| JayF | I mean, I was being more restrictive about who I gave them out to at first, but I have plenty left so spreading the good pixie boots joy around the world | 00:00 |
|---|---|---|
| mrda | JayF, tonyb: Next time I mneet up with the local AU team (we're all geographically dispersed) I will distribute stickers to Ironic contributors as promised :) | 00:01 |
| jlvillal | JayF: /me would like a sticker at some point. If still available | 00:11 |
| JayF | jlvillal: toss me an email with an address | 00:12 |
| * jlvillal feels sad the he is going to get a new laptop and his stickers will be gone. But happy to go from 4GB to 16GB of memory. | 00:12 | |
| jlvillal | JayF: Will do | 00:12 |
| *** sacharya has joined #openstack-ironic | 00:17 | |
| *** Jabadia has quit IRC | 00:18 | |
| *** r-daneel has quit IRC | 00:19 | |
| TheJulia | JayF: btw, thank you for the appearance of pixie boots stickers in my mailbox :) | 00:19 |
| JayF | np :) figured I'd go to Summer for that so it'd be a surprise :) | 00:20 |
| TheJulia | It was a nice surprise when I got home from Chicago :) | 00:22 |
| *** sacharya has quit IRC | 00:23 | |
| *** piet has quit IRC | 00:27 | |
| *** rama_y has quit IRC | 00:27 | |
| *** piet has joined #openstack-ironic | 00:27 | |
| openstackgerrit | Jay Faulkner proposed openstack/ironic-specs: Implement Keystone Policy Support https://review.openstack.org/327437 | 00:29 |
| JayF | devananda: ^ about to WIP it as incomplete, as I got pretty deep down a couple of rabbitholes today, but if you want to take a look and/or contribute feel free | 00:29 |
| *** mtanino has quit IRC | 00:33 | |
| *** hoangcx has joined #openstack-ironic | 00:37 | |
| *** sdake has quit IRC | 00:42 | |
| *** piet has quit IRC | 00:48 | |
| JayF | I made a thing | 00:52 |
| JayF | http://graphite.openstack.org/dashboard/#ironic-job-duration | 00:52 |
| *** Sukhdev has quit IRC | 00:55 | |
| *** phuongnh has joined #openstack-ironic | 00:55 | |
| *** baoli has joined #openstack-ironic | 01:03 | |
| gumbyblake | hi, need some help with overcloud deploy. Getting error 500 no host available, but my physical nodes register fine & have heaps of resource | 01:36 |
| *** mtanino has joined #openstack-ironic | 01:49 | |
| *** ccarmack has quit IRC | 01:49 | |
| *** thrash is now known as thrash|g0ne | 01:57 | |
| *** sacharya has joined #openstack-ironic | 01:59 | |
| *** PollyZ has joined #openstack-ironic | 02:01 | |
| *** sacharya has quit IRC | 02:07 | |
| *** sacharya has joined #openstack-ironic | 02:08 | |
| *** vishwanathj has joined #openstack-ironic | 02:08 | |
| *** vishwanathj has quit IRC | 02:09 | |
| *** vishwanathj has joined #openstack-ironic | 02:09 | |
| *** ayoung has quit IRC | 02:17 | |
| *** alex_xu has quit IRC | 02:21 | |
| *** PollyZ has quit IRC | 02:21 | |
| *** alex_xu has joined #openstack-ironic | 02:23 | |
| *** ayoung has joined #openstack-ironic | 02:25 | |
| *** rama_y has joined #openstack-ironic | 02:26 | |
| *** ayoung has quit IRC | 02:29 | |
| *** rbudden has joined #openstack-ironic | 02:49 | |
| *** yuanying has quit IRC | 02:50 | |
| gumbyblake | any suggestions on how can troubleshoot failure to poweron node during deploy | 02:57 |
| *** ijw has quit IRC | 03:05 | |
| *** piet has joined #openstack-ironic | 03:09 | |
| *** sdake has joined #openstack-ironic | 03:10 | |
| *** rloo has quit IRC | 03:18 | |
| *** piet has quit IRC | 03:20 | |
| *** piet has joined #openstack-ironic | 03:25 | |
| *** sdake has quit IRC | 03:26 | |
| *** ChrisAusten has joined #openstack-ironic | 03:27 | |
| *** amotoki has quit IRC | 03:40 | |
| *** piet has quit IRC | 03:41 | |
| *** links has joined #openstack-ironic | 03:46 | |
| *** rama_y has quit IRC | 03:48 | |
| *** yuanying has joined #openstack-ironic | 03:48 | |
| *** rbudden has quit IRC | 03:48 | |
| *** amotoki has joined #openstack-ironic | 03:54 | |
| *** sukhdev has joined #openstack-ironic | 03:58 | |
| *** amotoki has quit IRC | 04:01 | |
| *** amotoki has joined #openstack-ironic | 04:05 | |
| *** kalpase has joined #openstack-ironic | 04:08 | |
| *** baoli has quit IRC | 04:17 | |
| *** vishwanathj has quit IRC | 04:25 | |
| *** Sukhdev_ has joined #openstack-ironic | 04:26 | |
| *** vishwanathj has joined #openstack-ironic | 04:26 | |
| *** amotoki has quit IRC | 04:31 | |
| *** amotoki has joined #openstack-ironic | 04:31 | |
| *** amotoki has quit IRC | 04:33 | |
| *** vmud213 has joined #openstack-ironic | 04:38 | |
| *** amotoki has joined #openstack-ironic | 04:43 | |
| *** greghaynes has quit IRC | 04:48 | |
| *** rama_y has joined #openstack-ironic | 04:51 | |
| *** baremetal27 has joined #openstack-ironic | 04:54 | |
| *** harlowja_at_home has joined #openstack-ironic | 04:56 | |
| *** harlowja_at_home has quit IRC | 04:56 | |
| *** harlowja_at_home has joined #openstack-ironic | 04:56 | |
| *** greghaynes has joined #openstack-ironic | 04:59 | |
| *** harlowja_at_home has quit IRC | 05:08 | |
| *** rcernin|off has joined #openstack-ironic | 05:15 | |
| *** Sukhdev_ has quit IRC | 05:16 | |
| *** vdrok has quit IRC | 05:23 | |
| *** mgkwill has quit IRC | 05:23 | |
| *** francoblanco has quit IRC | 05:23 | |
| *** sukhdev has quit IRC | 05:23 | |
| *** sergek has quit IRC | 05:23 | |
| *** betherly has quit IRC | 05:23 | |
| *** evgenyl has quit IRC | 05:24 | |
| *** rajinir has quit IRC | 05:24 | |
| *** romcheg has quit IRC | 05:24 | |
| *** izaakk has quit IRC | 05:24 | |
| *** serverascode has quit IRC | 05:25 | |
| *** cppforlife_ has quit IRC | 05:25 | |
| *** sergek has joined #openstack-ironic | 05:25 | |
| *** yonglihe has quit IRC | 05:25 | |
| *** NobodyCam has quit IRC | 05:25 | |
| *** tdurakov has quit IRC | 05:25 | |
| *** kozhukalov has quit IRC | 05:25 | |
| *** boris-42 has quit IRC | 05:25 | |
| *** fyxim has quit IRC | 05:25 | |
| *** ChrisMead has quit IRC | 05:25 | |
| *** BadCub has quit IRC | 05:25 | |
| *** agordeev has quit IRC | 05:25 | |
| *** patrickeast has quit IRC | 05:25 | |
| *** zhenguo_ has quit IRC | 05:25 | |
| *** rama_y has quit IRC | 05:26 | |
| *** rama_y has joined #openstack-ironic | 05:26 | |
| *** rcernin|off has quit IRC | 05:26 | |
| *** tdurakov has joined #openstack-ironic | 05:28 | |
| *** NobodyCam has joined #openstack-ironic | 05:28 | |
| *** izaakk has joined #openstack-ironic | 05:28 | |
| *** yonglihe has joined #openstack-ironic | 05:28 | |
| *** cppforlife_ has joined #openstack-ironic | 05:28 | |
| *** fyxim has joined #openstack-ironic | 05:29 | |
| *** agordeev has joined #openstack-ironic | 05:29 | |
| *** moshele has joined #openstack-ironic | 05:30 | |
| *** evgenyl has joined #openstack-ironic | 05:30 | |
| *** vdrok has joined #openstack-ironic | 05:30 | |
| *** patrickeast has joined #openstack-ironic | 05:30 | |
| *** kozhukalov has joined #openstack-ironic | 05:30 | |
| *** zhenguo_ has joined #openstack-ironic | 05:30 | |
| *** ijw has joined #openstack-ironic | 05:32 | |
| *** izaakk has quit IRC | 05:33 | |
| *** yonglihe has quit IRC | 05:33 | |
| *** BadCub has joined #openstack-ironic | 05:33 | |
| *** rama_y has quit IRC | 05:33 | |
| *** francoblanco has joined #openstack-ironic | 05:33 | |
| *** rajinir has joined #openstack-ironic | 05:33 | |
| *** rama_y has joined #openstack-ironic | 05:34 | |
| *** boris-42 has joined #openstack-ironic | 05:34 | |
| *** serverascode has joined #openstack-ironic | 05:34 | |
| *** ChrisMead has joined #openstack-ironic | 05:34 | |
| *** mgkwill has joined #openstack-ironic | 05:34 | |
| *** izaakk has joined #openstack-ironic | 05:35 | |
| *** yonglihe has joined #openstack-ironic | 05:35 | |
| *** romcheg has joined #openstack-ironic | 05:35 | |
| *** sukhdev has joined #openstack-ironic | 05:35 | |
| *** betherly has joined #openstack-ironic | 05:37 | |
| *** rama_y has quit IRC | 05:48 | |
| *** rama_y has joined #openstack-ironic | 05:49 | |
| *** ChrisAusten has quit IRC | 05:53 | |
| *** rama_y has quit IRC | 05:53 | |
| *** rama_y has joined #openstack-ironic | 05:54 | |
| *** Nisha_away has joined #openstack-ironic | 05:59 | |
| *** rbartal has joined #openstack-ironic | 06:00 | |
| *** rama_y has quit IRC | 06:03 | |
| *** yolanda has joined #openstack-ironic | 06:05 | |
| *** izaakk has quit IRC | 06:08 | |
| *** betherly has quit IRC | 06:08 | |
| *** sukhdev has quit IRC | 06:08 | |
| *** izaakk has joined #openstack-ironic | 06:08 | |
| *** yolanda has quit IRC | 06:10 | |
| *** yolanda has joined #openstack-ironic | 06:10 | |
| *** sukhdev has joined #openstack-ironic | 06:11 | |
| *** betherly has joined #openstack-ironic | 06:12 | |
| *** sukhdev has quit IRC | 06:13 | |
| baremetal27 | no valid host was found | 06:14 |
| *** rcernin|off has joined #openstack-ironic | 06:15 | |
| *** yolanda has quit IRC | 06:16 | |
| *** yolanda has joined #openstack-ironic | 06:18 | |
| *** rcernin|off is now known as rcernin | 06:21 | |
| *** baoli has joined #openstack-ironic | 06:29 | |
| *** sacharya has quit IRC | 06:31 | |
| *** baoli has quit IRC | 06:33 | |
| openstackgerrit | Satoru Moriya proposed openstack/ironic-specs: Volume connection information for Ironic nodes https://review.openstack.org/200496 | 06:41 |
| openstackgerrit | Nisha Agarwal proposed openstack/ironic: Add support for building ISO for deploy ramdisk https://review.openstack.org/264590 | 06:42 |
| *** ChubYann has quit IRC | 06:44 | |
| openstackgerrit | Nisha Agarwal proposed openstack/ironic: Config variable to configure [glance] section https://review.openstack.org/266803 | 06:47 |
| *** gumbyblake has quit IRC | 06:47 | |
| *** hoangcx has quit IRC | 06:50 | |
| *** itamarl has joined #openstack-ironic | 06:51 | |
| *** hoangcx has joined #openstack-ironic | 06:53 | |
| *** baremetal27 has quit IRC | 06:58 | |
| pas-ha | morning Ironicers | 06:59 |
| vsaienko | Morning Ironic'ers | 06:59 |
| *** mjura has joined #openstack-ironic | 07:01 | |
| vsaienko | o/ pas-ha | 07:01 |
| *** ijw has quit IRC | 07:07 | |
| *** pcaruana has joined #openstack-ironic | 07:09 | |
| openstackgerrit | Anton Arefiev proposed openstack/ironic-inspector: WIP: Fix tempest https://review.openstack.org/327492 | 07:11 |
| *** aarefiev_ has joined #openstack-ironic | 07:12 | |
| *** mtanino has quit IRC | 07:15 | |
| *** dtantsur|afk is now known as dtantsur | 07:18 | |
| dtantsur | Morning Ironic | 07:19 |
| aarefiev_ | morning! | 07:24 |
| aarefiev_ | morning dtantsur, I'm trying to figure out why we return 200 on rule import | 07:25 |
| dtantsur | aarefiev_, morning, probably we just missed it? | 07:26 |
| aarefiev_ | dtantsur: more interesting question how tests worked before ( | 07:27 |
| dtantsur | yeah... maybe tempest didn't check it previously? | 07:28 |
| *** mjura has quit IRC | 07:28 | |
| *** hoangcx has quit IRC | 07:30 | |
| *** mjura has joined #openstack-ironic | 07:31 | |
| *** sacharya has joined #openstack-ironic | 07:31 | |
| *** stendulker has joined #openstack-ironic | 07:32 | |
| aarefiev_ | dtantsur: look like it didn't work, and was fixed here https://github.com/openstack/tempest/commit/ee05cf7fe897577ec40fae18a14f04e219a5e6be | 07:32 |
| dtantsur | hmm, yeah | 07:32 |
| dtantsur | I wonder if we need to fix our API... | 07:32 |
| aarefiev_ | dtantsur: I've started work on this https://review.openstack.org/#/c/327492/ | 07:33 |
| dtantsur | thanks! | 07:36 |
| *** sacharya has quit IRC | 07:37 | |
| *** vishwana_ has joined #openstack-ironic | 07:39 | |
| *** vishwanathj has quit IRC | 07:39 | |
| *** jtomasek_ has joined #openstack-ironic | 07:39 | |
| *** ifarkas has joined #openstack-ironic | 07:42 | |
| vsaienko | morning dtantsur, aarefiev_ | 07:44 |
| aarefiev_ | morning vsaienko | 07:45 |
| dtantsur | morning vsaienko | 07:45 |
| *** hoonetorg has quit IRC | 07:47 | |
| *** vishwana_ has quit IRC | 07:51 | |
| *** tesseract has joined #openstack-ironic | 07:56 | |
| *** hoonetorg has joined #openstack-ironic | 07:58 | |
| *** zzzeek has quit IRC | 08:00 | |
| *** snecklifter has joined #openstack-ironic | 08:00 | |
| *** zzzeek has joined #openstack-ironic | 08:00 | |
| snecklifter | Hello. I have compute nodes with 2 x HDDs but no RAID controller. Can I configure software RAID with ironic? | 08:01 |
| *** openstackgerrit has quit IRC | 08:03 | |
| *** openstackgerrit has joined #openstack-ironic | 08:03 | |
| *** jtomasek_ has quit IRC | 08:05 | |
| *** ohamada has joined #openstack-ironic | 08:06 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/ironic: [WIP] Testing latest u-c https://review.openstack.org/318440 | 08:11 |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/ironic: [WIP] Testing latest u-c https://review.openstack.org/318440 | 08:11 |
| *** electrofelix has joined #openstack-ironic | 08:15 | |
| alineb | morning all! | 08:15 |
| dtantsur | morning alineb | 08:19 |
| *** Nisha_away has quit IRC | 08:19 | |
| dtantsur | snecklifter, hi. ironic does not have software RAID support out of box, sorry | 08:20 |
| *** dtantsur is now known as dtantsur|lunch | 08:20 | |
| *** lucas-dinner is now known as lucasagomes | 08:20 | |
| lucasagomes | morning all | 08:21 |
| dtantsur|lunch | morning lucasagomes | 08:21 |
| snecklifter | dtantsur: thanks for the response! Any idea on how to get it working? Customize the deployment image? | 08:21 |
