a pina colada right now | 10:30 |
| mgould | but yeah, probably just a nasty cold :-( | 10:30 |
| TheJulia | mgould: that is my understanding as well | 10:30 |
| *** moshele has quit IRC | 10:32 | |
| *** vinsh has quit IRC | 10:34 | |
| sambetts | I think its what we call in the UK "Freshers Flu" which is the cold/flu/illness that everyone gets during the first week of university because your suddenly thrown into a pool of people from all over the place with all their alien germs | 10:34 |
| lucasagomes | dtantsur, quick q: In project config those excludes do not accept a regex/wildcard right ? Like, we have agent_ssh and pxe_ssh I think we probably need to exclude both from xenial. It would be nice to be able to do "driver: '*_ssh;" | 10:36 |
| * lucasagomes might be asking too much | 10:36 | |
| * milan freaks out: alien germs taking over this planet ;) | 10:36 | |
| dtantsur | lucasagomes, I think they don't | 10:36 |
| lucasagomes | yeah I couldn't find any example | 10:36 |
| sambetts | lucasagomes: I have found that in project config it only realises its a regex if it begins with ^ and ends with $ | 10:37 |
| sambetts | lucasagomes: maybe give that a go | 10:37 |
| lucasagomes | sambetts, oh, lemme check | 10:37 |
| mgould | sambetts: I think that's what happened to me on the way back from Kyrgyzstan | 10:39 |
| mgould | travelled through Istanbul airport, full of people from all over the world travelling to or from the Hajj, when I was already tired | 10:40 |
| *** links has quit IRC | 10:40 | |
| mgould | hence my somewhat delayed return to work :-( | 10:41 |
| sambetts | :( | 10:42 |
| * milan 's devstack got sick too | 10:43 | |
| *** moshele has joined #openstack-ironic | 10:43 | |
| milan | anyone successfully running devstack on a Fedora host? | 10:43 |
| *** tuanluong has quit IRC | 10:43 | |
| lucasagomes | milan, I did it before the summit I think, on f24 | 10:44 |
| sambetts | milan: nope ubuntu 16 here | 10:44 |
| * milan git pulled and now cries | 10:44 | |
| milan | for some reason, /etc/ dirs get created with myself as the owner instead of stack | 10:45 |
| sambetts | :/ | 10:45 |
| milan | cp: cannot create regular file '/etc/keystone/policy.json': Permission denied | 10:45 |
| milan | yeah | 10:45 |
| milan | devstack influenza | 10:45 |
| sambetts | I always use a ubuntu vm with the main user called stack so I don't think i'd see that anyway | 10:46 |
| milan | sudo install -d -o mkovacik /etc/keystone | 10:46 |
| milan | that's what the log says | 10:46 |
| milan | sambetts, so you sudo stack then ./stack.sh? | 10:46 |
| sambetts | milan: no, I just create the VM with the main user as stack | 10:47 |
| milan | sambetts, ah ok | 10:47 |
| TheJulia | bcornec: I took a look at 391593, I think you just need to nuke the explicit gate definition, but I'm by no means an expert on it | 10:48 |
| * milan will try w/ stack user; -> lunch -> | 10:49 | |
| mgould | milan: I tend to run devstack on centos | 10:49 |
| sambetts | milan: doing sudo su stack would probably work too, but I don't think you should need to do it :/ | 10:49 |
| milan | sambetts, that was my understanding too | 10:49 |
| milan | mgould, I suppose all is well there ;) | 10:49 |
| milan | mgould, you run as you or as stack? | 10:49 |
| mat128 | good morning guys | 10:50 |
| milan | morning mat128! :) | 10:50 |
| TheJulia | I need to fire up a fresh devstack | 10:50 |
| dtantsur | morning mat128 | 10:50 |
| lucasagomes | milan, you are following: http://docs.openstack.org/developer/ironic/dev/dev-quickstart.html#deploying-ironic-with-devstack ? | 10:50 |
| lucasagomes | right ? | 10:50 |
| milan | lucasagomes, heh, so I have to sudo -u stack it | 10:52 |
| milan | must be the hunger! | 10:52 |
| * milan runs away before lucasagomes finds more issues :P | 10:53 | |
| milan | thanks lucasagomes :) | 10:53 |
| * milan gone food | 10:54 | |
| *** milan has quit IRC | 10:54 | |
| *** phuongnh has quit IRC | 11:02 | |
| TheJulia | dtantsur: I may have proposed something a little crazy on https://review.openstack.org/#/c/184653/15/specs/approved/ironic-redfish.rst@92 :) | 11:02 |
| patchbot | patch 184653 - ironic-specs - Propose Redfish support into Ironic | 11:02 |
| * dtantsur likes crazy proposals | 11:02 | |
| mat128 | TheJulia: not sure I like having validate modify driver_info, but we'll see how others react | 11:03 |
| TheJulia | I'm not exactly a fan of it either | 11:03 |
| mgould | milan: as stack | 11:04 |
| mgould | and yeah, it seems to work fine | 11:04 |
| mat128 | TheJulia: maybe there cant be any default for this value? | 11:04 |
| TheJulia | but if it is a single node, recording it would kind of make sense. driver_internal_info is another possibility, but users can't change that | 11:04 |
| TheJulia | mat128: I think that would make dtantsur and myself sad. :) | 11:04 |
| mgould | I thought we were holding off supporting redfish until it's feature-complete? | 11:05 |
| dtantsur | TheJulia, we can cache it locally in a conductor even | 11:05 |
| TheJulia | Well, they released 1.0 | 11:05 |
| *** Nisha1 has joined #openstack-ironic | 11:05 | |
| dtantsur | TheJulia, the only thing I don't like it using validate() for that, as it's always synchronous | 11:05 |
| mat128 | TheJulia: maybe we can default to "the" system if we find only and exactly 1 entry? | 11:05 |
| mgould | TheJulia: right, but bcornec was saying that it's still missing stuff we need and hence encouraging incompatible vendor extensions to breed | 11:05 |
| dtantsur | mgould, well, we can't support inspection and raid | 11:06 |
| TheJulia | mgould: what dtantsur said | 11:06 |
| mgould | OK, cool | 11:06 |
| dtantsur | but it seems like we can support power and management without vendor extensions | 11:06 |
| mat128 | mgould: tbh I dont even think redfish will solve that vendor extensions problem | 11:06 |
| * dtantsur neither | 11:06 | |
| TheJulia | mgould: possibly, but I would hate to see multiple api exchanges to take any action. | 11:06 |
| TheJulia | any single action that is | 11:06 |
| mat128 | TheJulia: there is no way around it, unless you can tell redfish "operate on the default system" | 11:07 |
| mgould | mat128: so I was arguing with bcornec about this at the summit | 11:07 |
| TheJulia | I don't think it will solve it, as vendors are vendors, but this actually fits in as base interfaces for driver composition later :) | 11:07 |
| mgould | I'd be happy-ish with a default way of asking "which set of vendor extensions do you support?" | 11:07 |
| TheJulia | mat128: Well, if we can do it once and store it somehow, and validation seems to be the logical place as that would provide a nice error back to the user enrolling the node | 11:08 |
| mat128 | mgould: but then you'd get "hp-ilo-kvm-console", "dell-idrac-console" | 11:08 |
| mat128 | all different :( | 11:08 |
| mgould | the nightmare I want to avoid is http://esr.ibiblio.org/?p=801 | 11:08 |
| TheJulia | dtantsur: your right though, it is synchronous.... :\ | 11:08 |
| dtantsur | TheJulia, well, we can just do it on every request, but with local caching, I guess | 11:08 |
| mat128 | TheJulia: so validate() should ensure the thing works, but we cant rely on validate having run to cache the value | 11:08 |
| mgould | "I need to use vendor extensions, and I don't know which ones are available, so I have to cleverly infer what I'm dealing with" | 11:08 |
| TheJulia | dtantsur: Do we really really want to add caching layer? | 11:09 |
| *** mmedvede has quit IRC | 11:09 | |
| TheJulia | mat128: well, most validate calls actually just check that the values are present, because it is a synchronous call | 11:09 |
| dtantsur | TheJulia, well, I don't see anything particularly bad in caching the URL for a few minutes. I don't see anything too bad in NOT caching either. | 11:09 |
| sambetts | so we went through a similar discuss to this with the oneview guys | 11:10 |
| mat128 | dtantsur, TheJulia: caching and the caching ttl can be configureable I guess | 11:10 |
| sambetts | basically we decided to add a periodic task that cached the information, then validation only ever read the cached information | 11:10 |
| TheJulia | mat128: I guess I'm hoping to avoid things like cache invalidation issues :) | 11:10 |
| TheJulia | sambetts: I thought they were storing all of the data, interesting. | 11:11 |
| *** mjturek has quit IRC | 11:11 | |
| mat128 | sambetts: so validate fails until the task runs? | 11:12 |
| mat128 | not sure I'd like to have more of those "add node to ironic, wait for nova's tasks to run, then boot" | 11:12 |