| dtantsur|lunch | snecklifter, yes, I think your only bet here is to customize IPA | 08:21 |
| *** alex_xu has quit IRC | 08:22 | |
| dtantsur|lunch | sorry, need to get some food :) | 08:22 |
| snecklifter | no problem, thanks again! | 08:22 |
| dtantsur|lunch | lucasagomes, plz review this code removal party: https://review.openstack.org/327083 :) | 08:23 |
| * dtantsur|lunch really goes for lunch now | 08:23 | |
| *** alex_xu has joined #openstack-ironic | 08:25 | |
| *** daemontool has joined #openstack-ironic | 08:25 | |
| milan | #pixiesay Morning Ironic -m flexing | 08:27 |
| PixieBoots | ᕙʕ⇀ᴥ⇀ʔᕗ: Morning Ironic | 08:27 |
| milan | morning dtantsur|lunch lucasagomes vsaienko aarefiev vdrok and all :) | 08:28 |
| lucasagomes | dtantsur|lunch, hah ack | 08:31 |
| *** Nisha_away has joined #openstack-ironic | 08:38 | |
| *** jtomasek_ has joined #openstack-ironic | 08:38 | |
| *** Romanenko_K has joined #openstack-ironic | 08:38 | |
| *** kromanenko__ has joined #openstack-ironic | 08:38 | |
| *** kromanenko_ has joined #openstack-ironic | 08:38 | |
| *** vishwanathj has joined #openstack-ironic | 08:50 | |
| *** vishwanathj has quit IRC | 08:54 | |
| *** phuongnh has quit IRC | 08:56 | |
| *** jtomasek_ has quit IRC | 09:03 | |
| *** e0ne has joined #openstack-ironic | 09:04 | |
| *** athomas has joined #openstack-ironic | 09:10 | |
| *** SiRiuS__ has joined #openstack-ironic | 09:10 | |
| *** sambetts|afk is now known as sambetts | 09:12 | |
| sambetts | Morning all | 09:13 |
| openstackgerrit | Merged openstack/ironic: Updated from global requirements https://review.openstack.org/327367 | 09:13 |
| openstackgerrit | Merged openstack/ironic: Fix nits related to Ports api-ref https://review.openstack.org/326978 | 09:15 |
| *** mjura has quit IRC | 09:16 | |
| aarefiev | morning lucasagomes, milan, sambetts | 09:17 |
| milan | sambetts, morning :) | 09:17 |
| sambetts | o/ aarefiev, milan | 09:17 |
| lucasagomes | aarefiev, milan sambetts morning :-) | 09:20 |
| sambetts | o/ lucasagomes | 09:20 |
| milan | #pixiesay morning lucasagomes! | 09:21 |
| PixieBoots | ʕ•͡ᴥ•ʔ: morning lucasagomes! | 09:21 |
| milan | :D | 09:21 |
| *** SiRiuS__ has quit IRC | 09:21 | |
| *** SiRiuS__ has joined #openstack-ironic | 09:22 | |
| *** mjura has joined #openstack-ironic | 09:24 | |
| *** mbound has joined #openstack-ironic | 09:27 | |
| *** dmk0202 has joined #openstack-ironic | 09:28 | |
| *** vishwanathj has joined #openstack-ironic | 09:29 | |
| *** moshele has quit IRC | 09:30 | |
| *** derekh has joined #openstack-ironic | 09:30 | |
| *** sacharya has joined #openstack-ironic | 09:33 | |
| *** sacharya has quit IRC | 09:37 | |
| *** Nisha_away has quit IRC | 09:39 | |
| *** amotoki has quit IRC | 09:46 | |
| *** vishwanathj has quit IRC | 09:52 | |
| *** vishwanathj has joined #openstack-ironic | 09:53 | |
| lucasagomes | #pixiesay -m meh | 09:55 |
| PixieBoots | ¯\_(ツ)_/¯ | 09:55 |
| lucasagomes | ^ me when pep8 complains: "'Connection._add_nodes_filters' is too complex" | 09:56 |
| openstackgerrit | Aline Bousquet proposed openstack/python-ironicclient: Fix error returned by ironic --json node-validate https://review.openstack.org/303964 | 09:57 |
| *** hoangcx has joined #openstack-ironic | 10:01 | |
| *** mgould|afk is now known as mgould | 10:03 | |
| vdrok | morning ironic, milan lucasagomes and dtantsur|lunch :) | 10:06 |
| openstackgerrit | Sam Betts proposed openstack/ironic-inspector: Add generic LLDP processing hook https://review.openstack.org/321082 | 10:06 |
| vdrok | and good morning sambetts | 10:06 |
| sambetts | lucasagomes: haha, I hate that pep8 error... | 10:07 |
| milan | lucasagomes, :D lovely | 10:07 |
| milan | vdrok, morning :) | 10:07 |
| sambetts | lucasagomes: I thought we abolished that rule from our tox.ini | 10:07 |
| mgould | morning vdrok milan lucasagomes dtantsur|lunch sambetts and everyone else | 10:08 |
| milan | mgould, hi :) | 10:08 |
| vdrok | morning mgould | 10:08 |
| sambetts | o/ mgould | 10:09 |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/ironic: [WIP] Testing latest u-c https://review.openstack.org/318440 | 10:10 |
| *** piet has joined #openstack-ironic | 10:19 | |
| *** stendulker has quit IRC | 10:44 | |
| *** piet has quit IRC | 10:44 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: DevStack: Configure port setup delay for IPMI + Virtual https://review.openstack.org/327589 | 10:52 |
| *** moshele has joined #openstack-ironic | 10:52 | |
| *** dtantsur|lunch is now known as dtantsur | 10:54 | |
| dtantsur | morning sambetts, milan, vdrok, mgould | 10:54 |
| milan | hey dtantsur :) | 10:54 |
| *** piet has joined #openstack-ironic | 10:55 | |
| *** vishwanathj has quit IRC | 11:00 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic-inspector: Add an option to not fail when root device is not supplied https://review.openstack.org/326620 | 11:01 |
| *** vishwanathj has joined #openstack-ironic | 11:02 | |
| *** piet has quit IRC | 11:04 | |
| *** hoangcx has quit IRC | 11:11 | |
| openstackgerrit | Anton Arefiev proposed openstack/ironic-inspector: Fix response code for rule creating API https://review.openstack.org/327492 | 11:11 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic-specs: Boot management for in-band inspection https://review.openstack.org/305864 | 11:23 |
| *** lucasagomes is now known as lucas-hungry | 11:32 | |
| snecklifter | dtantsur: do you have any pointers on how to configure a custom IPA to deploy with LVM/software raid? Struggling to find any documentation regarding this. | 11:42 |
| openstackgerrit | Yuiko Takada Mori proposed openstack/ironic: Fix parameter order of create-node.sh https://review.openstack.org/325732 | 11:42 |
| dtantsur | snecklifter, there are several approaches you can take, depending on whether you want it to be built in advance (easier) or during deployment (much harder) | 11:43 |
| snecklifter | dtantsur: thanks, built in advance is fine | 11:43 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic: Remove support for the old ramdisk (DIB deploy-ironic element) https://review.openstack.org/327083 | 11:44 |
| dtantsur | snecklifter, then you can create a custom hardware manager like in: http://docs.openstack.org/developer/ironic-python-agent/#hardware-managers | 11:44 |
| dtantsur | snecklifter, this is a good example, this is how HPE folks implement their RAID: https://github.com/openstack/proliantutils/blob/master/proliantutils/ipa_hw_manager/hardware_manager.py | 11:45 |
| *** baoli has joined #openstack-ironic | 11:45 | |
| dtantsur | snecklifter, then you build your IPA image with this hw manager and follow http://docs.openstack.org/developer/ironic/deploy/raid.html | 11:45 |
| snecklifter | dtantsur: thanks, and this would work for software raid with no hardware controller? | 11:46 |
| *** baoli_ has joined #openstack-ironic | 11:46 | |
| dtantsur | snecklifter, depending on how you implement this hardware manager :) note that I personally would like to have software RAID upstream | 11:46 |
| dtantsur | snecklifter, so if you could in parallel file an RFE upstream and work through the spec process to get it landed in IPA, that would be beneficial for everyone | 11:47 |
| snecklifter | dtantsur: yes, so would I, I dont think it fits the hardware manager/custom IPA model | 11:47 |
| dtantsur | if you're fine with a bit longer process - awesome | 11:47 |
| snecklifter | dtantsur: I'll do that now, I have not done one before so ok to add you as a reference? | 11:48 |
| snecklifter | I need this for production deployment sadly but may need it again the future | 11:48 |
| dtantsur | snecklifter, the change you will make upstream is still similar: you will extend the default hardware manager to have these clean steps | 11:48 |
| snecklifter | so happy to file RFE if you are happy to answer the odd question about the spec | 11:48 |
| dtantsur | snecklifter, surely! | 11:48 |
| snecklifter | ok, thanks very much for responding, leave it with me | 11:49 |
| dtantsur | snecklifter, this is our RFE guide: http://docs.openstack.org/developer/ironic/dev/code-contribution-guide.html#adding-new-features | 11:49 |
| snecklifter | dtantsur: ah great. that is helpful | 11:49 |
| *** baoli has quit IRC | 11:49 | |
| openstackgerrit | Szymon Borkowski proposed openstack/ironic-python-agent: Add PCI devices collector to inspector https://review.openstack.org/326318 | 11:50 |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic: Add node provsioning methods to the tempest plugin https://review.openstack.org/327238 | 11:51 |
| dtantsur | TheJulia, morning (when you're here ofc). the folks we discussed yesterday came up with an upstream bug/rfe: https://bugs.launchpad.net/ironic-python-agent/+bug/1590606 | 11:52 |
| openstack | Launchpad bug 1590606 in ironic-python-agent "Introspection fails with - missing: ['local_gb']' in case of remote (via bios) iscsi " [Undecided,Incomplete] - Assigned to Yossi Ovadia (jabadia) | 11:52 |
| jroll | morning everyone | 11:52 |
| sambetts | Morning jroll | 11:53 |
| jroll | dtantsur: mind if I -2 your bash ramdisk removal until grenade is in the check queue? | 11:53 |
| dtantsur | morning jroll. I suspect you want to see 1000 lines of code going away from ironic, don't you? ;) | 11:53 |
| dtantsur | jroll, sure-sure | 11:53 |
| jroll | I do, looking already :) | 11:53 |
| jroll | cool, I'll review first since you have a wip anyway | 11:53 |
| openstackgerrit | Szymon Borkowski proposed openstack/ironic-python-agent: Add PCI devices collector to inspector https://review.openstack.org/326318 | 11:53 |
| jroll | btw, I don't think we test ironic/drivers/*.py | 11:53 |
| dtantsur | I don't expect it to break upgrades, but it's always good to check | 11:53 |
| jroll | ditto | 11:54 |
| vdrok | morning jroll | 11:54 |
| dtantsur | jroll, it's no longer wip, so feel free to -2 now | 11:54 |
| jroll | aha | 11:54 |
| dtantsur | +1 for "PLEASE DO REVIEW THIS ANYWAY" :) | 11:55 |
| dtantsur | would be awesome if we don't land too many patches conflicting with it as well | 11:55 |
| vdrok | jroll: re tempest scenarios without nova, is the intention to disable nova in the gate jobs running these scenarios? | 11:55 |
| jroll | vdrok: yeah | 11:56 |
| *** sdake has joined #openstack-ironic | 11:56 | |
| *** nico-afk is now known as nicodemos | 11:56 | |
| * dtantsur would love the majority of tests to use only ironic (e.g. bifrost), and then one big test with nova,neutron,swift,etc | 11:56 | |
| jroll | vdrok: I'm working on writing up a formal plan right now, if you want to mostly wait until I post that | 11:56 |
| * jroll doesn't think we need bifrost for standalone tests :) | 11:56 | |
| vdrok | jroll: cool, I'll just make sure the old things that I have work, and stop on that for now :) | 11:57 |
| jroll | I think we will need glance and neutron for most of them, though | 11:57 |
| jroll | vdrok++ thank you | 11:57 |
| vdrok | yes, at least neutron | 11:57 |
| snecklifter | dtantsur: https://bugs.launchpad.net/ironic/+bug/1590749 <-- anything else you need adding in here? | 12:02 |
| openstack | Launchpad bug 1590749 in Ironic "RFE: LVM/Software RAID support in ironic-python-agent" [Undecided,New] | 12:02 |
| dtantsur | jroll, wdyt https://bugs.launchpad.net/python-ironicclient/+bug/1590752 ? | 12:03 |
| openstack | Launchpad bug 1590752 in python-ironicclient "[RFE] --wait flag for node-set-provision-state" [Wishlist,Confirmed] | 12:03 |
| dtantsur | snecklifter, looking | 12:03 |
| dtantsur | snecklifter, I think you should mention that you're going implement it as a clean step in the IPA default hardware manager | 12:04 |
| dtantsur | snecklifter, next, the problem we've faced the last time we talked about it as the following: we wipe the hard drive every time after deployment. | 12:04 |
| dtantsur | snecklifter, so maybe we need to explicitly state that this will be an automated clean step running after wiping hard drive | 12:04 |
| dtantsur | jroll, JayF, re soft raid ^^^ | 12:05 |
| jroll | ++ | 12:05 |
| TheJulia | Good morning | 12:05 |
| *** kalpase has left #openstack-ironic | 12:06 | |
| snecklifter | dtantsur: sure, I'd assume and expect that it would be wiped on each deployment | 12:06 |
| jroll | dtantsur: replied on your rfe, I like it | 12:07 |
| snecklifter | dtantsur: ok, have updated | 12:07 |
| *** trown|outtypewww is now known as trown | 12:08 | |
| dtantsur | jroll, my primary use case for delay_function are things like eventlet (when people are not monkey-patching world) | 12:09 |
| dtantsur | jroll, so that people can use a greenlet-friendly sleep | 12:09 |
| jroll | dtantsur: ah, that also makes sense | 12:10 |
| dtantsur | jroll, +1 for poll_interval, will update | 12:10 |
| jroll | I do like the thought of using it in nova, where the delay_function would be something that logs and sleeps | 12:10 |
| jroll | (if we log today) | 12:10 |
| * jroll checks | 12:10 | |
| jroll | oh yeah, that would be useful https://github.com/openstack/nova/blob/master/nova/virt/ironic/driver.py#L379 | 12:11 |
| jroll | with the instance.refresh() | 12:11 |
| dtantsur | jroll, updated the RFE | 12:12 |
| jroll | dtantsur: ++ | 12:13 |
| dtantsur | jroll, now, the soft RAID stuff. I suspect it needs a spec, wdyt? | 12:14 |
| *** lucas-hungry is now known as lucasagomes | 12:14 | |
| openstackgerrit | Anton Arefiev proposed openstack/ironic-inspector: Fix response code for rule creating API https://review.openstack.org/327492 | 12:15 |
| jroll | dtantsur: yeah, definitely, that's going to get hairy | 12:15 |
| *** vmud213 has quit IRC | 12:15 | |
| dtantsur | snecklifter, ^^^. please follow the spec process https://wiki.openstack.org/wiki/Ironic/Specs_Process for your proposal | 12:16 |
| jroll | that's still a wiki page? O_o | 12:16 |
| dtantsur | yes, and that's what I was going to say as well | 12:18 |
| dtantsur | jroll, wanna move the whole thing to contributing docs or should I? | 12:19 |
| jroll | dtantsur: I don't right this moment :) | 12:19 |
| jroll | feel free to just file a bug, that's good low hanging fruit | 12:19 |
| dtantsur | good idea | 12:19 |
| *** Goneri has joined #openstack-ironic | 12:19 | |
| *** rbudden has joined #openstack-ironic | 12:21 | |
| dtantsur | here it goes: https://bugs.launchpad.net/ironic/+bug/1590759 | 12:21 |
| openstack | Launchpad bug 1590759 in Ironic "Move spec process description to our primary documentation" [Medium,Triaged] | 12:21 |
| jroll | thanks | 12:23 |
| *** thrash|g0ne is now known as thrash | 12:27 | |
| *** ohamada_ has joined #openstack-ironic | 12:29 | |
| *** ohamada has quit IRC | 12:29 | |
| snecklifter | erm, perhaps whilst that documentation is being moved it could be ... simplified a bit :) | 12:31 |
| *** jjohnson2 has joined #openstack-ironic | 12:31 | |
| jroll | snecklifter: heh, ++ | 12:34 |
| jroll | it's actually fairly simple | 12:34 |
| jroll | copy the template and go through it | 12:34 |
| dtantsur | yeah, maybe we need to make it shorter actually :) | 12:34 |
| jroll | and there's a symlink too | 12:34 |
| jroll | here's an example: https://github.com/openstack/ironic-specs/commit/ec3b9071b1fa90bf7fbf0dfacc7317795810b80e | 12:34 |
| lucasagomes | yeah, maybe we just need to remove the "historical" references | 12:35 |
| lucasagomes | Kilo bits, blueprints etc | 12:35 |
| dtantsur | Kilo is dead, lets remove it :) | 12:35 |
| jroll | so your spec goes in specs/approved/ and then `cd specs/not-implemented/ && ln -s ../approved/your-spec.rst .` | 12:35 |
| *** mgould has quit IRC | 12:40 | |
| *** mgould has joined #openstack-ironic | 12:40 | |
| *** thrash has quit IRC | 12:41 | |
| *** thrash has joined #openstack-ironic | 12:46 | |
| *** thrash has quit IRC | 12:46 | |
| *** thrash has joined #openstack-ironic | 12:46 | |
| openstackgerrit | Merged openstack/virtualbmc: Set IPMI session timeout for the virtual BMCs https://review.openstack.org/316022 | 12:47 |
| *** sdake has quit IRC | 12:48 | |
| *** ccarmack has joined #openstack-ironic | 12:53 | |
| *** links has quit IRC | 12:57 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Add power_transition_started_at to the nodes table https://review.openstack.org/327642 | 12:58 |
| *** rloo has joined #openstack-ironic | 12:58 | |
| openstackgerrit | Christopher Brown proposed openstack/ironic-specs: Implements: LVM/Software RAID support in ironic-python-agent Closes: https://bugs.launchpad.net/ironic/+bug/1590749 https://review.openstack.org/327643 | 12:58 |
| openstack | Launchpad bug 1590749 in Ironic "RFE: LVM/Software RAID support in ironic-python-agent" [Wishlist,Confirmed] | 12:58 |
| snecklifter | dtantsur: ^^^ | 12:59 |
| snecklifter | first ever spec so go easy on me! | 12:59 |
| dtantsur | sure :) | 12:59 |
| sambetts | snecklifter: Thanks for the spec, btw the format of commit messages should follow: https://wiki.openstack.org/wiki/GitCommitMessages | 13:01 |
| *** cdearborn has joined #openstack-ironic | 13:01 | |
| *** absubram has quit IRC | 13:05 | |
| dtantsur | I really wonder how to make a prebuilt RAID play well with out partitioning | 13:05 |
| sambetts | dtantsur: as in wholedisk image with RAID prebuilt? | 13:06 |
| dtantsur | oh, that's a good way to implement software RAID :D | 13:06 |
| dtantsur | sambetts, no, I'm rather thinking that we rebuild the partition table on every deploy | 13:06 |
| jroll | ++ | 13:07 |
| jroll | so, we do sw raid in onmetal | 13:07 |
| jroll | our cleaning destroys the raid, wipes the disks, rebuilds the raid | 13:07 |
| jroll | and uses whole disk images | 13:07 |
| jroll | (and craziness with bootloaders in those images, iirc) | 13:08 |
| sambetts | I think software RAID inside whole disk makes more sense than building it ourselves then its customised my the person creating the image | 13:08 |
| jroll | well, then we're done, right? :) | 13:08 |
| openstackgerrit | Yosef Hoffman proposed openstack/ironic-python-agent: Add configuration options for DISK_WAIT https://review.openstack.org/324909 | 13:08 |
| lucasagomes | jroll, sambetts dtantsur one tricky bit is network boot after deployment with sw RAID | 13:08 |
| jroll | some deployers don't want to put that on the image building folks | 13:08 |
| dtantsur | jroll, heh, yeah, I guess we already have everything for whole disk images. but what about partition images? | 13:09 |
| dtantsur | lucasagomes, what is tricky there? | 13:09 |
| jroll | dtantsur: yep | 13:09 |
| sambetts | and I guess if we support it for whole disk, we should support it for partition I guess | 13:09 |
| *** Nisha_away has joined #openstack-ironic | 13:09 | |
| *** jcoufal has joined #openstack-ironic | 13:09 | |
| jroll | but what I'm saying is it would be nice to support deploying a whole disk image without built-in raid to a software raid we build | 13:09 |
| jroll | (hi, that's my use case) | 13:09 |
| lucasagomes | dtantsur, because ironic right now just get the filesystem UUID with blkid as pass it in the kernel cmdline as being the root device | 13:10 |
| rbartal | TheJulia, good morning | 13:10 |
| sambetts | jroll: hmm interesting | 13:10 |
| dtantsur | yeah, interesting indeed | 13:10 |
| TheJulia | rbartal: good morning | 13:10 |
| dtantsur | wanna add it to midcycle topics? | 13:10 |
| lucasagomes | but the root device is the lvm volume for sw raid (with lvm) | 13:10 |
| jroll | is it not there already? | 13:10 |
| TheJulia | depends on the image | 13:11 |
| *** ayoung has joined #openstack-ironic | 13:11 | |
| jroll | oh, it isn't, sorry | 13:11 |
| sambetts | jroll: how would that work if your whole disk image had partitions defined in it that aren't compatible with the requested raid (is that even possible?) | 13:12 |
| rbartal | TheJulia, Hi, I installed the ipa image (http://paste.openstack.org/show/509170/) and try the erase_device on again, it fails with same error | 13:12 |
| rbartal | TheJulia, dtantsur advised the new code/patch is included in this version | 13:13 |
| lucasagomes | jroll, sambetts yeah, I'm also wondering how does that work | 13:13 |
| lucasagomes | and benefits as well | 13:13 |
| jroll | sambetts: I don't think that's possible, dunno | 13:14 |
| *** Goneri has quit IRC | 13:14 | |
| jroll | we treat the raid device as a disk, not a partition iirc | 13:15 |
| lucasagomes | jroll, even if it works, it seems like an extra overhead for no gain at all... Since one of the main benefits of using logical volumes is to not have to keep estimating the partition sizes prior to the dpeloyment | 13:15 |
| TheJulia | dtantsur: did some reading regarding the buzilla link you sent me yesterday and re-read it this morning. \o/ for deprecated option names :) | 13:16 |
| sambetts | jroll: I guess it depends on the FS in the whole disk image right? btfs vs lvm vs zfs or whatever | 13:16 |
| sambetts | lucasagomes: ^ | 13:16 |
| jroll | lucasagomes: well, we use intel's RSTe RAID, which is a weird thing that is exposed as mdraid on linux, and whatever windows softraid on windows, and the goal was to be able to build the RAID the same way for both linux and window | 13:16 |
| jroll | s | 13:16 |
| lucasagomes | sambetts, lvm is not a fs | 13:16 |
| lucasagomes | actually, fileystem like btrfs basically can kill lvm | 13:16 |
| lucasagomes | it can do things better, like bit rot on file blocks | 13:17 |
| sambetts | lucasagomes: right, I meant ext4+lvm vs btfs vs zfs | 13:17 |
| dtantsur | TheJulia, hehe | 13:17 |
| lucasagomes | sambetts, right | 13:17 |
| *** links has joined #openstack-ironic | 13:17 | |
| lucasagomes | jroll, ok, yeah I have to read more about that | 13:18 |
| * lucasagomes checks intel rste stuff | 13:18 | |
| jroll | lucasagomes: I *think* the only special thing we do when creating it is --metadata=imsm | 13:18 |
| TheJulia | rbartal: did you try booting another something else and checking to see if the drive is still frozen? | 13:18 |
| lucasagomes | sambetts, you will like this btw, http://arstechnica.com/information-technology/2014/01/bitrot-and-atomic-cows-inside-next-gen-filesystems/ | 13:18 |
| dtantsur | I'm putting this RAID discussion to https://etherpad.openstack.org/p/ironic-newton-midcycle please contribute | 13:18 |
| lucasagomes | sambetts, covers the next gen fs's (brtfs) vs logical volumes | 13:19 |
| dtantsur | I would like to have it in ironic, but I'm already completely lost | 13:19 |
| sambetts | lucasagomes: Oh awesome /me needs to read up | 13:19 |
| lucasagomes | sambetts, it's long, but it's a great overview why next gen fs are superior | 13:19 |
| TheJulia | I suspect it might not entirely be able to, and we have to remember some people do actually deploy windows with ironic in whole disk images.... | 13:19 |
| sambetts | tbh the RAID stuff for me I want to be able to create/destroy RAID on the fly via OOB, but I think we need to standardise the interface first | 13:20 |
| jroll | well, that's done, right? | 13:20 |
| jroll | drac driver is doing that now | 13:20 |
| sambetts | jroll: on the fly? based on flavor? | 13:21 |
| jroll | ah | 13:21 |
| jroll | ignore me :) | 13:21 |
| sambetts | heh ;) | 13:21 |
| TheJulia | sambetts: that is a big ask from some of our internal users | 13:21 |
| rbartal | TheJulia, yes, deployed overcloud on the node and re-run sudo hdparm -I /dev/sda2, it look the same as yesterday | 13:21 |
| sambetts | jroll, TheJulia: I added a section the RAID midcycle thingy talking about it | 13:21 |
| jroll | cool | 13:22 |
| TheJulia | awesome | 13:22 |
| TheJulia | rbartal: did you look through the hardware bios settings to see if it was something the bios was doing, since that is commonly what causes it | 13:22 |
| *** Nisha_away has quit IRC | 13:22 | |
| *** links has quit IRC | 13:23 | |
| rbartal | TheJulia, i looked at password and in general, do you have any specific pointer? | 13:23 |
| dtantsur | mm, I see rloo is here. g'morning rloo :) | 13:23 |
| rloo | hi dtantsur | 13:24 |
| TheJulia | rbartal: sadly I do not, since there is a massive variety out there, and it is something that the bios would basically be having to issue upon boot-up | 13:24 |
| sambetts | o/ rloo | 13:24 |
| rloo | oh. right. hi and good morning everyone, sambetts, TheJulia, rbartal, jroll, lucasagomes :) | 13:24 |
| lucasagomes | rloo, morning | 13:25 |
| TheJulia | Good morning rloo | 13:25 |
| rbartal | rloo, Good morning | 13:25 |
| jroll | morning rloo :) | 13:25 |
| lucasagomes | sambetts, I think the iLO machines does not support oob raid creation/deletion | 13:25 |
| dtantsur | correct | 13:26 |
| dtantsur | we're planning on OOB RAID for DELL | 13:26 |
| *** fellypefca has joined #openstack-ironic | 13:26 | |
| lucasagomes | jroll, and, intersting intel rste thingy, started reading the manual but it has 70+ pages so I will finish later | 13:26 |
| lucasagomes | http://www.intel.com/content/dam/support/us/en/documents/chipsets/rste/sb/intelr_rste_linux.pdf (for those interested) | 13:26 |
| snecklifter | sambetts: k, thanks, will update with feedback | 13:27 |
| rbartal | TheJulia, next week I can try different servers, what kind of HW would you recommend? HP ? | 13:27 |
| TheJulia | lucasagomes: thank you for the link | 13:27 |
| lucasagomes | yw | 13:27 |
| jroll | lucasagomes: heh, yeah the main thing for us is 1) hardware accelerated but behaves like mdraid, and 2) the same raid can work natively on windows and linux | 13:27 |
| NobodyCam | Good Morning Ironicers | 13:27 |
| jroll | and 3) our customers won't be quite as upset when they lose a boot disk ;) | 13:27 |
| dtantsur | morning NobodyCam | 13:28 |
| jroll | ohai NobodyCam :) | 13:28 |
| lucasagomes | jroll, yeah, it's neat | 13:28 |
| NobodyCam | morning dtantsur jroll lucasagomes TheJulia :) | 13:28 |
| TheJulia | rbartal: what exactly is your goal? | 13:28 |
| *** dprince has joined #openstack-ironic | 13:29 | |
| rbartal | TheJulia, I'm QA tester, and I try to check the cleaning steps are working, currently we have 1 step and that is the erase device, this step fails for me. | 13:30 |
| rbartal | TheJulia, so I want to make sure this is HW problem and not ipa problem | 13:31 |
| TheJulia | well, you could use an HP server. If a hardware raid controller is in use then the storage will likely be presented as scsi block devices to the OS, so secure erase functionality will not be used, and the shred command will be invoked. | 13:33 |