| *** absubram has joined #openstack-ironic | 11:13 | |
| TheJulia | validate would be a nice way to force it, then again power validation could just do the same thing, and the power state checking is a periodic task in ironic. | 11:13 |
| TheJulia | mat128: That makes sense, so then your waiting twice :\ | 11:13 |
| sambetts | TheJulia: they wanted the validation for the "is this node used by someone outside of Ironic?" use case, but that meant making a external call on validation, so adding a pretty regular periodic tasks that populated a internel field on the node that was like "used=True/False" then validation look at that | 11:14 |
| *** absubram_ has joined #openstack-ironic | 11:14 | |
| TheJulia | sambetts: ahh, yes! | 11:14 |
| mat128 | mgould: omg that GPS link... never knew it was such a nightmare | 11:14 |
| TheJulia | sambetts: I was thinking purely in memory, not driver_internal_info | 11:14 |
| dtantsur | TheJulia, mat128, I see it more like the URL is not mandatory during validation, but if it's not provided, then on every power operation we fetch the first system URL and use it (with optional caching) | 11:15 |
| mgould | mat128: ikr? | 11:15 |
| mat128 | mgould: all I knew was "oh, simple NMEA strings, text protocol -- I've seen worse" | 11:15 |
| mat128 | but when you dig into it.. omg | 11:15 |
| mgould | mat128: yeah | 11:16 |
| mgould | I think they started with a protocol designed for a very different use-case, and extended it in multiple, incompatible, stupid ways | 11:16 |
| *** dprince has joined #openstack-ironic | 11:16 | |
| mat128 | mgould: "That is, unless you buffer, in which case the altitude you see could be up to one second stale and associated with a previous fix." | 11:17 |
| sambetts | its similar to the LLDP type 127 for vendor extensions, basically every vendor value is 127, then they use vendor specifc sub-types and sub-sub-types to indicate what they actually mean | 11:17 |
| mat128 | mgould: that explains why altitude displayed on my car's GPS is always late by a second or two | 11:17 |
| mgould | mat128: aaaargh | 11:17 |
| mat128 | TIL I guess | 11:17 |
| *** absubram has quit IRC | 11:17 | |
| *** absubram_ is now known as absubram | 11:17 | |
| mat128 | dtantsur: I like your idea | 11:18 |
| mat128 | dtantsur: calls that need it will go fetch it | 11:18 |
| mat128 | and we can tackle on caching if reqd | 11:18 |
| mgould | mat128: and why I've had to spend time on cold mountain summits waiting for the altitude reading to settle :-) | 11:18 |
| TheJulia | dtantsur: I think that would work stored in an internal field, I just worry that someone might break something by changing settings in the controller, but if they do that, there is really not much we can do. | 11:18 |
| mat128 | mgould: :P | 11:18 |
| dtantsur | yep.. | 11:18 |
| mat128 | TheJulia: if they know they're going to do this often, they can lower the TTL and/or disable caching | 11:19 |
| mat128 | wdyt? | 11:19 |
| sambetts | the devstack plugin's are branched right?? | 11:19 |
| TheJulia | I guess the other thing is if it multiple systems behind the redfish url, power sync check could fail fast | 11:19 |
| mat128 | sambetts: they're in-tree, so yeah | 11:19 |
| mat128 | TheJulia: what will "systems" look like irl? | 11:20 |
| mat128 | like subsystems within the machine? | 11:20 |
| sambetts | mat128: so's the tempest plugin, but we always run the plugin from master against all versions of Ironic @_@ | 11:20 |
| mat128 | hmm | 11:20 |
| TheJulia | mat128: each system is a distinct path in systeminventory, which is /redfish/v1/Systems/ | 11:20 |
| mat128 | TheJulia: what I mean is, in a typical machines what will /Systems look like | 11:20 |
| mgould | sambetts, mat128: the tempest folks want us to move the tempest plugin out into a new branchless ironic-tempest-plugin repo | 11:20 |
| TheJulia | so actions upon it would be /redfish/v1/Systems/firstandonlysystem | 11:21 |
| sambetts | mgould: yeah, didn't know if the same thing for devstack or not, I assume no? | 11:21 |
| mgould | because tempest is really a cloud-smoketester that does CI as a hobby | 11:21 |
| vdrok | good morning everyone! dtantsur lucasagomes mgould sambetts TheJulia mat128 | 11:21 |
| mat128 | sambetts, mgould: I guess tempest has code to discover features and fail gracefully if they're not present? | 11:21 |
| mgould | sambetts: dunno about devstack, sorry | 11:21 |
| mat128 | o/ vdrok | 11:21 |
| TheJulia | mat128: I believe it is an object that is returned listing the inventory of machines known to the controller | 11:21 |
| lucasagomes | vdrok, hi there! | 11:21 |
| TheJulia | good morning vdrok | 11:21 |
| mgould | mat128: that's the whole point of tempest :-) | 11:21 |
| mat128 | TheJulia: Sorry I phrased it wrong, I got that | 11:22 |
| dtantsur | morning vdrok | 11:22 |
| mat128 | TheJulia: I mean, will I only see a single system in there for a typical 1U server? | 11:22 |
| mgould | morning vdrok | 11:22 |
| TheJulia | mat128: I believe so yes | 11:22 |
| mat128 | TheJulia: i.e. was this for chassis with blades? | 11:22 |
| mat128 | ok | 11:22 |
| TheJulia | mat128: pulling up the spec for the protocol | 11:22 |
| dtantsur | mat128, "tempest has code to discover features and fail gracefully if they're not present" no. you have to set it in configuration and pray :/ | 11:22 |
| *** mmedvede has joined #openstack-ironic | 11:22 | |
| TheJulia | mat128: basically, yes, it was built with carriages/blades in mind | 11:22 |
| mat128 | dtantsur: ooh | 11:22 |
| mat128 | TheJulia: ok, so we're more likely to see a single entry there | 11:23 |
| mat128 | thats good | 11:23 |
| mgould | dtantsur: AIUI, tempest is meant to perform tests according to reported API versions supported | 11:23 |
| *** jkilpatr has joined #openstack-ironic | 11:24 | |
| dtantsur | mgould, API versions is not everything. or rather: it's nearly nothing :) | 11:24 |
| mgould | WIP summit notes, search for "Tempest": https://www.dropbox.com/home/lj?preview=ocata_summit.md | 11:24 |
| mgould | dtantsur: OK, the tempest folks seem to have a higher opinion of API versions than you | 11:25 |
| mat128 | mgould: not sure we can read this | 11:25 |
| mat128 | mgould: I think you have to get a "share url" | 11:25 |
| mgould | mat128: sorry, try https://www.dropbox.com/s/xwvk1g93s71ejys/ocata_summit.md?dl=0 | 11:25 |
| sambetts | sigh... /me has just run into the symantics of "ironic network interface" vs "network interface used for ironic" | 11:25 |
| mgould | dtantsur: seriously, I spent a lot of time talking to dmellado at the summit trying to make sure I had this straight | 11:26 |
| dtantsur | mgould, well, API versions don't tell us which nova backend supports what.. they're only discussion reporting such capabilities | 11:26 |
| mat128 | mgould: and... no search in Dropbox ui :( | 11:26 |
| mgould | mat128: gah! scroll down to "Thursday" | 11:26 |
| *** jerrygb has joined #openstack-ironic | 11:26 | |
| mgould | dtantsur: can you talk me through a test you want to run? | 11:27 |
| mgould | I think the answer is going to be "don't use tempest", but I think I need a concrete example to think about | 11:27 |
| jroll | sambetts: yes, devstack is branched thus devstack plugins are branched | 11:27 |
| jroll | morning :) | 11:28 |
| *** awiddersheim has quit IRC | 11:28 | |
| mgould | morning jroll | 11:28 |
| sambetts | jroll: awesome, then I don't have to worry about certain API things not existing while I write the plugin right? | 11:28 |
| mgould | btw, what's the standard thing to do with summit notes? ML, blog post, /dev/null? | 11:28 |
| dtantsur | mgould, well, from tempest point of view we should be able to not use some features when it uses the ironic backend | 11:29 |
| dtantsur | morning jroll! I plan on dracclient release pretty soon btw | 11:29 |
| jroll | sambetts: right | 11:29 |
| sambetts | \o/ | 11:29 |
| TheJulia | mat128: https://www.dmtf.org/standards/redfish if you want to bulk load the data into your brain and then let a background process try and parse it :) | 11:29 |
| jroll | dtantsur: wanna race between that release and dropping it from governance? :) | 11:29 |
| mgould | dtantsur: I +Wed a bunch of stuff in dracclient last night, but you probably saw that | 11:29 |
| jroll | dtantsur: (which means we can't use the releases repo (yet)) | 11:29 |
| TheJulia | good morning jroll | 11:29 |
| jroll | \o TheJulia | 11:30 |
| mgould | dtantsur: riiiiiiight | 11:30 |
| mat128 | TheJulia: I've seen that before when I wanted to write a simulator for it (not knowing it was already in the works) and it was way too abstract for me to even simulate a real machine | 11:30 |