| mat128 | rbartal: from what I understand your drive is locked | 13:34 |
| dtantsur | using a vm would be even easier :) | 13:34 |
| TheJulia | mat128: frozen upon boot based on the hdparm output supplied yesterday | 13:34 |
| rloo | hey, doesn't anyone remember from one of the design sessions/austin summit. jroll volunteered to ... was it, show people how to debug a failed test? so that folks didn't just do 'recheck'? | 13:34 |
| mat128 | rbartal: Most BIOS lock the drives to prevent evil code from wiping your drive with 1 command (what a bad idea) | 13:34 |
| dtantsur | rbartal, that's fair point actually, you can try it on vm, use hard disk of e.g. 4 GiB | 13:34 |
| rloo | I mean s/doesn't/does/ | 13:34 |
| TheJulia | dtantsur: ++ | 13:35 |
| dtantsur | rloo, I remember something like that | 13:35 |
| dtantsur | but since then our gate improved a lot IMO | 13:35 |
| rloo | dtantsur: right, until the next *thing* happens. | 13:35 |
| dtantsur | of course :) | 13:36 |
| jroll | rloo: I volunteered to do that? :| | 13:36 |
| rloo | dtantsur: i'm going to put it down as a possible thing-to-do in midcycle. | 13:36 |
| rloo | jroll: according to dtantsur :) | 13:36 |
| dtantsur | lol | 13:36 |
| sambetts | rloo: that would be a cool session1 | 13:36 |
| lucasagomes | jroll, make a video of it :-) | 13:36 |
| dtantsur | yeah, there was a complaint that people got used to just issuing "recheck" every time | 13:36 |
| jroll | >.> | 13:36 |
| *** Goneri has joined #openstack-ironic | 13:36 | |
| dtantsur | c'mon! :D | 13:37 |
| lucasagomes | heh | 13:37 |
| dtantsur | debugging gate is easy! | 13:37 |
| * dtantsur ducks | 13:37 | |
| rbartal | dtantsur, mat128 I did try on vm and got the same, I don't think it support secure delete, as for the lock please see output of hdparm http://paste.openstack.org/show/508952/ | 13:37 |
| openstackgerrit | Milan Kováčik proposed openstack/ironic-inspector: [WIP]Introducing Inspector Grenade testing https://review.openstack.org/327667 | 13:37 |
| * lucasagomes remembers someone volunteered to show how elastic recheck works | 13:37 | |
| sambetts | I think devananda, said he might give an overview of how to use the query thing | 13:37 |
| sambetts | lucasagomes: ^ | 13:37 |
| lucasagomes | sambetts, yup | 13:38 |
| dtantsur | rbartal, yeah, vm does not support secure erase, but it should fall back to long and boring erase in this case | 13:38 |
| * dtantsur \o/ at inspector grenade testing | 13:38 | |
| jroll | nice | 13:38 |
| jroll | milan++ | 13:38 |
| rbartal | dtantsur, yes, we are waiting for this patch(fallback), still I like to see secure erase works | 13:39 |
| *** sdake has joined #openstack-ironic | 13:39 | |
| rloo | sambetts, lucasagomes: I mentioned in the midcycle etherpad, about elastic recheck :) | 13:39 |
| dtantsur | rbartal, sure, but if you want to see cleaning working, it should be enough to test with fallback | 13:39 |
| lucasagomes | rloo, cool, ty | 13:40 |
| sambetts | rloo: I see it :D | 13:40 |
| NobodyCam | morning rloo :) | 13:40 |
| lucasagomes | JayF, mat128 dtantsur if you guys have some time, mind taking a peek at https://review.openstack.org/#/c/323511/ ? | 13:40 |
| milan | jroll, dtantsur thx but you know, very first revision :D | 13:41 |
| * milan didn't run it yet | 13:41 | |
| TheJulia | rbartal: if there is no security capability, then the code that has been in IPA for quite a while should fall back to shredding | 13:41 |
| rloo | morning NobodyCam! | 13:41 |
| NobodyCam | :) | 13:41 |
| dtantsur | milan, when you get it more ready, ping jlvillal for review | 13:41 |
| milan | and morning jroll, rloo, NobodyCam TheJulia mat128 rbartal snecklifter :) | 13:42 |
| NobodyCam | morning milan :) | 13:42 |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: DO NOT REVIEW https://review.openstack.org/296432 | 13:42 |
| milan | dtantsur, sure thing :) | 13:42 |
| mat128 | morning milan | 13:42 |
| sambetts | milan: oh nice :D, are you going for full grenade first? or partial (inspector only) first? | 13:42 |
| milan | sambetts, good question :D | 13:43 |
| dtantsur | lucasagomes, looks great to me, lets see what other folks say | 13:43 |
| rbartal | TheJulia, from my testing it fails the cleaning all together, and we have this https://bugs.launchpad.net/ironic-python-agent/+bug/1536695 | 13:43 |
| openstack | Launchpad bug 1536695 in ironic-python-agent "Node cleaning fails if the _ata_erase fails" [High,In progress] - Assigned to Julia Kreger (juliaashleykreger) | 13:43 |
| milan | sambetts, seems it's going to be whatever Ironic is doing, since atm it's copy&paste O:-) | 13:43 |
| * milan hopes nobody notice :D | 13:43 | |
| mat128 | rbartal: running on a VM normally means the secure erase capability isn't even there, thus the code falls back to shred | 13:43 |
| dtantsur | lol | 13:43 |
| rloo | hi milan, vsaienko, mat128 :) | 13:43 |
| rbartal | mat128, ^^ from my testing it fails the cleaning all together, and we have this https://bugs.launchpad.net/ironic-python-agent/+bug/1536695 | 13:44 |
| openstack | Launchpad bug 1536695 in ironic-python-agent "Node cleaning fails if the _ata_erase fails" [High,In progress] - Assigned to Julia Kreger (juliaashleykreger) | 13:44 |
| mat128 | rbartal: the issue TheJulia is working on is falling back to shred in the case drive reports being ATA eraseable but fails for any reason | 13:44 |
| mat128 | rbartal: I'm well aware of this bug :) | 13:44 |
| TheJulia | dtantsur: rbartal https://review.openstack.org/#/c/302819/ is related to that, conductor side setting, but we need more information if it is failing with in VMs | 13:45 |
| *** sabeen has joined #openstack-ironic | 13:45 | |
| sambetts | milan: heh, full grenade then :D | 13:47 |
| rbartal | mat128, I may not understand correctly, this is the error I have after erase_device fails http://paste.openstack.org/show/508947/ | 13:47 |
| mat128 | rbartal: Block device /dev/sda is frozen and cannot be erased | 13:48 |
| rbartal | mat128, can I expect the fall back in this case? | 13:48 |
| mat128 | rbartal: that means the drive reports being ATA eraseable, but since it's frozen it cant | 13:48 |
| mat128 | rbartal: only with TheJulia's code | 13:48 |
| openstackgerrit | Christopher Brown proposed openstack/ironic-specs: LVM/Software RAID support in ironic-python-agent https://review.openstack.org/327643 | 13:48 |
| mat128 | rbartal: without it, you are a victim of https://bugs.launchpad.net/ironic-python-agent/+bug/1536695 | 13:49 |
| openstack | Launchpad bug 1536695 in ironic-python-agent "Node cleaning fails if the _ata_erase fails" [High,In progress] - Assigned to Julia Kreger (juliaashleykreger) | 13:49 |
| lucasagomes | dtantsur, cheerios | 13:49 |
| *** ametts has joined #openstack-ironic | 13:49 | |
| TheJulia | rbartal: fallback is essentially opt-in only because of the damage it can do to SSD's | 13:49 |
| *** piet has joined #openstack-ironic | 13:50 | |
| openstackgerrit | Merged openstack/python-oneviewclient: Remove insecure connection warnings https://review.openstack.org/323873 | 13:51 |
| rbartal | mat128, TheJulia if I sum up the issue: the SDD report ATA eraseable, hdparm show it not frozen, but IPA find it frozen | 13:52 |
| mat128 | rbartal: I found it frozen in your paste | 13:52 |
| TheJulia | what mat128 said | 13:52 |
| mat128 | http://paste.openstack.org/show/508952/ line 88 | 13:53 |
| mat128 | rbartal: ^ | 13:53 |
| vsaienko | morning rloo | 13:53 |
| rbartal | TheJulia, mat128 right, my mistake, it was not frozen when I run this procedure on the same SDD on other server http://www.intel.com/content/www/us/en/support/software/000006094.html | 13:54 |
| openstackgerrit | Christopher Brown proposed openstack/ironic-specs: Add missing symlink https://review.openstack.org/327679 | 13:54 |
| mat128 | rbartal: Intel bioses normally offer locking as an option | 13:55 |
| openstackgerrit | Andre Aranha proposed openstack/python-oneviewclient: Extend python-oneviewclient to Network https://review.openstack.org/294559 | 13:55 |
| rbartal | TheJulia, mat128 is it possible ipa lock the SSD? | 13:56 |
| openstackgerrit | Andre Aranha proposed openstack/python-oneviewclient: Extend python-oneviewclient to Network https://review.openstack.org/294559 | 13:56 |
| mat128 | rbartal: it doesnt | 13:56 |
| mat128 | rbartal: what it does it set a password on it, because that's the only way you can issue a secure erase | 13:56 |
| rbartal | mat128, but I see the password is not set after ipa try the secure delete, can it fail to set password? | 13:57 |
| snecklifter | milan: morning | 13:57 |
| mat128 | rbartal: it could fail for any reason, but the reason is returned in the error message and from your case it's simply saying "/dev/sda is frozen and cannot be erased" | 13:57 |
| mat128 | https://github.com/openstack/ironic-python-agent/blob/031593614eca21d4c88451fcb120f23de56e3f97/ironic_python_agent/hardware.py#L751 | 13:58 |
| mat128 | rbartal: if hdparm does not report "not frozen", then you get that message | 13:58 |
| snecklifter | dtantsur: thanks for review, have updated, any better? | 13:58 |
| rbartal | mat128, I'm trying to make it work, is there any action I should preform? btw this is conductor log with the error http://paste.openstack.org/show/508645/ | 14:00 |
| mat128 | rbartal: It is possible that your hardware does not allow you to secure erase devices. This happens on some of our devices downstream. In that case, you will either want to leverage TheJulia's configuration option (not landed yet) or use different hardware. | 14:01 |
| mat128 | Using a hard drive that does not claim to support ATA secure erase would allow the code to gracefully fall back to shred, even without TheJulia's patch. | 14:01 |
| dtantsur | snecklifter, I'll take a look sooner or later. we need to figure out what we plan to see for software RAID, as you can see there are plenty of variants | 14:01 |
| mat128 | That would be the case of a virtual hard drive, such as the one in a virtual machine | 14:01 |
| mat128 | rbartal: that 404 is weird, are you sure it's related to cleaning? | 14:02 |
| mat128 | rbartal: neutronclient raised that while trying to authenticate with keystone | 14:02 |
| dtantsur | snecklifter, do you plan to be at the virtual midcycle? https://etherpad.openstack.org/p/ironic-newton-midcycle | 14:02 |
| rbartal | mat128, sorry this is log form were the ease_device is disabled | 14:03 |
| *** PollyZ has joined #openstack-ironic | 14:03 | |
| snecklifter | dtantsur: virtual? yes, i'll add it to my diary. Happy to attend if you think it will help | 14:04 |
| dtantsur | snecklifter, I hope to discuss this software RAID there | 14:05 |
| snecklifter | dtantsur: thanks for the invite! | 14:07 |
| rbartal | mat128, this is the log http://paste.openstack.org/show/509219/ | 14:08 |
| mat128 | rbartal: "in delete_cleaning_ports" | 14:09 |
| mat128 | rbartal: do you have the right setup to use neutron cleaning ports? | 14:09 |
| mat128 | rbartal: it seems it is unable to authenticate with keystone, in an attempt to interact with neutron | 14:10 |
| johnthetubaguy | jroll: you know I said about making it work when the user creates their own port, and passes it to nova, when booting their instance, I think this should make that work: https://review.openstack.org/#/c/327009/3 | 14:10 |
| rbartal | mat128, so this is the tear down problem, its the new patch, I will try testing again. | 14:11 |
| mat128 | rbartal: ok, I might be away for an hour or two, but mention me and I will be able to read it when I get in | 14:12 |
| rbartal | mat128, TheJulia, one more question :) Why is there erase_devices_priority in the [deploy] section and clean_priority_erase_devices in the the [ilo] section ? | 14:12 |
| rbartal | mat128, thanks for your help, I'm on GMT +3 so it end of day for me :) | 14:12 |
| mat128 | rbartal, TheJulia: I remember the former being under [agent]. Is it possible we simply haven't removed the ilo one? | 14:13 |
| sambetts | johnthetubaguy: https://review.openstack.org/#/c/277853/4/specs/approved/VLAN-aware-baremetal-instances.rst that problem won't be an issue any more when this spec merges, because it proposes making Ironic override the mac address at the point when we bind during the deploy process | 14:13 |
| sambetts | johnthetubaguy: also its down to the network driver, some network drivers might not override the mac address, because some hardware supports virtual mac addresses | 14:14 |
| johnthetubaguy | sambetts: yeah, the other change is really to allow a different refactor in that code | 14:14 |
| mat128 | rbartal: https://bugs.launchpad.net/ironic/+bug/1515871 | 14:14 |
| openstack | Launchpad bug 1515871 in Ironic "agent_ilo driver uses it's own CONF variable for erase device priority" [Medium,In progress] - Assigned to Shuquan Huang (shuquan) | 14:14 |
| johnthetubaguy | sambetts: agreed the details are a bit tricky though, its probably going to need unassigned mac addresses and things, in some cases | 14:14 |
| mat128 | rbartal: the [ilo] one should go away and the [deploy] one should be used. | 14:15 |
| rbartal | mat128, thank | 14:15 |
| *** rama_y has joined #openstack-ironic | 14:15 | |