| mat128 | o/ jroll | 11:30 |
| *** AlexeyAbashkin has quit IRC | 11:30 | |
| *** links has joined #openstack-ironic | 11:30 | |
| mgould | so this is a "can't use existing tempest tests against Ironic" problem, not a "can't test the Ironic API with Tempest" problem | 11:30 |
| TheJulia | mat128: Yeah, that. :( | 11:30 |
| dtantsur | jroll, right. I'd prefer to do it the easy way, as doing it manually is going to be fun with signing, etc | 11:30 |
| jroll | morning mat128, mgould, sambetts, dtantsur, anyone else I didn't say hi to yet | 11:31 |
| mat128 | mgould: I read through your notes and it was a great (short) read. I suggest you post it to the ML and ask others to do the same :) | 11:31 |
| jroll | dtantsur: yeah, you can go first :) | 11:31 |
| mgould | mat128: thanks! I'll try to do that this afternoon :-) | 11:31 |
| dtantsur | cool | 11:31 |
| openstackgerrit | Anton Arefiev proposed openstack/ironic-inspector: Allow to configure cache clean up period https://review.openstack.org/392943 | 11:32 |
| *** roeyc has joined #openstack-ironic | 11:32 | |
| *** AlexeyAbashkin has joined #openstack-ironic | 11:35 | |
| openstackgerrit | Julia Kreger proposed openstack/ironic: Add volume_connector table to save connector information https://review.openstack.org/200983 | 11:35 |
| openstackgerrit | Anton Arefiev proposed openstack/ironic-inspector: Allow to configure cache clean up period https://review.openstack.org/392943 | 11:35 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: DNM: Test patch to verify bond https://review.openstack.org/388660 | 11:37 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Ironic devstack portgroup support https://review.openstack.org/381743 | 11:37 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Rely on portgroup standalone_ports_supported https://review.openstack.org/360647 | 11:37 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Tempest tests for portgroups https://review.openstack.org/382476 | 11:37 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Add PortGroups API https://review.openstack.org/347549 | 11:37 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Integrate portgroups with ports to support LAG https://review.openstack.org/332177 | 11:37 |
| mgould | mat128: I'm having real difficulty finding actual routes in the Redfish API schema :-( | 11:40 |
| bfournie | morning ironic | 11:40 |
| dtantsur | morning bfournie | 11:41 |
| TheJulia | mgould: that was my complaint to bcornec | 11:42 |
| bfournie | morning dtantsur | 11:42 |
| mgould | TheJulia: is it input to some REST-API building tool? | 11:42 |
| TheJulia | I think so yes :( | 11:42 |
| mgould | bah | 11:43 |
| mgould | still useful if it can also generate some basic tests we can use, mind | 11:43 |
| TheJulia | There is XML structural representations out there as well for that purpose, I think what put some of it together for me mentally was the in-progress conversion stuff they are building | 11:43 |
| TheJulia | s/conversion/translation/ | 11:43 |
| *** jaosorior_lunch is now known as jaosorior | 11:44 | |
| *** AlexeyAbashkin has quit IRC | 11:49 | |
| *** AlexeyAbashkin has joined #openstack-ironic | 11:50 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: DevStack: Support for creating UEFI VMs https://review.openstack.org/374346 | 11:55 |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: DevStack: Configure nodes/environment to boot in UEFI mode https://review.openstack.org/374988 | 11:55 |
| *** aparnav has quit IRC | 11:55 | |
| *** rbrud has joined #openstack-ironic | 12:00 | |
| *** bfournie has quit IRC | 12:00 | |
| openstackgerrit | Merged openstack/python-ironic-inspector-client: Updated from global requirements https://review.openstack.org/389938 | 12:00 |
| *** bfournie1 has joined #openstack-ironic | 12:02 | |
| *** krypto has joined #openstack-ironic | 12:03 | |
| *** xavierr has joined #openstack-ironic | 12:04 | |
| openstackgerrit | Merged openstack/bifrost: playbooks: Do not serialize ironic deployment https://review.openstack.org/389079 | 12:04 |
| *** Nisha1 has quit IRC | 12:04 | |
| openstackgerrit | Merged openstack/bifrost: Don't include openstack/common in flake8 exclude list https://review.openstack.org/391703 | 12:04 |
| *** absubram has quit IRC | 12:06 | |
| dtantsur | jroll, I still plan on applying http://governance.openstack.org/reference/tags/assert_supports-upgrade.html to ironic-inspector. What does Ironic have for "A procedure for general upgrades of the project is defined and does not change substantially from cycle to cycle."? | 12:09 |
| jroll | dtantsur: http://docs.openstack.org/developer/ironic/deploy/upgrade-guide.html | 12:09 |
| dtantsur | aha! I guess I'll cargo-cult it to inspector docs, and then apply for the tag | 12:10 |
| jroll | :) | 12:10 |
| *** cdaniel has joined #openstack-ironic | 12:11 | |
| dtantsur | jroll, btw it does not mention ordering with IPA | 12:13 |
| dtantsur | do we expect IPA to usually be upgraded before or after ironic{,-inspector}? | 12:14 |
| jroll | dtantsur: we try to support both, we only test ironic upgrade first | 12:14 |
| jroll | s/first/before IPA/ | 12:14 |
| dtantsur | interesting, I guess that's what we should recommend then | 12:15 |
| sambetts | my spec for IPA versioning should include testing n-1 and n+1 | 12:15 |
| * dtantsur thinks what should be upgraded first: ironic or ironic-inspector | 12:16 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic-inspector: Document upgrade procedure https://review.openstack.org/393275 | 12:18 |
| dtantsur | sambetts, jroll, wdyt ^^^ | 12:18 |
| jroll | uno momento | 12:19 |
| sambetts | dtantsur: looks good, we'll have to always make sure new ironic inspector interface works with old inspector though | 12:20 |
| sambetts | dtantsur: so we'll not be able to add a feature to ironic inspector interface for at least a cycle | 12:20 |
| sambetts | for example the manage_boot API | 12:21 |
| dtantsur | mmmmmm | 12:21 |
| *** yogi has quit IRC | 12:22 | |
| dtantsur | well, we should be careful to use API discovery or something similar. or just not enable new features by default | 12:22 |
| sambetts | yeah, we'll have to ping inspector API version in ironic or something | 12:22 |
| sambetts | pin* | 12:22 |
| dtantsur | I think we do already | 12:23 |
| sambetts | or have a warning on a config option that like "don't turn this on unless your inspector is greater than this version" | 12:23 |
| dtantsur | yeah | 12:23 |
| *** nicodemos has joined #openstack-ironic | 12:25 | |
| *** ralonsoh_ has joined #openstack-ironic | 12:30 | |
| *** jerrygb has quit IRC | 12:30 | |
| nicodemos | good morning, ironicers. =D | 12:30 |
| sambetts | o/ nicodemos | 12:31 |
| *** ralonsoh has quit IRC | 12:32 | |
| nicodemos | hey, sambetts. \o | 12:32 |
| *** lucasagomes is now known as lucas-hungry | 12:35 | |
| *** daemontool has joined #openstack-ironic | 12:37 | |
| *** hw_wutianwei has quit IRC | 12:41 | |
| *** dtantsur is now known as dtantsur|brb | 12:41 | |
| *** jheroux has joined #openstack-ironic | 12:43 | |
| openstackgerrit | Sam Betts proposed openstack/ironic: Rework Ironic devstack baremetal network simulation https://review.openstack.org/392959 | 12:44 |
| *** rbudden has joined #openstack-ironic | 12:48 | |
| *** syed_ has joined #openstack-ironic | 12:49 | |
| *** ccamacho is now known as ccamacho|lunch | 12:49 | |
| *** baoli has joined #openstack-ironic | 12:54 | |
| *** david-lyle has quit IRC | 12:55 | |
| *** baoli_ has joined #openstack-ironic | 12:56 | |
| *** baoli has quit IRC | 13:00 | |
| *** rbartal has quit IRC | 13:01 | |
| *** mjturek has joined #openstack-ironic | 13:04 | |
| *** jcoufal has joined #openstack-ironic | 13:04 | |
| *** mjturek has quit IRC | 13:08 | |
| *** rloo has joined #openstack-ironic | 13:09 | |
| *** jerrygb has joined #openstack-ironic | 13:10 | |
| *** rbartal has joined #openstack-ironic | 13:13 | |
| mgould | mat128: just posted my notes to the ML | 13:14 |
| mgould | added a sentence or two about each session, so they're no longer as short :-( | 13:14 |
| *** vikrant has quit IRC | 13:15 | |
| *** ppiela has joined #openstack-ironic | 13:19 | |
| *** jerrygb_ has joined #openstack-ironic | 13:21 | |
| *** jerrygb has quit IRC | 13:22 | |
| *** milan has joined #openstack-ironic | 13:22 | |
| *** mjturek has joined #openstack-ironic | 13:22 | |
| *** Alexey_Abashkin has joined #openstack-ironic | 13:27 | |
| openstackgerrit | Aline Bousquet proposed openstack/ironic: Add possibility to remove chassis_uuid from a node https://review.openstack.org/390558 | 13:28 |
| *** Goneri has joined #openstack-ironic | 13:29 | |
| *** dtantsur|brb is now known as dtantsur | 13:30 | |
| *** AlexeyAbashkin has quit IRC | 13:30 | |
| rloo | hi and morning all ironic'ers. (that reminds me, have to give feedback on angry bear) | 13:31 |
| nicodemos | morning, rloo. | 13:31 |
| dtantsur | morning rloo | 13:32 |
| rloo | alineb: i saw you just updated 390558. I want to make sure you saw the comment i added just as you updated. | 13:32 |
| rloo | hi nicodemos, dtantsur, mgould | 13:32 |
| rloo | alineb: thx for rebasing | 13:32 |
| alineb | rloo: yes, i just saw your comment: thanks for pointing that out, i will update the patch | 13:34 |
| rloo | thx alineb! | 13:34 |
| alineb | rloo: regarding your question, the code does allow to change chassis_uuid and to create a node without one | 13:34 |
| *** lucas-hungry is now known as lucasagomes | 13:35 | |
| rloo | alineb: ok, i couldn't remember. so the version is just for removing/unsetting it - just need to document that. thx. | 13:35 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Cleanup hung iscsi session https://review.openstack.org/388092 | 13:35 |
| alineb | rloo: ok, thanks a lot for your comments! | 13:35 |
| rloo | alineb: yw :) | 13:36 |
| rloo | alineb: sorry we didn't see you in barcelona | 13:36 |
| openstackgerrit | Julia Kreger proposed openstack/ironic: Introduce VolumeConnector object https://review.openstack.org/214584 | 13:36 |
| openstackgerrit | Julia Kreger proposed openstack/ironic: Introduce new RPCs to support volume connector operation https://review.openstack.org/214585 | 13:37 |
| openstackgerrit | Julia Kreger proposed openstack/ironic: Add volume_targets table to save target information https://review.openstack.org/285218 | 13:37 |
| alineb | rloo: yes, it would have been great to go, maybe next summit :) | 13:39 |
| rloo | alineb: yes! | 13:39 |
| aarefiev | morning rloo | 13:40 |
| rloo | hi aarefiev! | 13:40 |
| aarefiev | dtantsur: we can discover current inspector api version, right? | 13:41 |
| dtantsur | aarefiev, we can, yes | 13:41 |
| aarefiev | what if check it before calling inspector | 13:41 |
| aarefiev | and rollback to previous one | 13:42 |
| aarefiev | in case not supported | 13:42 |
| milan | morning alineb, nicodemos, rloo! :) | 13:42 |
| *** krtaylor has joined #openstack-ironic | 13:44 | |
| dtantsur | aarefiev, that's what I call API discovery, yes | 13:45 |
| rloo | hiya milan | 13:45 |
| openstackgerrit | Aline Bousquet proposed openstack/ironic: Add possibility to remove chassis_uuid from a node https://review.openstack.org/390558 | 13:45 |
| milan | guys, anyone seeing this in devstack: cp: cannot stat '/opt/stack/nova/etc/nova/policy.json': No such file or directory | 13:49 |
| * milan tries tox -egenpolicy with no luck | 13:49 | |
| *** amotoki has joined #openstack-ironic | 13:50 | |
| aarefiev | dtantsur: is there issues with it now? | 13:51 |
| aarefiev | why not using it | 13:51 |
| *** links has quit IRC | 13:52 | |
| *** shuquan has joined #openstack-ironic | 13:53 | |
| *** pcaruana has joined #openstack-ironic | 13:53 | |
| dtantsur | aarefiev, we never had to yet | 13:53 |
| dtantsur | the manage_boot case will be the first | 13:54 |
| aarefiev | ok, I think we need to do it before land manage_boot feature | 13:56 |
| openstackgerrit | Michael Turek proposed openstack/ironic: Enable PXE for systems using petitboot https://review.openstack.org/185987 | 13:57 |
| *** jerrygb has joined #openstack-ironic | 13:57 | |
| *** shuquan_ has joined #openstack-ironic | 13:58 | |
| *** shuquan has quit IRC | 13:59 | |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Cleanup hung iscsi session https://review.openstack.org/388092 | 13:59 |
| mgould | morning rloo | 14:00 |
| sambetts | TheJulia: when using ironic standalone how do you manage the DHCP configuration for PXE? | 14:00 |
| *** anush has joined #openstack-ironic | 14:00 | |
| *** jerrygb_ has quit IRC | 14:00 | |
| TheJulia | sambetts: we explicitly point all pxe booting clients to the initial pxe script which if memory serves upgrades them to ipxe, ipxe is replied to with the ipxe script | 14:01 |
| dtantsur | sambetts, mind checking https://review.openstack.org/374381 again please when you have a minute? | 14:03 |
| patchbot | patch 374381 - ironic-inspector-specs - Add LLDP processing hook and new CLI commands | 14:03 |
| sambetts | TheJulia: so you have a pxe all rule not per mac address, I guess you rely on the cleanup stage of the pxe boot interface to prevent the node re-pxeing into the ramdisk ? | 14:03 |
| sambetts | dtantsur: sure | 14:04 |
| rloo | dtantsur (or anyone else that recalls microversions, i know dmitry loves them). don't we raise exception.NotAcceptable if a request is made with the 'wrong' microversion? https://review.openstack.org/#/c/347549/36..37/ironic/api/controllers/v1/portgroup.py | 14:05 |
| patchbot | patch 347549 - ironic - Add PortGroups API | 14:05 |
| mgould | rloo: yep | 14:06 |
| jroll | rloo: we raise what makes the most sense - e.g. here we do NotFound because in lower versions that's what you would get | 14:06 |
| dtantsur | rloo, usually. except for the case that confuses you: when a completely new endpoint is introduced | 14:06 |
| dtantsur | rloo, that was raised on the lookup API patches too: https://github.com/openstack/ironic/blob/master/ironic/api/controllers/v1/ramdisk.py#L98-L99 | 14:07 |
| mgould | we also raise NotAcceptable in a few other situations :-( | 14:07 |
| dtantsur | I can say a lot of nasty things about trying to fix all possible API errors in a very limited set of HTTP codes.. | 14:07 |
| rloo | jroll, dtantsur: oh, honestly, it doesn't make sense to me what we return. i just thought that we always returned NotAcceptable for mcroversion-related changes. | 14:08 |
| mgould | ah, wait, maybe not | 14:08 |
| dtantsur | rloo, I used to think the same :) | 14:08 |
| rloo | dtantsur: it was so much easier to remember one rule. i wonder if we've messed things up now. | 14:08 |
| mgould | the thing I thought was a non-microversion use was actually checking microversions internally | 14:09 |
| rloo | (that's what we get for not documenting these things) | 14:09 |
| jroll | rloo: part of the point of microversions is to keep from changing behavior in a given version | 14:09 |
| rloo | jroll: then why would we ever return NotAcceptable? | 14:09 |
| dtantsur | welcome to the confused-by-microversions camp, rloo :) | 14:10 |
| rloo | jroll: Bad Request would make more sense | 14:10 |
| rloo | dtantsur: i was always in that camp; just wanted to be a team player though :) | 14:10 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: DNM: Test patch to verify bond https://review.openstack.org/388660 | 14:10 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Ironic devstack portgroup support https://review.openstack.org/381743 | 14:10 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Rely on portgroup standalone_ports_supported https://review.openstack.org/360647 | 14:10 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Tempest tests for portgroups https://review.openstack.org/382476 | 14:10 |
| jroll | rloo: I really don't remember why we decided NotAcceptable was the best thing to do for many of the changes | 14:10 |
| dtantsur | jroll, well, what else? | 14:11 |
| dtantsur | Bad Request is fine, but too broad | 14:11 |
| jroll | dtantsur: right, I don't disagree | 14:11 |
| jroll | rloo asked why NotAcceptable, I don't recall | 14:11 |
| dtantsur | in particular, returning a specific code for versioning issues allows for some heuristic on client side | 14:11 |
| dtantsur | ... which we break by returning NotFound :( | 14:11 |
| rloo | ^^ which is why i thought we were always using NotAcceptable | 14:12 |
| jroll | mmm. | 14:12 |
| rloo | (although we could be nice and add a message why something isn't acceptable... i think that is a TODO...) | 14:12 |
| dtantsur | yeah.. though ironicclient actually composes a nice message itself | 14:12 |
| rloo | i thought there was some xproject microversion spec but didn't find it. was wondering if it was mentioned elsewhere. | 14:13 |
| dtantsur | rloo, there used to be. I think only some smaller part was merged | 14:13 |
| rloo | anyway, jroll, dtantsur: you are good with NotFound in the portgroup case? i don't want to hold that patch up. | 14:13 |
| TheJulia | sambetts: Technically I guess we do, although any default failure with bifrost pxe boots the agent for inspector usage | 14:13 |