| sambetts | johnthetubaguy: thats why I want move all the logic into Ironic, so code in Nova on longer defines any of that info, nova will create the port, but with the new Ironic network stuff we don't bind in nova any more so Ironic can override anything we want before we bind during deploy | 14:17 |
| *** joprovost has joined #openstack-ironic | 14:17 | |
| johnthetubaguy | sambetts: I am fine extracting the logic, if we can make that work, this refactoring is to make routed networks work, where we need to create the ports before we call the scheduler | 14:19 |
| sambetts | johnthetubaguy: myself and sukdev also discussed at the summit allowing ports of the baremetal type to exist without mac addresses, this is something that would be ideal for the case where nova creates the port, and also I want it for this spec here: https://review.openstack.org/#/c/255240/4/specs/newton/dhcp-serve-unknown-hosts.rst were I allow a port to reserve a ip address, but DHCP to serve | 14:19 |
| sambetts | unknown hosts | 14:19 |
| johnthetubaguy | sambetts: yeah, I like the port without a mac address | 14:20 |
| johnthetubaguy | routed networks will have ports without ips | 14:20 |
| sambetts | johnthetubaguy: to represent a bond interface? | 14:20 |
| johnthetubaguy | no, when IPs are local to certain sub sets of hosts | 14:21 |
| sambetts | johnthetubaguy: oh :/ | 14:22 |
| * sambetts needs to go and re-read that spec | 14:22 | |
| *** rbartal has quit IRC | 14:23 | |
| *** ccarmack has quit IRC | 14:25 | |
| sambetts | johnthetubaguy: the code I'm proposing in my VLAN aware BM spec, basicly just makes it so that the mac address the port has whether given to it by a user, or generated by neutron, will be ignored because Ironic will replace it anyways, but there is also cases when we want a mac address generated by neutron, because some hardware can be programed with mac addresses, so we don't always want to | 14:26 |
| sambetts | not have a generated mac address | 14:26 |
| *** ccarmack has joined #openstack-ironic | 14:29 | |
| *** PollyZ has quit IRC | 14:37 | |
| jroll | johnthetubaguy: will have a look, thanks | 14:39 |
| openstackgerrit | Ruby Loo proposed openstack/ironic: Replace dict.get(key) with dict[key] https://review.openstack.org/327726 | 14:41 |
| jroll | johnthetubaguy: seems like it should work, at a glance | 14:41 |
| *** pcaruana has quit IRC | 14:41 | |
| *** piet has quit IRC | 14:42 | |
| *** PollyZ has joined #openstack-ironic | 14:42 | |
| *** dulek has joined #openstack-ironic | 14:44 | |
| openstackgerrit | Moshe Levi proposed openstack/ironic-python-agent: Add vendor id and vendor name to interface information https://review.openstack.org/279940 | 14:53 |
| vsaienko | jroll moving discussion about smoke test on ironic master here :). Ironic'ers the question is next. We are going to switch grenade jobs to nv per commit mode. With grenade smoke tests are launched instead of baremetal basic ops. I think that wee need to have a job that will launch smoke tests on the ironic master stage without upgrade. We need to know at least if we failed due to incorrect test configuration or due to upgrade. What do you think | 14:53 |
| openstackgerrit | Yosef Hoffman proposed openstack/ironic-python-agent: Add configuration options for DISK_WAIT https://review.openstack.org/324909 | 14:53 |
| *** ccarmack1 has joined #openstack-ironic | 14:53 | |
| jroll | vsaienko: well, the logs can tell us that, but | 14:54 |
| jroll | maybe our regular test jobs can just add smoke to the regex? | 14:54 |
| jroll | regex=(ironic|\bsmoke\b) | 14:54 |
| jroll | or something | 14:54 |
| vsaienko | might be we need to add baremetal basic ops to smoke suite and always launch smoke scenario? | 14:55 |
| jroll | honestly now that we can boot an instance with "normal" tempest tests... do we need the baremetal scenario? | 14:55 |
| jroll | yeah, that could work too | 14:55 |
| jroll | we'd still need api tests included too | 14:55 |
| sambetts | I think we could just ditch baremetal basic ops right? | 14:55 |
| * jroll wonders how much work tempest-full is from here | 14:55 | |
| jroll | sambetts: yeah, that's kinda what I'm thinking | 14:56 |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic: Add node provsioning methods to the tempest plugin https://review.openstack.org/327238 | 14:56 |
| *** ccarmack has quit IRC | 14:56 | |
| *** thiagop has joined #openstack-ironic | 14:57 | |
| sambetts | vsaienko, jroll: makes sense to me, smoke covers the same and more than basic ops does, so | 14:58 |
| vsaienko | jroll I will try to figure out what we need to have green full tempest job | 14:59 |
| jroll | vsaienko: you rock | 14:59 |
| openstackgerrit | Anton Arefiev proposed openstack/ironic-inspector: Tempest: don't rely on tempest ironic client https://review.openstack.org/327743 | 14:59 |
| *** waltz has joined #openstack-ironic | 14:59 | |
| sambetts | \o/ | 14:59 |
| jroll | vsaienko: the whole "wait for cleaning to finish" will be involved, so maybe turn off cleaning to start? | 15:00 |
| *** mtanino has joined #openstack-ironic | 15:00 | |
| waltz | Hello everyone, Is it possible to pass kickstart files to servers being booted by ironic | 15:01 |
| waltz | I am trying to use Ironic to automate exsi builds | 15:01 |
| vsaienko | jroll, we need to have tests for manual cleaning first, I will update my patches | 15:02 |
| sambetts | waltz: we don't install an OS like cobbler, we install an image. I would look into creating a prebuilt whole disk image of exsi | 15:02 |
| vsaienko | jroll so launching smoke tests takes near 25 minutes http://logs.openstack.org/09/311109/66/experimental/gate-grenade-dsvm-ironic/e2b7fed/console.html#_2016-06-07_17_57_12_940 | 15:02 |
| jroll | vsaienko: well, if you're just trying to explore what's needed, you could try without cleaning :) | 15:02 |
| jroll | yeah | 15:02 |
| vsaienko | it is the worst case I think | 15:02 |
| waltz | sambetts: thanks for this | 15:03 |
| thiagop | Morning folks | 15:03 |
| JayF | Given we are using tinyipa, and less ram | 15:03 |
| JayF | can we start to investigate some additional parallelism for these tests? | 15:03 |
| jroll | JayF: so there's a couple things there | 15:03 |
| JayF | even if it just means more nodes so that the tempest cleaning/capacity race isn't an issue | 15:03 |
| jroll | 1) max is 7 VMs, I think, which isn't enough to satisfy tempest-full quite yet | 15:04 |
| sambetts | jroll, JayF: I have had a recent tinyipa booted with 256mb RAM locally, not tried in the gate though | 15:04 |
| JayF | sambetts: we could make an experimental job for it if you want | 15:04 |
| JayF | sambetts: I can get a patch up for that in like 5m if you think it's useful | 15:04 |
| jroll | 2) we need to split away our api tests from these scenario tests, because they kind of stomp on each other in parallel mode | 15:04 |
| vsaienko | jroll baremetal basic ops takes near 5 min. And the overall job takes near 1h, while allowed time is 3h. I think it makes sense to switch to smoke tests | 15:04 |
| jroll | vsaienko: agree, my long-term goal is tempest-full, that's why I'm curious :) | 15:05 |
| sambetts | vsaienko: why does smoke tests take long to launch? | 15:05 |
| jroll | vsaienko: no rush though | 15:05 |
| jroll | JayF: let's get grenade stuff merged before we ask infra core for anything else :P | 15:05 |
| * jroll is reminded to bug again | 15:05 | |
| JayF | jroll: in this case; I'm just saying we could configure to launch more nodes; say 6 instead of 3, and you completely erradicate the issue of having to wait for a node to exit cleaning to readd the capacity before we move on | 15:05 |
| jroll | actually, gonig to send an email instead | 15:06 |
| sambetts | vsaienko: devstack time? | 15:06 |
| jroll | JayF: sure, we do 7 today | 15:06 |
| JayF | jroll: no we don't | 15:06 |
| JayF | jroll: not for basic jobs | 15:06 |
| sambetts | JayF, jroll: how many VM would we manage if we reduced the ram even more? | 15:06 |
| jroll | JayF: basic jobs only boot one instance, nothing to parallelize | 15:06 |
| JayF | but it was said above ^ the reason we wouldn't change that to smoke | 15:06 |
| JayF | is because of that cleaning race | 15:07 |
| jroll | sambetts: I think we could 10-11 at 256mb | 15:07 |
| JayF | If I'm just missing something that's fine, it's early and brain isn't warmed up yet, lol | 15:07 |
| jroll | JayF: I was talking about full, not smoke | 15:07 |
| jroll | which boots a metric ton of instances | 15:07 |
| vsaienko | JayF, when we have tests for cleaning https://review.openstack.org/#/c/304469/. We can disable automated cleaning at all. | 15:07 |
| sambetts | how many do we need for full? is it down to how fast we can recycle the nodes? | 15:07 |
| jroll | sambetts: I don't know, but it's quite a few :) | 15:08 |
| *** thiagop is now known as thiagop-planning | 15:08 | |
| vdrok | sambetts: yes, and also keeping the time in 3 hr bound | 15:08 |
| vsaienko | JayF: It will save a time and allow to launch tests in parallel | 15:08 |
| *** rama_y has quit IRC | 15:09 | |
| *** itamarl has quit IRC | 15:10 | |
| sambetts | if we ditch basic ops upstream in favor of smoke, what do we expect the third party CIs to run? Right now we're using basic ops and it takes ~1.5hrs to run booting on instance, and I don't have enough hardware to throw at booting multiple instances in parallel | 15:10 |
| vdrok | maybe this one? https://review.openstack.org/327238 | 15:11 |
| vdrok | it does basically the same thing | 15:11 |
| jroll | side note: still confused why it takes longer to run baremetal basic ops on hardware | 15:11 |
| sambetts | jroll: real hardware posts are slow | 15:11 |
| sambetts | fast network, slow post :( | 15:11 |
| jroll | sambetts: but 30-45 minutes slow? :/ | 15:12 |
| jroll | though I guess you probably don't have all the optimizations of pre-built devstack images, local pip caches, etc | 15:12 |
| dtantsur | 5-10 minutes for every boot already gives 15-30 minutes with deploy-boot-clean flow | 15:12 |
| jroll | aha, fair enough | 15:12 |
| sambetts | ^ | 15:12 |
| dtantsur | (well, 10 is a bit too much, but 5 is pretty possible) | 15:12 |
| jroll | yeah, we have 3-5 minute POSTs iirc | 15:12 |
| dtantsur | throw in inspection, and you get even more | 15:12 |
| sambetts | I've seen >10 if the machine decides to do a full health and power check | 15:13 |
| sambetts | which when its going on and off again all day it likes to do | 15:13 |
| jroll | ouch | 15:14 |
| *** jaybeale has joined #openstack-ironic | 15:14 | |
| sambetts | yeah, thats why boot from volume can be such a massive optimisation, because it removes a boot cycle | 15:15 |
| JayF | So I feel like this is a dumb question | 15:15 |
| JayF | but how does nova get away from being vm-in-vm? | 15:15 |
| JayF | They obviously have to build instances somehow | 15:16 |
| moshele | jroll: is it possible to remove the -2 from https://review.openstack.org/#/c/279940/ I can update the commit message that it just adds the interface interface vendor id and vendor name and remove the dependency to the infiniband use case as there are other use cases to it like tripleO | 15:16 |
| jroll | JayF: they don't, but they only need to boot cirros, and make sure ssh works | 15:16 |
| sambetts | JayF: they only have to do one boot per instance, also not pxe, and also not dding things, I've found IOPs is v.slow in nested vms | 15:16 |
| jroll | rather than booting a large image and doing a bunch of disk io | 15:16 |
| JayF | jroll: ^ fwiw I agree with moshele. If this had been something wanting to be added to inspection we woulc've already merged it | 15:17 |
| JayF | sambetts: jroll okay, that makes a lot of sense | 15:17 |
| *** mbound has quit IRC | 15:17 | |
| jroll | moshele: JayF done | 15:18 |
| JayF | moshele: it'll be on my list to review today | 15:18 |
| openstackgerrit | Christopher Brown proposed openstack/ironic-specs: LVM/Software RAID support in ironic-python-agent Closes-bug: #1590749 https://review.openstack.org/327769 | 15:18 |
| openstack | bug 1590749 in Ironic "RFE: LVM/Software RAID support in ironic-python-agent" [Wishlist,In progress] https://launchpad.net/bugs/1590749 - Assigned to Christopher Brown (snecklifter) | 15:18 |
| moshele | JayF: thanks :) | 15:18 |
| *** mbound has joined #openstack-ironic | 15:22 | |
| dtantsur | rloo, I know you like reviewing RFEs, especially talking about client modifications, mind reviewing this short one https://bugs.launchpad.net/python-ironicclient/+bug/1590752 ? | 15:23 |
| openstack | Launchpad bug 1590752 in python-ironicclient "[RFE] --wait flag for node-set-provision-state" [Wishlist,Confirmed] | 15:23 |
| dtantsur | :) | 15:23 |
| rloo | dtantsur: s/like/hate/ ? in a meeting, will look at it in a few hours. sigh. seems like i have a lot of meetings today. so will look later today. | 15:24 |
| dtantsur | rloo, no worries then, I'll ping someone else | 15:26 |
| *** ccarmack has joined #openstack-ironic | 15:26 | |
| rloo | dtantsur: even better then! ;) | 15:26 |
| dtantsur | lucasagomes, still around? | 15:26 |
| *** Nisha_away has joined #openstack-ironic | 15:27 | |
| jroll | sent an email asking for reviews on the grenade things, btw | 15:28 |