| dtantsur | rloo, unfortunately, this is consistent with the lookup and heartbeat, so yes. I am. | 14:13 |
| jroll | rloo: yes | 14:13 |
| vsaienk0 | rloo, it was decided with devananda to return NotFound() when we perform call to endpoint that shouldn't exist in old version | 14:14 |
| rloo | dtantsur: i think it is inconsistent with the RAID ones :-( | 14:14 |
| TheJulia | sambetts: sorry for the delay responding, on another computer working on expense report stuffs | 14:14 |
| rloo | vsaienk0: sigh. we should document these things. | 14:14 |
| dtantsur | we've screwed up V1, let's drop it this cycle!!1 | 14:14 |
| sambetts | TheJulia: hehe same | 14:14 |
| rloo | dtantsur: i'm on it! | 14:14 |
| sambetts | kill it with fire | 14:14 |
| *** shuquan_ has quit IRC | 14:15 | |
| vsaienk0 | rloo document where? | 14:15 |
| TheJulia | sambetts: I don't think I've ever been someplace that had such a desire for cash transactions | 14:15 |
| *** mtanino has joined #openstack-ironic | 14:15 | |
| TheJulia | vdrok: tribal-knowledge.rst? | 14:15 |
| sambetts | TheJulia: I know its crazy... so many reciepts to track | 14:15 |
| TheJulia | err | 14:15 |
| TheJulia | vsaienk0: | 14:15 |
| rloo | TheJulia: more like, we don't have dev docs on microversions :-( | 14:15 |
| TheJulia | rloo: hmmmmmm | 14:16 |
| rloo | TheJulia: eg, yesterday JayF was asking *when* something needed a version bump | 14:16 |
| TheJulia | contextual documentation to communicate how/why we leverage ? | 14:16 |
| *** jaosorior has quit IRC | 14:17 | |
| TheJulia | yeah, that was tribal context at some point in the past, and has been debated many times here in IRC. We should have documented it in something, anything even a manifesto if needed. :( | 14:17 |
| * dtantsur would love to see a spec *before* jumping into implementing something like that | 14:17 | |
| * dtantsur is probably just too naive | 14:17 | |
| *** vinsh has joined #openstack-ironic | 14:18 | |
| *** shobhar_ has joined #openstack-ironic | 14:20 | |
| rloo | vsaienk0: I +2'd the portgroup patch but ... i think the release notes ought to be updated. | 14:22 |
| *** yolanda has quit IRC | 14:22 | |
| *** yolanda has joined #openstack-ironic | 14:23 | |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: DNM: Test patch to verify bond https://review.openstack.org/388660 | 14:27 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Ironic devstack portgroup support https://review.openstack.org/381743 | 14:27 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Rely on portgroup standalone_ports_supported https://review.openstack.org/360647 | 14:27 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Tempest tests for portgroups https://review.openstack.org/382476 | 14:27 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Add PortGroups API https://review.openstack.org/347549 | 14:27 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Integrate portgroups with ports to support LAG https://review.openstack.org/332177 | 14:27 |
| *** absubram has joined #openstack-ironic | 14:28 | |
| vsaienk0 | rloo: just uploaded a new version | 14:30 |
| *** sabeen has joined #openstack-ironic | 14:30 | |
| rloo | thx vsaienk0! | 14:32 |
| *** krypto has quit IRC | 14:33 | |
| *** yolanda has quit IRC | 14:33 | |
| *** sabeen has quit IRC | 14:35 | |
| rloo | yuriyz, JayF: if you have time to review 347549 today? maybe we can +A it. | 14:35 |
| yuriyz | hi rloo ironic will do today | 14:36 |
| rloo | yuriyz: thx. btw, wrt notifications, is the coding done? | 14:37 |
| *** sabeen has joined #openstack-ironic | 14:37 | |
| yuriyz | rloo only provision states, CRUD etc in progress | 14:37 |
| * mariojv will re-review provision state today | 14:38 | |
| rloo | yuriyz: oh. i think i meant, is the coding ready to be reviewed :) | 14:38 |
| *** yolanda has joined #openstack-ironic | 14:38 | |
| rloo | yuriyz: oh, wait, i am confused. got it now. | 14:38 |
| *** athomas has quit IRC | 14:39 | |
| *** livelace has joined #openstack-ironic | 14:39 | |
| dtantsur | jroll, have you seen anything like https://bugs.launchpad.net/ironic/+bug/1637210 ? | 14:40 |
| openstack | Launchpad bug 1637210 in Ironic "Conductor's periodic tasks stops executing" [Undecided,New] | 14:40 |
| openstackgerrit | Félix Bouliane proposed openstack/ironic: DONOTMERGE-change IRONIC_DEPLOY_DRIVER to pxe_snmp https://review.openstack.org/388168 | 14:41 |
| *** ccamacho|lunch is now known as ccamacho | 14:42 | |
| *** athomas has joined #openstack-ironic | 14:46 | |
| *** shobhar_ has quit IRC | 14:47 | |
| jroll | dtantsur: no, I have not O_o | 14:49 |
| dtantsur | strange... | 14:50 |
| jroll | sambetts: hey, nova attach/detach BP is fine without a spec, they'd like the ironic spec merged before approving it, so let's get that done next week | 14:50 |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: DevStack: Fix standard PXE on Ubuntu Xenial https://review.openstack.org/393353 | 14:50 |
| sambetts | jroll: oh awesome :D | 14:51 |
| sambetts | jroll: thanks for chasing that, if there was a nova meeting i totally missed it... | 14:51 |
| jroll | sambetts: no worries, I told you monday I'd attend :) | 14:52 |
| *** amotoki has quit IRC | 14:53 | |
| *** hw_wutianwei has joined #openstack-ironic | 14:56 | |
| *** d0ugal has quit IRC | 14:57 | |
| jroll | reminder that cirros is super relevant today: | 14:58 |
| jroll | login as 'cirros' user. default password: 'cubswin:)'. use 'sudo' for root. | 14:58 |
| sambetts | jroll: ;) that joke has literally just passed through our team too | 14:59 |
| *** e0ne has joined #openstack-ironic | 14:59 | |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Update iptables rules and services IPs for multinode https://review.openstack.org/368648 | 14:59 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Add devstack setup_vxlan_network() https://review.openstack.org/368647 | 14:59 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Skip some steps for multinode case https://review.openstack.org/368646 | 14:59 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Update ironic node names for multinode case https://review.openstack.org/368645 | 14:59 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Update devstack provision net config for multihost https://review.openstack.org/368644 | 14:59 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Skip db configuration on subnodes https://review.openstack.org/353303 | 14:59 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Testing multinode stuff https://review.openstack.org/368173 | 14:59 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Ignore required_services for multinode topology https://review.openstack.org/352793 | 14:59 |
| sambetts | vsaienk0: this started off as a rework of your patch, but then its changed quite dramtically so I turned it into its own patch https://review.openstack.org/#/c/392959 | 15:00 |
| patchbot | patch 392959 - ironic - Rework Ironic devstack baremetal network simulation | 15:00 |
| vsaienk0 | sorry for the spam, ironic-multinode related devstack-gate chain will be merged soon, so I decided to rebase multinode related patches :) bash experts reviews are welcome :) | 15:00 |
| vsaienk0 | sambetts: I saw your patch, but didn't have time to dig enough | 15:01 |
| vsaienk0 | sambetts: one thing I'm concerned, why not left ironic nodes connected to ovs? ovs is better emulation of hardware switch than linux bridges. | 15:02 |
| sambetts | vsaienk0: I'm planning on making it optional, so we can emulate with linux bridge or OVS, depending on what people have installed | 15:03 |
| vsaienk0 | sambetts: we still may have ironic nodes connected to ovs event when we using Neutron+LB | 15:04 |
| * rloo doesn't get the cirros joke | 15:04 | |
| sambetts | vsaienk0: its possible yes but some people don't install OVS if they are using neutron+lb | 15:04 |
| jroll | rloo: "cubs win :)" | 15:05 |
| sambetts | vsaienk0: so I wanted to provide a no-OVS option | 15:05 |
| vsaienk0 | sambetts: but they using hardware servers not VMs? | 15:05 |
| jroll | rloo: big news of the day, cubs won the world series for the first time in 108(?) years | 15:05 |
| rloo | jroll: oh, chicago cubs. baseball. is it the *final* game today? | 15:05 |
| jroll | rloo: last night was the final game | 15:06 |
| jroll | went into extra innings and everything, such excitement | 15:06 |
| rloo | jroll: OH. so they won. even better! :D | 15:07 |
| JayF | comeback from 3 games back, won in extra innings after a rain delay | 15:07 |
| JayF | a little bit of everything | 15:07 |
| rloo | bet it was really exciting then! | 15:07 |
| jroll | twas :) | 15:08 |
| * rloo happy. cleveland beat toronto blue jays so i'm glad they lost. | 15:08 | |
| jroll | heh | 15:08 |
| *** krtaylor has quit IRC | 15:08 | |
| *** Nisha1 has joined #openstack-ironic | 15:09 | |
| sambetts | vsaienk0: not nessesaraly, I think its possible that if you select neutron+lb it won't install ovs | 15:09 |
| sambetts | vsaienk0: but I havn't been able to confirm yet | 15:09 |
| *** jerrygb_ has joined #openstack-ironic | 15:12 | |
| vsaienk0 | sambetts: openvswitch is a dependency for neutron-openvswitch-agent package http://packages.ubuntu.com/yakkety/neutron-openvswitch-agent so if it is not installed, ovs is not installed also | 15:12 |
| *** jkilpatr_ has joined #openstack-ironic | 15:12 | |
| *** d0ugal has joined #openstack-ironic | 15:12 | |
| *** jkilpatr_ has quit IRC | 15:13 | |
| sambetts | vsaienk0: so if you pick the neutron linux bridge agent then I'd expect that it won't install that package and therefore not install OVS | 15:13 |
| sambetts | vsaienk0: but we always have LB | 15:13 |
| *** jerrygb has quit IRC | 15:13 | |
| vsaienk0 | sambetts: LB is bad emulation of real switch, so if we want to use VMs with Neutron+LB I prefer to left them connected via ovs | 15:14 |
| *** jkilpatr has quit IRC | 15:14 | |
| NobodyCam | Good Morning Ironic'ers | 15:14 |
| vsaienk0 | sambetts: we can add requirement for ovs here: https://github.com/openstack/ironic/blob/master/devstack/files/debs/ironic | 15:15 |
| sambetts | vsaienk0: then we have a hard dependecy on OVS for testing even when we don't need it for flat network simulation? | 15:15 |
| mgould | morning NobodyCam | 15:16 |
| NobodyCam | Good Morning mgould :) | 15:16 |
| *** jkilpatr has joined #openstack-ironic | 15:16 | |
| *** jaosorior has joined #openstack-ironic | 15:17 | |
| vsaienk0 | sambetts: technically it is possible to do not use ovs and plug VMs directly to linux bridge from flat network, but in real world hardware servers always plugged to switch, not matter it is flat network or not. | 15:19 |
| *** rpioso has joined #openstack-ironic | 15:19 | |
| *** anush has quit IRC | 15:20 | |
| vsaienk0 | sambetts: so event with flat case, I think we should left simulation of hardware switch | 15:21 |
| vsaienk0 | *even | 15:21 |
| sambetts | vsaienk0: the linux bridge emulates a dumb l2 switch and the vms are pluged into that, and then a virtual patch cable plugs the virtual switch into neutron | 15:22 |
| vsaienk0 | sambetts: ovs is better emulation of the switch than linuxbridge | 15:23 |
| *** KennethWilke has joined #openstack-ironic | 15:24 | |
| sambetts | OVS has more switch features like built in vlan etc, but for flat networks I don't see much difference | 15:26 |
| sambetts | its even pretty easy to emulate staticly defined vlan networks using it | 15:27 |
| vsaienk0 | sambetts: we should test both flat and vlan, might be vxlan in future, why not keep setup the same for all cases? | 15:27 |
| *** anush has joined #openstack-ironic | 15:28 | |
| *** rbrud has quit IRC | 15:28 | |
| vsaienk0 | sambetts: linuxbridges doesn't support vlans natively, they way how neutron combines LB and tagged subinterfaces looks hacky | 15:29 |
| sambetts | vsaienk0: I use that method for doing vlans with lb in my patch | 15:30 |
| *** e0ne has quit IRC | 15:30 | |
| vsaienk0 | sambetts: that is my concern, LB is bad emulation of the switch, because it doesn't support natively basic functions like vlans | 15:30 |
| vsaienk0 | sambetts: I like the idea to be possible to use Neutron+LB, but I still think that ironic VMs should be connected via ovs. | 15:32 |
| sambetts | vsaienk0: makes sense, I'll edit my patch to use OVS as the default switch emulator and see if it all hooks up | 15:32 |
| *** causten has joined #openstack-ironic | 15:36 | |
| *** hw_wutianwei has quit IRC | 15:37 | |
| *** hw_wutianwei has joined #openstack-ironic | 15:38 | |
| *** d0ugal has quit IRC | 15:39 | |
| *** hw_wutianwei has quit IRC | 15:41 | |
| *** hw_wutianwei has joined #openstack-ironic | 15:41 | |
| *** jtomasek has quit IRC | 15:44 | |
| *** rbudden has quit IRC | 15:46 | |
| *** jkilpatr has quit IRC | 15:47 | |
| *** krtaylor has joined #openstack-ironic | 15:48 | |
| *** krtaylor has quit IRC | 15:53 | |
| *** roeyc has quit IRC | 15:53 | |
| *** jaosorior has quit IRC | 15:54 | |
| *** harlowja_at_home has joined #openstack-ironic | 15:57 | |
| *** jkilpatr has joined #openstack-ironic | 16:00 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: DevStack: Fix standard PXE on Ubuntu Xenial https://review.openstack.org/393353 | 16:00 |
| *** causten has quit IRC | 16:02 | |
| *** rcernin has quit IRC | 16:02 | |
| *** causten has joined #openstack-ironic | 16:03 | |
| *** causten_ has joined #openstack-ironic | 16:05 | |
| *** shobhar_ has joined #openstack-ironic | 16:06 | |
| *** causten has quit IRC | 16:07 | |
| *** jjohnson2 has quit IRC | 16:08 | |
| *** rajinir has joined #openstack-ironic | 16:12 | |
| *** pcaruana has quit IRC | 16:20 | |
| *** mjura_ has quit IRC | 16:21 | |
| *** rcernin has joined #openstack-ironic | 16:25 | |
| *** shobhar_ has quit IRC | 16:31 | |
| *** shobhar_ has joined #openstack-ironic | 16:34 | |
| *** aNuposic has joined #openstack-ironic | 16:38 | |
| *** aNuposic has quit IRC | 16:41 | |
| *** aNuposic has joined #openstack-ironic | 16:41 | |
| *** nicodemos has quit IRC | 16:42 | |
| *** sabeen has quit IRC | 16:49 | |
| *** moshele has quit IRC | 16:50 | |
| *** jtomasek has joined #openstack-ironic | 16:52 | |
| *** Alexey_Abashkin_ has joined #openstack-ironic | 16:56 | |
| *** ohamada has quit IRC | 16:57 | |
| *** pester has quit IRC | 16:57 | |
| *** Alexey_Abashkin has quit IRC | 16:59 | |
| *** harlowja_at_home has quit IRC | 17:00 | |
| *** hw_wutianwei has quit IRC | 17:00 | |
| *** g3ek has quit IRC | 17:01 | |
| *** rbudden has joined #openstack-ironic | 17:02 | |
| *** maratoid_ has joined #openstack-ironic | 17:03 | |
| *** rcernin has quit IRC | 17:03 | |
| *** rcernin has joined #openstack-ironic | 17:03 | |
| *** oberezovskyi has joined #openstack-ironic | 17:07 | |
| *** g3ek has joined #openstack-ironic | 17:07 | |
| *** rcernin has quit IRC | 17:08 | |
| *** nicodemos has joined #openstack-ironic | 17:08 | |
| *** strigazi is now known as strigazi_AFK | 17:16 | |
| milan | good night Ironic! :) | 17:18 |
| *** milan has quit IRC | 17:18 | |
| *** ralonsoh_ is now known as ralonsoh | 17:18 | |
| *** rcernin has joined #openstack-ironic | 17:19 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: DevStack: Fix standard PXE on Ubuntu Xenial https://review.openstack.org/393353 | 17:19 |
| *** e0ne has joined #openstack-ironic | 17:20 | |
| *** rcernin has quit IRC | 17:24 | |
| *** anush has quit IRC | 17:24 | |
| *** daemontool has quit IRC | 17:28 | |
| *** mgould is now known as mgould|afk | 17:31 | |
| *** rbartal has quit IRC | 17:31 | |
| *** dtantsur is now known as dtantsur|afk | 17:32 | |
| *** aarefiev has quit IRC | 17:34 | |
| *** aarefiev has joined #openstack-ironic | 17:34 | |
| *** aarefiev_ has joined #openstack-ironic | 17:35 | |
| *** aarefiev has quit IRC | 17:35 | |
| *** aarefiev has joined #openstack-ironic | 17:35 | |
| *** aarefiev_ has quit IRC | 17:36 | |
| *** oberezovskyi has quit IRC | 17:36 | |
| *** aarefiev_ has joined #openstack-ironic | 17:38 | |
| *** fragatina has quit IRC | 17:39 | |
| *** aarefiev is now known as aarefiev_afk | 17:41 | |
| *** aarefiev_ has quit IRC | 17:42 | |
| *** harlowja has quit IRC | 17:43 | |
| *** harlowja has joined #openstack-ironic | 17:46 | |
| *** srobert has joined #openstack-ironic | 17:46 | |
| *** moshele has joined #openstack-ironic | 17:48 | |
| *** anush has joined #openstack-ironic | 17:52 | |
| *** derekh has quit IRC | 17:56 | |
| *** Guest25209 has quit IRC | 17:57 | |
| lucasagomes | calling it a day, have a great evening all! | 17:59 |
| *** lucasagomes is now known as lucas-afk | 17:59 | |
| JayF | o/ | 18:00 |
| *** ralonsoh has quit IRC | 18:01 | |
| *** portdirect is now known as portdirect_away | 18:02 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: DevStack: Fix standard PXE on Ubuntu Xenial https://review.openstack.org/393353 | 18:03 |
| lucas-afk | (ok last patch-set, now tested (-: ) | 18:03 |
| * lucas-afk goes | 18:03 | |
| *** rcernin has joined #openstack-ironic | 18:05 | |
| *** sdake has joined #openstack-ironic | 18:05 | |
| *** athomas has quit IRC | 18:05 | |
| *** Nisha1 has quit IRC | 18:05 | |