| *** ccarmack1 has quit IRC | 15:29 | |
| dtantsur | lucasagomes, when you're here, we need someone more to take a loot at easy RFE: https://bugs.launchpad.net/python-ironicclient/+bug/1590752 | 15:29 |
| openstack | Launchpad bug 1590752 in python-ironicclient "[RFE] --wait flag for node-set-provision-state" [Wishlist,Confirmed] | 15:29 |
| openstackgerrit | Moshe Levi proposed openstack/ironic-python-agent: Add vendor id and vendor name to interface information https://review.openstack.org/279940 | 15:30 |
| dtantsur | this is again part of my general "make tripleo reinvent less wheels" effort :) | 15:30 |
| *** jaybeale has quit IRC | 15:30 | |
| *** tesseract has quit IRC | 15:30 | |
| *** mbound has quit IRC | 15:31 | |
| lucasagomes | dtantsur, yup | 15:32 |
| lucasagomes | dtantsur, 1 sec will do | 15:32 |
| *** rcernin has quit IRC | 15:32 | |
| *** wajdi has joined #openstack-ironic | 15:32 | |
| *** PollyZ_ has joined #openstack-ironic | 15:35 | |
| *** dprince has quit IRC | 15:36 | |
| lucasagomes | dtantsur, jroll the idea LGTM | 15:37 |
| lucasagomes | dtantsur, that --wait flag, how the syntax will look like? --wait OR --wait=<integer> ? | 15:37 |
| lucasagomes | not sure I've seems something like that | 15:37 |
| jroll | lucasagomes: it says that in the spec, both | 15:37 |
| dtantsur | lucasagomes, both. --wait will mean --wait=infinity | 15:37 |
| jroll | ^^ | 15:37 |
| *** sacharya has joined #openstack-ironic | 15:37 | |
| lucasagomes | yeah I know, just wondering if it's something common | 15:38 |
| dtantsur | lucasagomes, I've seen such things here and there, and iirc argparse supports it | 15:38 |
| *** PollyZ has quit IRC | 15:38 | |
| lucasagomes | usually --<flag> are True/False | 15:38 |
| lucasagomes | dtantsur, ok so +1 | 15:38 |
| *** Nisha_brb has joined #openstack-ironic | 15:38 | |
| *** PollyZ_ has quit IRC | 15:38 | |
| *** Goneri has quit IRC | 15:39 | |
| *** dmk0202 has quit IRC | 15:41 | |
| lucasagomes | dtantsur, yeah nargs='?' supports it | 15:41 |
| *** Nisha_away has quit IRC | 15:41 | |
| * lucasagomes wanted to confirm :-) | 15:41 | |
| openstackgerrit | Kyrylo Romanenko proposed openstack/python-ironicclient: Add tests for openstackclient plugin https://review.openstack.org/301687 | 15:42 |
| dtantsur | thanks, I'll start hacking on it soon (probably tomorrow) | 15:42 |
| lucasagomes | dtantsur, cool | 15:44 |
| *** PollyZ has joined #openstack-ironic | 15:49 | |
| devananda | g'morning, all | 15:51 |
| NobodyCam | Good morning devananda | 15:51 |
| TheJulia | good morning devananda | 15:51 |
| sambetts | Hi NobodyCam, devananda | 15:51 |
| NobodyCam | good morning sambetts :) | 15:51 |
| * devananda catches up on scrollback | 15:51 | |
| openstackgerrit | Kyrylo Romanenko proposed openstack/python-ironicclient: Negative tests for port-related commands https://review.openstack.org/259354 | 15:52 |
| thiagop-planning | morning devananda | 15:52 |
| dtantsur | morning devananda | 15:52 |
| NobodyCam | morning thiagop-planning | 15:52 |
| devananda | sambetts, lucasagomes: re elasticrecheck, yea, I'll be happy to do a screen share or walk folks through it during the midcycle | 15:52 |
| sambetts | devananda: :D awesome, thanks! | 15:53 |
| lucasagomes | devananda, morning, yeah that would be great | 15:53 |
| *** mjura has quit IRC | 15:53 | |
| dtantsur | ++ for walk through | 15:54 |
| *** bnemec has quit IRC | 15:55 | |
| *** vishwana_ has joined #openstack-ironic | 15:59 | |
| *** vishwanathj has quit IRC | 16:02 | |
| *** amotoki has joined #openstack-ironic | 16:03 | |
| *** bnemec has joined #openstack-ironic | 16:04 | |
| *** Jabadia has joined #openstack-ironic | 16:04 | |
| *** rama_y has joined #openstack-ironic | 16:05 | |
| *** ijw has joined #openstack-ironic | 16:06 | |
| thiagop-planning | mornin' NobodyCam | 16:07 |
| openstackgerrit | Merged openstack/ironic: Add config option for ATA erase fallback in agent https://review.openstack.org/302819 | 16:08 |
| *** PollyZ has quit IRC | 16:11 | |
| openstackgerrit | Christopher Brown proposed openstack/ironic-specs: LVM/Software RAID support in ironic-python-agent Closes-bug: #1590749 https://review.openstack.org/327769 | 16:12 |
| openstack | bug 1590749 in Ironic "RFE: LVM/Software RAID support in ironic-python-agent" [Wishlist,In progress] https://launchpad.net/bugs/1590749 - Assigned to Christopher Brown (snecklifter) | 16:12 |
| jroll | harlowja: I just brought my flamethrower to your party :D | 16:13 |
| *** bcornec1 has joined #openstack-ironic | 16:14 | |
| * NobodyCam *ducks* | 16:14 | |
| *** waltz has quit IRC | 16:14 | |
| * sambetts isn't sure he was invited to this party :) | 16:15 | |
| *** PollyZ has joined #openstack-ironic | 16:16 | |
| TheJulia | On the mailing list | 16:17 |
| devananda | jroll: oh, nice! /me gets marshmallows | 16:18 |
| dtantsur | jroll, looks reasonable | 16:19 |
| *** fellypefca has quit IRC | 16:20 | |
| * dtantsur remembers his first attempt to install ironic manually from scratch | 16:20 | |
| dtantsur | well, now I would probably use puppet | 16:20 |
| *** vishwana_ has quit IRC | 16:20 | |
| sambetts | Looks good :D, jroll you comments about the nova driver align with this: https://review.openstack.org/#/c/317636/ | 16:21 |
| * sambetts is currently working on POC code for it | 16:21 | |
| dtantsur | JayF, TheJulia, oh interesting: https://bugs.launchpad.net/ironic-python-agent/+bug/1590606 looks like it's possible to detect if 'iscsistart -b' is of any use | 16:22 |
| openstack | Launchpad bug 1590606 in ironic-python-agent "[RFE] Support remove iSCSI shares as root devices via iscsistart" [Wishlist,Triaged] - Assigned to Yossi Ovadia (jabadia) | 16:22 |
| dtantsur | (see the last comment by the reporter) | 16:23 |
| *** sdake has quit IRC | 16:25 | |
| sambetts | dtantsur: seems to make sense to me | 16:26 |
| jroll | sambetts: indeed :) | 16:27 |
| jroll | dtantsur: when I spin up a new service, I always try to get it working by hand first, build automation to copy that, then blow it away and run the automation to create it from scratch again | 16:28 |
| * jroll afks to meeting | 16:28 | |
| *** alexpilotti has joined #openstack-ironic | 16:28 | |
| TheJulia | dtantsur: I was kind of suspecting that it might be a good idea to look at the pci bus device descriptions, but that may not be at all viable | 16:29 |
| sambetts | dtantsur: is this a reason to have the pci devices in the inventory? | 16:30 |
| sambetts | dtantsur: Re: the other spec | 16:30 |
| dtantsur | sambetts, it's happening on IPA side, we don't need to pass them arounfd | 16:30 |
| dtantsur | especially since it seems like there is a command to say yes/no: iscsistart -f | 16:31 |
| *** afaranha has joined #openstack-ironic | 16:31 | |
| dtantsur | I think we might probably have something like: if iscsistart -f; then iscsistart -b; fi (but in python) | 16:31 |
| Jabadia | Im just about to issue a pr on it | 16:32 |
| Jabadia | ( will be my first one. Please be gentle. ) | 16:32 |
| *** ijw_ has joined #openstack-ironic | 16:32 | |
| dtantsur | that would be interesting, thanks | 16:32 |
| harlowja | jroll oh goodie! | 16:33 |
| jroll | :D | 16:33 |
| dtantsur | I won't be around for review today, but will definitely take a look tomorrow | 16:33 |
| *** ijw_ has quit IRC | 16:33 | |
| harlowja | will check in a few :-P | 16:33 |
| *** ijw_ has joined #openstack-ironic | 16:33 | |
| *** Romanenko_K has quit IRC | 16:34 | |
| *** kromanenko__ has quit IRC | 16:34 | |
| JayF | jroll: I was promised a flame, but that email was actually constructive | 16:34 |
| *** kromanenko_ has quit IRC | 16:34 | |
| jroll | JayF: yeah, josh just kept talking about how he was starting a flame war there so I had to say it :P | 16:35 |
| openstackgerrit | Yossi Ovadia proposed openstack/ironic-python-agent: Handle diskless hardware connected to remote iscsi. https://review.openstack.org/327807 | 16:35 |
| harlowja | lol | 16:35 |
| dtantsur | aha, here's the patch | 16:35 |
| openstackgerrit | Andre Aranha proposed openstack/python-oneviewclient: Extend python-oneviewclient to Uplink Set https://review.openstack.org/295981 | 16:35 |
| harlowja | i like it spicay | 16:35 |
| harlowja | lol | 16:35 |
| *** ijw has quit IRC | 16:37 | |
| *** sdake has joined #openstack-ironic | 16:38 | |
| dtantsur | Jabadia, reviewed | 16:38 |
| JayF | I'm going to say this again at a different overlapping timezone; if anyone wants a pixie boots sticker for their laptop feel free to let me know | 16:38 |
| Jabadia | Tnx! | 16:39 |
| dtantsur | JayF, did you have a chance to send them to me already? | 16:39 |
| JayF | I believe you should already have them | 16:39 |
| JayF | if you don't resend me you address adn I'll resend them | 16:39 |
| dtantsur | interesting. I'll wait a bit more, czech post is no good :) | 16:39 |
| dtantsur | also going now, see you all tomorrow | 16:40 |
| *** dtantsur is now known as dtantsur|afk | 16:40 | |
| jlvillal | dtantsur|afk: Paka paka! | 16:40 |
| JayF | dtantsur|afk: I mean, I sent it like a month ago | 16:40 |
| JayF | dtantsur|afk: so I should probably resend :) | 16:40 |
| JayF | dtantsur|afk: would not surprise me if I shorted it on postage or smth by accident | 16:40 |
| *** daemontool has quit IRC | 16:41 | |
| harlowja | jroll thats not a flame, lol | 16:41 |
| dtantsur|afk | JayF, ah understood | 16:41 |
| openstackgerrit | Andre Aranha proposed openstack/python-oneviewclient: Extend python-oneviewclient to Uplink Set https://review.openstack.org/295981 | 16:41 |
| harlowja | i think u actually made sense | 16:42 |
| harlowja | so +1 | 16:42 |
| harlowja | lol | 16:42 |
| jroll | harlowja: I torched everything before I sent it so that it was all clear :P | 16:42 |
| jroll | also | 16:42 |
| harlowja | lol | 16:42 |
| openstackgerrit | Andre Aranha proposed openstack/python-oneviewclient: Extend python-oneviewclient to Uplink Set https://review.openstack.org/295981 | 16:42 |
| harlowja | jroll the k8s stuff i'm just waiting for someone to start trying to do it (might be me, others at godaddy) | 16:42 |
| jroll | wait until everybody just picks out the 'separate nova api out' part and that devolves into a flame war itself | 16:42 |
| harlowja | :-P | 16:43 |
| *** trown is now known as trown|lunch | 16:43 | |
| harlowja | true dat | 16:43 |
| *** d0ugal has quit IRC | 16:43 | |
| *** d0ugal has joined #openstack-ironic | 16:44 | |
| *** sdake_ has joined #openstack-ironic | 16:45 | |
| *** sdake has quit IRC | 16:45 | |
| rloo | lucasagomes, wrt https://review.openstack.org/#/c/302819/, if both you and yuriy prefer 'ata', i think we should change it. | 16:45 |
| lucasagomes | rloo, well, it's all good. I think it's just personal preferences | 16:46 |
| *** jaybeale has joined #openstack-ironic | 16:46 | |
| lucasagomes | I woulldn't +2 or +a if I I was strongly against the name | 16:46 |
| rloo | lucasagomes: ok, if you're sure. | 16:46 |
| lucasagomes | rloo, it's just comestics | 16:46 |
| lucasagomes | yeah | 16:46 |
| rloo | lucasagomes: don't need to be strongly against. this is what the user sees. so if 'ata' is better, we should do that. | 16:46 |
| *** ijw_ has quit IRC | 16:47 | |
| lucasagomes | rloo, right, yeah but I think it's one of those things were opinions may be divided and none are wrong | 16:47 |
| lucasagomes | so, either disk or ata works | 16:47 |
| *** ijw has joined #openstack-ironic | 16:48 | |
| rloo | lucasagomes: ok then. | 16:48 |
| *** e0ne has quit IRC | 16:48 | |
| *** sdake_ has quit IRC | 16:52 | |
| *** ifarkas has quit IRC | 16:55 | |
| vdrok | JayF: do you send to ukraine? :) | 16:56 |
| JayF | vdrok: I'll send stickers to anyone that wants/needs them :) | 16:57 |
| JayF | Ironic pays my bills, least I can do is ship some stickers around | 16:57 |
| *** amotoki has quit IRC | 16:57 | |
| JayF | just pm me and address and how many you'll need, just make sure you actually use all the ones I send since I only got 250 | 16:57 |
| vdrok | ok, thanks! | 16:57 |
| *** ccarmack has quit IRC | 17:01 | |
| *** ccarmack has joined #openstack-ironic | 17:01 | |
| *** alexpilotti has quit IRC | 17:01 | |
| Jabadia | help ! I was asked to create release notes for my commit. even got a link that explains how to. | 17:06 |
| Jabadia | this is what link explains "tox -e venv -- reno new version-foo: | 17:06 |
| Jabadia | what do I put there instead of 'version-foo' ? | 17:06 |
| vdrok | Jabadia: a very short (in a couple of words) of what changes | 17:07 |
| Jabadia | Allright , tnx! | 17:07 |
| vdrok | very short description I mean :) | 17:07 |
| *** alexpilotti has joined #openstack-ironic | 17:07 | |
| JayF | Jabadia: so for instance, I might do reno new add-support-for-chickens | 17:09 |
| JayF | if I were patching IPA to support deploying chickens | 17:09 |
| openstackgerrit | Moshe Levi proposed openstack/ironic-python-agent: Add vendor id and vendor name to interface information https://review.openstack.org/279940 | 17:09 |
| openstackgerrit | Moshe Levi proposed openstack/ironic-python-agent: tinyipa: git clone tags not branches https://review.openstack.org/327825 | 17:09 |
| Jabadia | got it , thanks :) | 17:09 |