| jcook | JayF https://review.openstack.org/#/c/391415/ | 18:12 |
| patchbot | patch 391415 - nova - Fix hypervisor-stats show invalid information abou... | 18:13 |
| *** dprince has quit IRC | 18:15 | |
| *** baoli_ has quit IRC | 18:15 | |
| *** rbudden has quit IRC | 18:19 | |
| *** rbudden has joined #openstack-ironic | 18:19 | |
| *** Sukhdev has joined #openstack-ironic | 18:20 | |
| *** cdearborn has joined #openstack-ironic | 18:22 | |
| *** sdake has quit IRC | 18:22 | |
| *** [1]cdearborn has quit IRC | 18:23 | |
| *** anush has quit IRC | 18:30 | |
| *** jpich has quit IRC | 18:37 | |
| *** dprince has joined #openstack-ironic | 18:41 | |
| *** sabeen has joined #openstack-ironic | 18:41 | |
| *** moshele has quit IRC | 18:41 | |
| *** jrcloud is now known as zackf | 18:46 | |
| openstackgerrit | Julia Kreger proposed openstack/ironic: Introduce VolumeTarget object https://review.openstack.org/285219 | 18:49 |
| *** rama_y has quit IRC | 18:56 | |
| *** shobhar__ has joined #openstack-ironic | 18:58 | |
| *** shobhar_ has quit IRC | 19:01 | |
| *** strigazi_AFK is now known as strigazi | 19:02 | |
| *** ccamacho has quit IRC | 19:05 | |
| rloo | JayF: what do you think of these new APIs? https://review.openstack.org/#/c/332177/ | 19:07 |
| patchbot | patch 332177 - ironic - Integrate portgroups with ports to support LAG | 19:07 |
| JayF | rloo: I'm in a meeting right now, I can look in around an hour | 19:07 |
| rloo | JayF: I was wondering why we want to add them | 19:07 |
| rloo | JayF: no worries. | 19:07 |
| JayF | ah I think I nkow that answer without reading it, I think those are the pieces needed to virtually combine two physical interfaces into one virtual bonded interface | 19:07 |
| rloo | JayF: we can discuss when you have time. /v1/nodes/<node>/portgroups doesn't seem useful if we already have /v1/portgroups/ ?node_uuide=X | 19:08 |
| JayF | ok; you may be right. I'll look at that patchset later if you want to put your concerns on it I can reply in patch | 19:09 |
| rloo | JayF: will do. am only asking you cuz you reviewed so maybe you know :) | 19:10 |
| *** ChubYann has joined #openstack-ironic | 19:10 | |
| *** e0ne has quit IRC | 19:12 | |
| *** baoli_ has joined #openstack-ironic | 19:18 | |
| rloo | vsaienk0: your rebasing or something seems to mess up the topic. I manually changed it for https://review.openstack.org/#/c/332177 | 19:18 |
| patchbot | patch 332177 - ironic - Integrate portgroups with ports to support LAG | 19:18 |
| *** shobhar__ has quit IRC | 19:19 | |
| *** shobhar__ has joined #openstack-ironic | 19:20 | |
| *** ccamacho has joined #openstack-ironic | 19:22 | |
| sambetts | rloo: I think thats related to the question I brought up in the API session at the summit | 19:28 |
| rloo | sambetts: remind me again | 19:29 |
| *** mmedvede has quit IRC | 19:29 | |
| rloo | sambetts: or just reply to the patch :) | 19:30 |
| sambetts | rloo: right now we have both /nodes/<uuid>/ports and /ports?node=<uuid>, and /ports/<port_id>, and there is inconsistancy with what is possible with the different endpoints, e.g. you get get a port using /nodes/<node_id>/ports/<port_id> but not update it, where as /ports/<port_id> works for getting and putting | 19:32 |
| openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic: WiP Serve iPXE configs from Ironic API https://review.openstack.org/369438 | 19:32 |
| openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic: DNM test enabled dynamic iPXE https://review.openstack.org/391943 | 19:32 |
| openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic: WiP Use Jinja2 for switched config as well https://review.openstack.org/393462 | 19:32 |
| rloo | sambetts: right. so i am wondering why we'd want to make it more messy with a similar pattern with portgroups | 19:33 |
| *** wajdi_ has joined #openstack-ironic | 19:33 | |
| *** jjohnson2 has joined #openstack-ironic | 19:33 | |
| *** strigazi is now known as strigazi_AFK | 19:33 | |
| openstackgerrit | Julia Kreger proposed openstack/ironic: Introduce new RPCs to support volume target operations https://review.openstack.org/285220 | 19:33 |
| *** glonlas has joined #openstack-ironic | 19:34 | |
| sambetts | rloo: ideally I would like ironic to use "resourceful" routing, and by that I refer to the fact that becausce ports are a sub-resource of a node, ports should always be retrieved and updated via a node | 19:34 |
| TheJulia | mariojv: I replied to your question on ^^^ on rev 30. | 19:34 |
| mariojv | thanks TheJulia, will take a look | 19:34 |
| sambetts | rloo: e.g. /ports doesn't exist, only /nodes/<node_id/ports exists | 19:34 |
| rloo | sambetts: OHHH. I preferred the other way, but what you say makes sense. | 19:35 |
| *** strigazi_AFK is now known as strigazi | 19:35 | |
| rloo | sambetts: oh. but how do you get a list of all ports, or does that not make any sense. | 19:35 |
| *** livelace has quit IRC | 19:35 | |
| rloo | if i hate API, does that mean i hate users? i hope not | 19:36 |
| openstackgerrit | Julia Kreger proposed openstack/ironic: Add volume_connectors and volume_targets to task https://review.openstack.org/366196 | 19:37 |
| *** causten_ has quit IRC | 19:38 | |
| sambetts | I don't feel like /ports is particularly useful by itself, I would always point someone in the direction of doing GET /nodes, followed by for node in nodes: GET /node/<node_id>/ports, although I'm not sure how often you need to view all ports without kow what node your looking to find out the list of ports for | 19:39 |
| *** glonlas has quit IRC | 19:40 | |
| rloo | sambetts: sigh. is this a similar thing for portgroups? I just approved the /v1/portgroups endpoint :-( | 19:41 |
| rloo | sambetts: not too late to stop it i think... | 19:41 |
| *** strigazi is now known as strigazi_AFK | 19:41 | |
| sambetts | I think we need to address all the API stuff all at once, so being consistent with the existing API right now is probably more important | 19:41 |
| *** causten has joined #openstack-ironic | 19:42 | |
| rloo | sambetts: hmm. i'd rather not add more mess to existing; it means more to deprecate if we ever do so. | 19:42 |
| sambetts | until the summit I wasn't sure redo the API to that level would be on the cards, so :/ | 19:45 |
| *** baoli_ has quit IRC | 19:45 | |
| *** rama_y has joined #openstack-ironic | 19:46 | |
| rloo | sambetts: sigh. and if we redo, lets do redo it all. so lets add more mess then. | 19:46 |
| *** baoli has joined #openstack-ironic | 19:48 | |
| *** baoli_ has joined #openstack-ironic | 19:51 | |
| *** jerrygb has joined #openstack-ironic | 19:51 | |
| *** baoli has quit IRC | 19:51 | |
| *** maratoid_ has quit IRC | 19:51 | |
| *** jerrygb_ has quit IRC | 19:54 | |
| *** baoli__ has joined #openstack-ironic | 19:55 | |
| *** baoli_ has quit IRC | 19:55 | |
| sambetts | rloo: yeah not ideal but we'll see what we discuss with devananda | 19:57 |
| openstackgerrit | Julia Kreger proposed openstack/ironic: Add storage_interface DB field and object https://review.openstack.org/348005 | 19:57 |
| * sambetts -> food then sleep | 19:57 | |
| sambetts | night all | 19:57 |
| rloo | sambetts: thx! have a great evening | 19:57 |
| *** sambetts is now known as sambetts|afk | 19:58 | |
| sambetts|afk | o/ rloo | 19:58 |
| *** fragatina has joined #openstack-ironic | 20:01 | |
| openstackgerrit | Julia Kreger proposed openstack/ironic: [WIP] Add storage_interface to base driver class https://review.openstack.org/348006 | 20:03 |
| devananda | sambetts|afk: some times, it's very useful to say GET /ports?address=aa.bb.cc | 20:04 |
| *** srobert has quit IRC | 20:05 | |
| devananda | I suppose we could implement that on the /nodes resource instead, though | 20:06 |
| TheJulia | Also, to that point, last time I checked, the ports are not deleted along with a node, they have to be independently deleted with the current api, so until that were to change, we would need /ports to be able to clean up nodes so we can re-add them. | 20:06 |
| devananda | TheJulia: oh really? I thought it cascaded :( | 20:06 |
| TheJulia | devananda: I don't think it does, but that might have changed in the past year | 20:06 |
| *** jkilpatr has quit IRC | 20:09 | |
| rloo | TheJulia, devananda: I am thinking that maybe we just implement the portgroups API stuff similar to ports, and if/clean it all up later? wdyt? | 20:09 |
| rloo | TheJulia, devananda: I don't like it, but I don't think we have a good handle/plan on existing ports/nodes API. | 20:10 |
| TheJulia | I kind of thought was always the plan for portgroups, mirror ports, and evolve later as necessary | 20:13 |
| rloo | TheJulia: Oh, I didn't know that was the plan. the spec is incomplete wrt portgroup APIs (i mean, between spec and the two patches i've looked at so far) | 20:14 |
| rloo | TheJulia: in that case, I'll just continue reviewing and approve the first patch for /v1/portgroups. | 20:15 |
| TheJulia | rloo: I could very well be offbase and have out of date context, but that is what I thought the plan was | 20:16 |
| rloo | TheJulia: what, i was just about to say that you said so! :) I don't see any other 'quick' choice besides that one right now. | 20:16 |
| TheJulia | \o/ consensus! | 20:17 |
| rloo | TheJulia: :D | 20:18 |
| *** dprince has quit IRC | 20:26 | |
| *** khamtamtun has joined #openstack-ironic | 20:26 | |
| *** Goneri has quit IRC | 20:27 | |
| openstackgerrit | Oleksandr Berezovskyi proposed openstack/ironic-python-agent: Add possibility to inspect systems with multiple CPUs https://review.openstack.org/391290 | 20:29 |
| *** jkilpatr has joined #openstack-ironic | 20:33 | |
| *** krtaylor has joined #openstack-ironic | 20:34 | |
| openstackgerrit | Félix Bouliane proposed openstack/ironic: WIP- Add virtualpdu to ironic devstack plugin https://review.openstack.org/388154 | 20:35 |
| openstackgerrit | Félix Bouliane proposed openstack/ironic: DONOTMERGE-change IRONIC_DEPLOY_DRIVER to pxe_snmp https://review.openstack.org/388168 | 20:35 |
| *** hamzy has quit IRC | 20:35 | |
| openstackgerrit | Oleksandr Berezovskyi proposed openstack/ironic-python-agent: Add possibility to inspect systems with multiple CPUs https://review.openstack.org/391290 | 20:42 |
| *** baoli__ has quit IRC | 20:43 | |
| *** baoli has joined #openstack-ironic | 20:43 | |
| *** rama_y has quit IRC | 20:46 | |
| *** mmedvede has joined #openstack-ironic | 20:48 | |
| *** fragatina has quit IRC | 20:49 | |
| *** rbudden_ has joined #openstack-ironic | 20:53 | |
| *** rbudden has quit IRC | 20:54 | |
| *** rbudden_ is now known as rbudden | 20:54 | |
| mrda | Morning Ironic | 20:55 |
| *** khamtamtun has quit IRC | 20:55 | |
| *** rcernin has quit IRC | 20:57 | |
| *** oberezovskyi has joined #openstack-ironic | 21:04 | |
| oberezovskyi | 2016-11-03 20:59:03.868 | [91mE: Failed to fetch http://http.us.debian.org/debian/pool/main/q/qemu/qemu-utils_2.7+dfsg-3~bpo8+1_amd64.deb 404 Not Found [IP: 64.50.236.52 80] | 21:04 |
| patchbot | Error: Missing "]". You may want to quote your arguments with double quotes in order to prevent extra brackets from being evaluated as nested commands. | 21:04 |
| *** jheroux has quit IRC | 21:08 | |
| *** rama_y has joined #openstack-ironic | 21:18 | |
| *** jcoufal has quit IRC | 21:23 | |
| TheJulia | oberezovskyi: Sounds like the mirror you got directed to was missing that file, it downloaded for me | 21:23 |
| *** khamtamtun has joined #openstack-ironic | 21:24 | |
| *** khamtamtun has quit IRC | 21:24 | |
| mariojv | TheJulia: about the comment regarding VolumeTargetNotFound on https://review.openstack.org/#/c/285220/30/ironic/conductor/rpcapi.py - i agree it won't be raised for internal rpc usage unless there's a bug / race somewhere | 21:26 |
| patchbot | patch 285220 - ironic - Introduce new RPCs to support volume target operat... | 21:26 |
| mariojv | TheJulia: but i'm curious about its behavior if an ironic user tries to delete the volume target via API, which hits that rpc call | 21:27 |
| mariojv | specifically L412 here https://review.openstack.org/#/c/285221/9/ironic/api/controllers/v1/volume_target.py | 21:27 |
| patchbot | patch 285221 - ironic - Add REST API for volume target operation | 21:27 |
| mariojv | would that just give back a 5xx error? | 21:27 |
| mariojv | mainly i just wanted a docstring added to the rpc bit so we remember to return a helpful status code to the user if they're trying to delete a non-existent volume target | 21:28 |
| TheJulia | I _think_ so, but the rpcapi interaction is not currently swapped into my brain on that level, as in the level between API service -> rpcapi to know for sure | 21:31 |
| mariojv | i guess that might be handled when you're retrieving the target, before that | 21:32 |
| * TheJulia looks at api code | 21:32 | |
| TheJulia | it _should_ be, but then you would have to have a race for it to occur | 21:32 |
| mariojv | right | 21:32 |
| mariojv | if there's something internal to ironic that would delete, at the same time a user tries to delete | 21:33 |
| mariojv | i haven't reviewed all the code for this yet so not sure if that's possible | 21:33 |
| *** shobhar__ has quit IRC | 21:34 | |
| *** khamtamtun has joined #openstack-ironic | 21:34 | |
| TheJulia | mariojv: the key would be to lock on the conductor :) | 21:34 |
| mariojv | ah, then you couldn't even get the node, or conductor couldn't get the lock before trying to delete the volume | 21:35 |
| mariojv | so, this shouldn't be an issue | 21:35 |
| TheJulia | you could get it, just not the lock | 21:35 |
| mariojv | right, so no destructive operations would be allowed | 21:35 |
| TheJulia | yup | 21:35 |
| *** e0ne has joined #openstack-ironic | 21:35 | |
| mariojv | assuming this works in the same order as the api code does (get volume target, get node, destroy volume target) | 21:36 |
| * mariojv removes -1 | 21:36 | |
| mariojv | thanks | 21:36 |
| TheJulia | Yeah, the API code needs work at this point | 21:38 |
| TheJulia | But we have a volunteer to work on that | 21:39 |
| oberezovskyi | TheJulia: the main problem that it error from gate job at IPA | 21:39 |
| TheJulia | oberezovskyi: I hate to suggest this, but it is likely transitory | 21:39 |
| TheJulia | :( | 21:40 |
| * TheJulia thought infra setup mirrors though | 21:40 | |
| JayF | TheJulia: we don't utilize those mirrors for the IPA -src builds | 21:40 |
| jroll | we don't use the mirrors for the docker build | 21:40 |
| TheJulia | right :( | 21:40 |
| JayF | jroll: not just for the docker build; also for the tinyipa build | 21:40 |
| oberezovskyi | It fails at devstack installation | 21:40 |
| jroll | well, yeah | 21:40 |
| jroll | but that was debian so I assumed docker :D | 21:40 |
| *** baoli has quit IRC | 21:41 | |
| *** sabeen has quit IRC | 21:43 | |
| *** khamtamtun has quit IRC | 21:45 | |
| openstackgerrit | Merged openstack/ironic: Add PortGroups API https://review.openstack.org/347549 | 21:46 |
| rloo | yay, one done, more to review ^^. which reminds me, i forgot to review the next one. | 21:47 |
| openstackgerrit | Merged openstack/python-ironicclient: Updated from global requirements https://review.openstack.org/392820 | 21:48 |
| openstackgerrit | Merged openstack/python-ironicclient: Use function import_versioned_module from oslo.utils https://review.openstack.org/392395 | 21:54 |
| openstackgerrit | Merged openstack/ironic-python-agent: Updated from global requirements https://review.openstack.org/389896 | 21:56 |
| *** harlowja has quit IRC | 21:56 | |
| *** david-lyle has joined #openstack-ironic | 22:03 | |
| *** shobhar__ has joined #openstack-ironic | 22:07 | |
| *** absubram has quit IRC | 22:19 | |
| *** baoli has joined #openstack-ironic | 22:21 | |
| *** sabeen has joined #openstack-ironic | 22:27 | |
| *** jerrygb_ has joined #openstack-ironic | 22:29 | |
| *** causten has quit IRC | 22:30 | |
| *** rbudden has quit IRC | 22:31 | |
| *** wajdi_ has quit IRC | 22:31 | |
| *** jerrygb has quit IRC | 22:32 | |
| openstackgerrit | Merged openstack/ironic: Updated from global requirements https://review.openstack.org/392730 | 22:33 |
| *** devananda is now known as devananda|away | 22:33 | |
| *** baoli has quit IRC | 22:38 | |
| *** baoli has joined #openstack-ironic | 22:41 | |
| *** rbrud has joined #openstack-ironic | 22:42 | |
| *** baoli_ has joined #openstack-ironic | 22:44 | |
| *** baoli has quit IRC | 22:46 | |
| *** baoli has joined #openstack-ironic | 22:51 | |
| *** baoli_ has quit IRC | 22:53 | |
| *** baoli has quit IRC | 22:57 | |
| *** bfournie1 has left #openstack-ironic | 23:02 | |
| *** rloo has quit IRC | 23:04 | |
| openstackgerrit | Merged openstack/ironic: Do not hide unexpected exceptions in inspection code https://review.openstack.org/385510 | 23:09 |
| *** oberezovskyi has quit IRC | 23:11 | |
| *** awiddersheim has joined #openstack-ironic | 23:12 | |
| *** bfournie has joined #openstack-ironic | 23:33 | |
| *** e0ne has quit IRC | 23:33 | |
| *** mtanino has quit IRC | 23:39 | |
| *** yuanying has joined #openstack-ironic | 23:41 | |
| *** hw_wutianwei has joined #openstack-ironic | 23:57 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!