| moshele | JayF, sambetts: bug in the tgt qemu cloning it was always clone master because the project use tags not branches I fixed ^^ | 17:10 |
| moshele | JayF, sambetts: let see if CI will pass | 17:10 |
| sambetts | moshele: --branch can take a tag | 17:12 |
| mat128 | TIL IPA will deploy chickens :) | 17:12 |
| mat128 | yet we were discussing deploying toasters the other day :P | 17:12 |
| moshele | sambetts: didn't work on my env | 17:12 |
| *** bsilverman has quit IRC | 17:12 | |
| JayF | mat128: if you can get a chicken to boot a ramdisk I can deploy to it | 17:12 |
| mat128 | :P | 17:13 |
| sambetts | --branch <name> | 17:13 |
| sambetts | -b <name> | 17:13 |
| sambetts | Instead of pointing the newly created HEAD to the branch pointed to by the cloned repository’s HEAD, point to <name> branch instead. In a non-bare repository, this is the branch that will be checked out. --branch can also take tags and detaches the HEAD at that commit in the resulting repository. | 17:13 |
| *** mbound has joined #openstack-ironic | 17:13 | |
| sambetts | moshele: what version of git are you using ? | 17:13 |
| *** vishwanathj has joined #openstack-ironic | 17:14 | |
| Jabadia | last question - I now edit the RN generated in the correct place ( e.g 'fixes' ) and adding what I've done ? then commit and push ? | 17:15 |
| mat128 | Jabadia: amend your commit and 'git review' again | 17:15 |
| Jabadia | K . tnx. | 17:15 |
| *** dprince has joined #openstack-ironic | 17:17 | |
| openstackgerrit | Nisha Agarwal proposed openstack/proliantutils: Fix the logic for fetching the gpu device count https://review.openstack.org/315402 | 17:17 |
| *** mbound has quit IRC | 17:18 | |
| lucasagomes | jroll, replied to that email (please don't punch me heh) | 17:18 |
| lucasagomes | and I'm calling it a day, I'm off tomorrow until tuesday (holidays o/) | 17:19 |
| mat128 | sambetts, moshele: git clone supports --branch since v1.6.5 (Aug 2009): https://github.com/git/git/commit/7a4ee28f41270bf032d0dd0bfb17f601b9b3971a | 17:19 |
| lucasagomes | so I see you all next week! | 17:19 |
| * jroll pushes a pint towards lucasagomes instead of punching | 17:19 | |
| jroll | have a good weekend dude :) | 17:19 |
| moshele | mat128: I use 1.7.1 and it not working Centos 6.7 | 17:20 |
| lucasagomes | jroll, cheers dude :-) enjoy the weekend too | 17:20 |
| moshele | mat128: I mean git 1.7.1 | 17:20 |
| lucasagomes | see you all! | 17:20 |
| *** lucasagomes is now known as lucas-afk | 17:20 | |
| mat128 | moshele: figured. Trying it in a centos container | 17:21 |
| *** derekh has quit IRC | 17:21 | |
| *** xavierr_afk is now known as xavierr | 17:21 | |
| *** ijw has quit IRC | 17:22 | |
| *** ijw has joined #openstack-ironic | 17:22 | |
| xavierr | good afternoon Ironic | 17:22 |
| *** PollyZ has quit IRC | 17:23 | |
| mat128 | moshele: http://paste.openstack.org/show/509285/ | 17:23 |
| mat128 | moshele: said remote branch not found in upstream origin, weird | 17:23 |
| Jabadia | sorry, the RN has many sections, do I just keep all section as is and modifying the 'fixes' one ? or do I delete all section leaving 'fixes' only ? ( or something else ..) | 17:24 |
| sambetts | mat128: dear centos why you break git... | 17:24 |
| NobodyCam | morning xavierr | 17:25 |
| *** ohamada_ has quit IRC | 17:25 | |
| xavierr | NobodyCam :) | 17:25 |
| JayF | Jabadia: you only need the relevant section | 17:25 |
| Jabadia | great. delete all others then ? | 17:25 |
| NobodyCam | :) | 17:25 |
| JayF | https://github.com/openstack/ironic/blob/master/releasenotes/notes/adopt-ironic-context-5e75540dc2b2f009.yaml | 17:25 |
| JayF | being one of many good examples in that dir | 17:26 |
| Jabadia | cool. checking. | 17:26 |
| Jabadia | thanks! | 17:26 |
| moshele | JayF: this is lshw http://www.ezix.org/project/wiki/HardwareLiSter as far as I could find | 17:27 |
| *** Goneri has joined #openstack-ironic | 17:28 | |
| JayF | moshele: hmm. Yeah I've been digging into that myself and tend to agree. Assuming lshw isn't packaged by TCL either? | 17:29 |
| JayF | moshele: that being the case, can you put a comment there linking to the page as the upstream for lshw? That way other folks like me might worry a bit less | 17:29 |
| moshele | JayF: sure | 17:30 |
| *** sdake has joined #openstack-ironic | 17:31 | |
| JayF | moshele: honestly that comment is more a reflection of how ehhh I am in general about how that build is done, and makes me wish we did more hash/gpg checking in the build and all that | 17:32 |
| mat128 | moshele: https://github.com/git/git/blob/master/Documentation/RelNotes/1.7.10.txt#L84 | 17:32 |
| mat128 | this is whats missing for your git 1.7.1 | 17:32 |
| mat128 | brb, 1h | 17:32 |
| *** mat128 is now known as mat128|away | 17:32 | |
| openstackgerrit | Nisha Agarwal proposed openstack/ironic: Add support for building ISO for deploy ramdisk https://review.openstack.org/264590 | 17:33 |
| moshele | sambetts, mat128: do you want me to undo https://review.openstack.org/#/c/327825/ | 17:33 |
| sambetts | moshele, JayF: re: lshw, could it replace or be replaced with dmidecode? | 17:33 |
| moshele | JayF: I check not TCL package | 17:33 |
| *** sdake_ has joined #openstack-ironic | 17:35 | |
| *** aginwala has joined #openstack-ironic | 17:35 | |
| moshele | sambetts: didn't fine with it the Mellanox Nic or pci device | 17:36 |
| *** sdake has quit IRC | 17:37 | |
| openstackgerrit | Nisha Agarwal proposed openstack/ironic: Config variable to configure [glance] section https://review.openstack.org/266803 | 17:37 |
| sambetts | moshele, JayF: if lshw does a superset of dmidecode then we could replace the dmidecode stuff with that right? | 17:37 |
| sambetts | dtantsur|afk: ^^ | 17:37 |
| JayF | What's the value in that? | 17:38 |
| sambetts | we don't have to install 2 packages that do mostly the same thing | 17:38 |
| *** david-lyle has joined #openstack-ironic | 17:38 | |
| *** mtanino has quit IRC | 17:39 | |
| JayF | I mean, I guess I don't care either way that much? But I certainly wouldn't block moshele's patch on it | 17:39 |
| JayF | would be possibly a good low-hanging-fruit bug though | 17:39 |
| JayF | and we always have folks looking for those | 17:39 |
| sambetts | JayF: just want to keep tinyipa tiny ;) | 17:39 |
| JayF | Yeah; but we do also have the gate proving that we aren't making it prohibitavely big as wel | 17:40 |
| JayF | now that all the gates are run in 384 I feel much less squishy about that | 17:40 |
| JayF | this makes me wish we had a 256 experimental job too though | 17:40 |
| * JayF going to write that patch | 17:40 | |
| sambetts | yeah, its the 256mb thats a fine knife edge right now | 17:40 |
| *** Nisha_away has joined #openstack-ironic | 17:40 | |
| sambetts | JayF: the tinyipa src job on IPA is still nv? | 17:41 |
| JayF | not at all | 17:41 |
| sambetts | gate-tempest-dsvm-ironic-pxe_ipa-tinyipa-src-nv | 17:41 |
| sambetts | ^ | 17:41 |
| JayF | https://review.openstack.org/#/c/324909/ is an example of a patch running the new jobs | 17:42 |
| JayF | only -nv job is inspector | 17:42 |
| thiagop-planning | sambetts: that's me | 17:42 |
| thiagop-planning | ops | 17:42 |
| thiagop-planning | that's not me | 17:42 |
| thiagop-planning | (almost equal, sorry) | 17:43 |
| sambetts | JayF: oh cool :) | 17:43 |
| *** Nisha_brb has quit IRC | 17:43 | |
| *** Sukhdev has joined #openstack-ironic | 17:43 | |
| *** athomas has quit IRC | 17:43 | |
| sambetts | JayF: oh wow, the last time the lshw patch had a passing CI run was may 4th thats why its the old jobs | 17:44 |
| JayF | aye | 17:45 |
| jroll | should recheck that :P | 17:45 |
| sambetts | its got a new patchset so its running now :) | 17:45 |
| jroll | coo | 17:45 |
| * jroll requires noms, bbl | 17:45 | |
| moshele | sambetts, JayF: I have to go now, so will work on your comments tomorrow | 17:49 |
| sambetts | moshele: sure, cya tomorrow | 17:49 |
| *** vishwanathj has quit IRC | 17:50 | |
| bcornec | Hello all | 17:50 |
| bcornec | Hello all | 17:50 |
| * sambetts -> home | 17:50 | |
| *** vishwanathj has joined #openstack-ironic | 17:50 | |
| bcornec | Wrt https://bugs.launchpad.net/ironic/+bug/1589627 I see the error in http://logs.openstack.org/79/326079/2/check/gate-ironic-python27-db/790f9a0/console.html, But I'm not understanding how to fix it (complete newbye in python sorry). | 17:51 |
| openstack | Launchpad bug 1589627 in Ironic "Documentation error on raid configuration" [Low,In progress] - Assigned to Bruno Cornec (bruno-cornec) | 17:51 |
| *** trown|lunch is now known as trown | 17:51 | |
| *** moshele has quit IRC | 17:51 | |
| sambetts | Night all | 17:51 |
| thiagop-planning | night sambetts | 17:52 |
| bcornec | It seems to me that the JSON schema is correct, proposing 2 values is_root_volume & share_physical_disks as booleans, and so I don't understand why when I removed the " around the True value in the test it now fails. From the python doc at https://docs.python.org/2/library/stdtypes.html, it seems it should be without " right ? | 17:54 |
| bcornec | It's probably obvious for you all sorry for that... | 17:54 |
| * mgould -> home, good night! | 17:55 | |
| *** mgould is now known as mgould|afk | 17:55 | |
| thiagop-planning | bcornec: this should be raising an exception and isn't | 17:57 |
| thiagop-planning | but isn't* | 17:58 |
| *** serverascode has quit IRC | 17:58 | |
| *** electrofelix has quit IRC | 17:59 | |
| *** serverascode has joined #openstack-ironic | 17:59 | |
| *** mgkwill has quit IRC | 18:00 | |
| *** electrofelix has joined #openstack-ironic | 18:00 | |
| *** alexpilotti has quit IRC | 18:01 | |
| thiagop-planning | bcornec: the change on raid_constants seems to be setting and invalid format to valid, so the exceptions expected to be raised on test_raid aren't being raised at all | 18:01 |
| *** electrofelix has quit IRC | 18:01 | |
| *** mgkwill has joined #openstack-ironic | 18:01 | |
| *** aginwala has quit IRC | 18:01 | |
| *** alexpilotti has joined #openstack-ironic | 18:01 | |
| *** aginwala has joined #openstack-ironic | 18:02 | |
| *** yolanda has quit IRC | 18:02 | |
| *** mtreinish has quit IRC | 18:03 | |
| *** mtreinish has joined #openstack-ironic | 18:03 | |
| bcornec | Excuse me, not sure I got it. Do you mean that the dict declared in ironic/tests/unit/raid_constants.py should be invalid ones to raise an exception ? | 18:04 |
| bcornec | So instead of true it should be an invalid value such as "something" e.g. ? | 18:05 |
| bcornec | Because in fact I changed back a value which was indeed incorrect, but in fact I should have left that one untouched, as it is expected to be incoreect there right ? | 18:05 |
| thiagop-planning | bcornec: yep | 18:06 |
| thiagop-planning | bcornec: the keys are saying that they're invalid | 18:06 |
| bcornec | Then if that's the case, I think it could be more clear to directly put invalid as a value to show it's done on purpose ;-) | 18:06 |
| bcornec | I understand that now you explained it to me ! | 18:06 |
| thiagop-planning | bcornec: the point is, people would think that it could be string True instead of true, both are valid on JSON | 18:07 |
| *** PollyZ has joined #openstack-ironic | 18:07 | |
| *** Sukhdev has quit IRC | 18:08 | |
| thiagop-planning | bcornec: but the expected one is true, so if someone puts string True it shall fail | 18:08 |
| bcornec | Hummm so what is the best appraoch: to put "is_root_volume": "True" (to the risk it's unclear for beginers like me) or "is_root_volume": invalid ? After all it's just a test right ? | 18:08 |
| thiagop-planning | bcornec: I'd leave it as it was. | 18:09 |
| thiagop-planning | :) | 18:09 |
| bcornec | Ok, you're the masters ;-) | 18:09 |
| thiagop-planning | bcornec: not me | 18:09 |
| * thiagop-planning ducks | 18:09 | |
| *** mtreinish has quit IRC | 18:10 | |
| bcornec | Just that when I did a recursive grep to find all invalid "True" values in the code, I found these which are in fact false positive in that case. Distutrbing... | 18:10 |
| *** Nisha_away has quit IRC | 18:12 | |
| *** vishwanathj has quit IRC | 18:12 | |
| *** mtreinish has joined #openstack-ironic | 18:13 | |
| openstackgerrit | Ruby Loo proposed openstack/ironic: Replace dict.get(key) with dict[key] https://review.openstack.org/327726 | 18:14 |
| *** mat128|away is now known as mat128 | 18:14 | |
| mat128 | moshele | 18:14 |
| mat128 | :( | 18:14 |
| *** mtanino has joined #openstack-ironic | 18:15 | |
| *** mgagne has quit IRC | 18:16 | |
| Jabadia | Hi, I have many failure in Jenkins check which ( I hope) are not related to my code change. | 18:16 |
| Jabadia | e.g " stderr: 'fatal: unable to connect to git.openstack.org" | 18:16 |
| Jabadia | should I be concern ? is that normal ? | 18:17 |
| *** mgagne has joined #openstack-ironic | 18:17 | |
| JayF | That looks like a temporary unrelated error | 18:17 |
| JayF | in those cases, you can leave a comment startnig with "recheck" to recheck it | 18:17 |
| JayF | I usually put a statement about why | 18:17 |
| JayF | so like "recheck connection failure to git.openstack.org" would be a good one | 18:17 |
| Jabadia | tnx! | 18:18 |
| JayF | but only recheck once you've checked logs and are sure it's unrelated + temporary (as you did here) | 18:18 |
| openstackgerrit | Yossi Ovadia proposed openstack/ironic-python-agent: Handle diskless hardware connected to remote iscsi. https://review.openstack.org/327807 | 18:19 |
| Jabadia | working on unitest now. | 18:19 |
| *** vishwanathj has joined #openstack-ironic | 18:28 | |
| *** vishwanathj has quit IRC | 18:28 | |
| *** alexpilotti has quit IRC | 18:29 | |
| *** ayoung has quit IRC | 18:33 | |
| *** alexpilotti has joined #openstack-ironic | 18:34 | |
| *** sdake_ has quit IRC | 18:38 | |
| *** bcornec1 has quit IRC | 18:44 | |
| *** ayoung has joined #openstack-ironic | 18:47 | |
| *** aNupoisc has joined #openstack-ironic | 18:49 | |
| *** Sukhdev has joined #openstack-ironic | 18:50 | |
| *** baoli_ has quit IRC | 18:53 | |
| *** anush_ has quit IRC | 18:54 | |
| *** stevebaker has quit IRC | 18:55 | |
| *** stevebaker has joined #openstack-ironic | 18:55 | |
| *** sdake has joined #openstack-ironic | 18:59 | |
| *** e0ne has joined #openstack-ironic | 18:59 | |
| *** vishwanathj has joined #openstack-ironic | 19:02 | |
| *** e0ne has quit IRC | 19:02 | |
| rloo | JayF: do you know, we are still going to support coreos & tinyipa, right? or did we decide to deprecate coreos? | 19:05 |
| rloo | JayF: I might just start a rumour for fun | 19:05 |
| JayF | both are supported and gated. | 19:05 |
| JayF | I want to get DIB gated as well. | 19:05 |
| JayF | The CoreOS image needs some love if it's going to be supported long-term | 19:06 |
| rloo | JayF: OH. do we know how much love? | 19:06 |
| JayF | Mainly it needs an upgrade for $security_reasons | 19:06 |
| JayF | that we avoided doing because it made the gate less reliable. | 19:06 |
| JayF | now that coreos is only run in ipxe jobs, against ipa, I think that's hedged against somewhat | 19:06 |
| rloo | JayF: less reliable == timed out? | 19:06 |
| JayF | and I'd like to reattempt and upgrade | 19:07 |
| JayF | aye | 19:07 |
| rloo | is there a bug open for that? | 19:07 |
| JayF | the image in newer versions is quite larger | 19:07 |
| *** e0ne has joined #openstack-ironic | 19:07 | |
| JayF | Hm. If not there should be | 19:07 |
| JayF | I'll look for one, and if not, file one | 19:07 |
| rloo | thx JayF | 19:07 |
| JayF | but basically the general pattern is: | 19:07 |
| Jabadia | can someone paste tux that tests specific test ? | 19:07 |
| Jabadia | tox* | 19:08 |
| JayF | all supported ramdisks are tested from source in IPA, and get ramdisks built and put on tarballs.openstack | 19:08 |
| rloo | Jabadia: should be documented. sec, let me find it. | 19:08 |
| JayF | the binary version of the ramdisk best suited to the gate (tinyipa today) is used in Ironic jobs | 19:08 |
| rloo | Jabadia: http://docs.openstack.org/developer/ironic/dev/dev-quickstart.html | 19:08 |
| rloo | Jabadia: look for tox -epy27 -- -r | 19:09 |
| JayF | so we could have DIB (possibly even multiple distros of ?), CoreOS, and TinyIPA all gating against IPA only with no pain to Ironic jobs | 19:09 |
| JayF | and since IPA gets a lot less code churn (and that should stay true with the advent of ironic-lib) the extra jobs cause much less pain | 19:09 |
| rloo | JayF: that sounds good to me. are there bugs or something to track all this work that you'd like to see done? | 19:09 |
| Jabadia | Giving a shot. thank rloo | 19:10 |
| JayF | Well I already did most of it, lol | 19:10 |
| JayF | the only piece out of that not done yet is DIB image | 19:10 |
| JayF | but that's more an RFE (officially support DIB) than a bug-bug | 19:10 |
| JayF | I will file the bug about the TLC coreos needs though | 19:10 |
| jroll | JayF: rloo: there's no bug for the coreos version, but here's the patch we had to revert https://review.openstack.org/#/c/191505/ | 19:11 |
| rloo | jroll: oh, right, that patch is bringing back memories. | 19:11 |
| *** aginwala has quit IRC | 19:11 | |
| jroll | heh | 19:11 |
| rloo | jroll: we shoudl have a bug to track it i think. | 19:11 |
| JayF | Yep; I'll file it | 19:12 |
| JayF | and revive that patch | 19:12 |
| rloo | JayF: and we should have an RFE for the DIB thing. | 19:12 |
| JayF | right now | 19:12 |
| JayF | rloo: I'll gladly file it, but to be clear my only care about DIB is that if people are using it we should actually support it. I've never so much as built an IPA DIB image | 19:12 |
| *** harlowja has quit IRC | 19:12 | |
| jroll | we have dib build support in devstack | 19:12 |
| jroll | I know ilo CI uses it | 19:12 |
| rloo | JayF: yeah, you don't need to do the work. | 19:13 |
| jroll | I think UEFI only works on DIB today | 19:13 |
| rloo | JayF: just want to track it somehow so we know it is missing. | 19:13 |
| JayF | jroll: no it doesn't, because we don't test it :D | 19:13 |
| jroll | >.> | 19:13 |
| JayF | all those are now on my todo list, will take care of them shortly | 19:14 |
| JayF | - Bugs for CoreOS TLC | 19:14 |
| JayF | - Patch for upgrade CoreOS to latest stable | 19:14 |
| JayF | - RFC Bug for DIB post job and gate on ipa | 19:14 |
| JayF | was there anything I missed? | 19:14 |
| *** vishwanathj has quit IRC | 19:14 | |
| *** ccarmack has quit IRC | 19:14 | |
| JayF | jroll: did we start filing bugs aganist IPA directly? | 19:15 |
| JayF | jroll: or is https://bugs.launchpad.net/ironic-python-agent/+bug/1590606 in the wrong place | 19:15 |
| openstack | Launchpad bug 1590606 in ironic-python-agent "[RFE] Support remote iSCSI shares as root devices via iscsistart" [Wishlist,In progress] - Assigned to Yossi Ovadia (jabadia) | 19:15 |
| jroll | JayF: we did | 19:15 |
| *** aginwala has joined #openstack-ironic | 19:15 | |
| jroll | RFEs though? dunno if we "officially" did, but it seems okay | 19:15 |
| *** penick has joined #openstack-ironic | 19:15 | |
| *** penick has quit IRC | 19:15 | |
| *** aginwala has quit IRC | 19:16 | |
| jroll | I think we generally said, file RFEs against ironic, and we'll tag the projects as needed, but I don't recall honestly | 19:16 |
| *** aginwala has joined #openstack-ironic | 19:16 | |
| JayF | https://bugs.launchpad.net/ironic-python-agent/+bug/1581673 was already filed around one of the symptoms caused by the image being old | 19:16 |
| openstack | Launchpad bug 1581673 in ironic-python-agent "Shipped CoreOS image is out of date" [High,In progress] - Assigned to Kris Lindgren (klindgren) | 19:16 |
| JayF | so I changed the title and will use it for that work | 19:17 |
| jroll | wfm | 19:17 |
| cinerama | bifrost allows folks the option to build IPA w/dib | 19:17 |
| JayF | it. | 19:17 |
| jroll | it? | 19:18 |
| *** moshele has joined #openstack-ironic | 19:18 | |
| jroll | JayF: might be worth referencing https://review.openstack.org/#/c/326203/ in your DIB RFE, we could revert that if needed and adjust | 19:19 |
| *** sdake has quit IRC | 19:19 | |
| JayF | my internet is screwing up, that's why I had the danglign comment | 19:19 |
| JayF | I think Comcast is having some v6 routing issues today | 19:19 |
| jroll | ah | 19:19 |
| cinerama | we also test in CI with building IPA | 19:19 |
| TheJulia | ++ | 19:20 |
| TheJulia | actually, that should be more \o/ | 19:20 |
| *** jaybeale has quit IRC | 19:20 | |
| TheJulia | except /o\ when mirrors or dib changes that are broken on debian strike | 19:21 |
| JayF | https://bugs.launchpad.net/ironic-python-agent/+bug/1590935 | 19:22 |
| openstack | Launchpad bug 1590935 in ironic-python-agent "DIB image should be officially supported in IPA" [Undecided,New] | 19:22 |
| * JayF adds RFE: prefix to that title | 19:22 | |
| JayF | TheJulia: ^ sounds like you might have extra insight to lend to that | 19:23 |
| jroll | JayF: this might be related to your networking troubles? http://downdetector.com/status/level3 | 19:26 |
| jroll | there's tweets in there that imply comcast having issues | 19:26 |
| TheJulia | JayF: for item #2, are you thinking voting or non-voting? | 19:26 |
| JayF | TheJulia: I think all supported IPA ramdisks should have voting CI. | 19:27 |
| TheJulia | greghaynes: ^^^ you might actually have an opinion on this as well | 19:27 |
| cinerama | JayF: my take on your part 1 is no opinion on #1, #2 is a good | 19:28 |
| * greghaynes perks up | 19:28 | |
| TheJulia | JayF: so, truthfully, I've seen breaking element changes in dib master take 2-3 weeks to get fixes landed for, and transient connectivity issues across clouds for building the images every so often. My understanding this may have been remedied | 19:28 |
| TheJulia | err, fixes for breaking element changes | 19:29 |
| cinerama | TheJulia: that is one reason i've been giving dib a lot of love | 19:29 |
| JayF | TheJulia: Yeah. That's why I haven't gone after this heavily in the past, but it's real problems we should solve. I don't know enough about DIB to have tehcnical ideas about how to fix it | 19:29 |
| TheJulia | so I'm not sure it would be the greatest idea to have it be voting | 19:29 |
| JayF | but I do know that unless we gate on it, we can't rely upon it to work | 19:29 |
| greghaynes | So, I am lacking some context... is there a tl;dr | 19:29 |
| JayF | and given how tightly IPA is tied to it's environment, I don't think having DIB be non-voting is OK | 19:29 |
| TheJulia | greghaynes: we want to officially support an IPA image built with DIB | 19:29 |
| cinerama | greghaynes: bug https://bugs.launchpad.net/ironic-python-agent/+bug/1590935 | 19:29 |
| openstack | Launchpad bug 1590935 in ironic-python-agent "RFE: DIB image should be officially supported in IPA" [Undecided,New] | 19:29 |
| JayF | because that implies we'd be OK with merging a change that breaks a supported ramdisk | 19:29 |
| *** aginwala has quit IRC | 19:30 | |
| JayF | which I'm extremely -1 to if we are supposed to be supporting it | 19:30 |
| TheJulia | JayF: I actually do agree with that as well, I'm wondering if we could put a test job in dib that is voting? | 19:30 |
| greghaynes | ah | 19:30 |
| JayF | I was asked to file the bug :) | 19:30 |
| * JayF gives all the sausagemaking equipment to TheJulia | 19:30 | |
| TheJulia | lol | 19:30 |
| *** Jabadia has quit IRC | 19:30 | |
| * TheJulia changes the letters on the house | 19:30 | |
| TheJulia | err, numbers | 19:30 |
| greghaynes | So what kind of bugs are we hoping to catch with this test? | 19:30 |
| *** MattMan has quit IRC | 19:31 | |
| *** MattMan has joined #openstack-ironic | 19:31 | |
| greghaynes | IME most of the bugs I see that get merged for processes like IPA are things that show up post-boot | 19:31 |
| jroll | JayF: TheJulia: I think that the DIB job should just make sure the image builds okay, so we don't take down DIB CI if we break things | 19:31 |
| jroll | but maybe not given what greg just said :) | 19:32 |
| TheJulia | greghaynes: that might be the past six months, prior to that it was rather rocky where a fedora change would be landed and it would break say the ubuntu use case | 19:32 |
| greghaynes | TheJulia: Yea, so theres absolutely external breakages that happen (and will continue to), but those arent really ipa specific | 19:32 |
| TheJulia | maybe this as not as big of an issue, for now, that cinerama has been paying a much more watchful eye on dib | 19:32 |
| TheJulia | cinerama: thank you by the way, you rock | 19:33 |
| TheJulia | yeah, dib is also a consumer of external things | 19:33 |
| TheJulia | so any of those things can break us | 19:33 |
| TheJulia | and it won't be dib's fault | 19:33 |
| greghaynes | My thinking is theres the class of general dib external breakage bugs, those should be covered by dib tests purely, theres the dib <-> ipa integration at image build time which is super minimal, then most everthing else is just at boot time | 19:33 |
| TheJulia | like... vendor cloud images magically changing | 19:33 |
| greghaynes | so really the two bits I would expect an ipa-specific dib test to cover are either dib <-> ipa integration bits but I am not sure I've ever seen an issue there - its just a pip install | 19:34 |
| TheJulia | I guess this is kind of why I'm perfectly content to leave the bifrost dib build job as non-voting, if a distribution breaks something for a couple days, we're not blocked | 19:34 |
| greghaynes | and copying a couple init scripts | 19:35 |
| openstackgerrit | Jay Faulkner proposed openstack/ironic-python-agent: Use latest CoreOS stable when building https://review.openstack.org/327891 | 19:35 |
| greghaynes | TheJulia: What all does that job do? (I don't know...) | 19:35 |
| greghaynes | does it boot an image or? | 19:35 |
| TheJulia | greghaynes: fires off the dib role to build an IPA image and then builds a guest os image, typically the breakages that I've seen in recent memory have been connectivity failures | 19:36 |
| openstackgerrit | Jay Faulkner proposed openstack/ironic-python-agent: Use latest CoreOS stable when building https://review.openstack.org/327891 | 19:36 |
| TheJulia | we then boot/deploy to ensure that it all works together | 19:36 |
| greghaynes | TheJulia: Yea, and theres a lot of bifrost <-> dib glue there | 19:36 |
| greghaynes | oh, you boot | 19:36 |
| JayF | The bigger concern I have there really is more around added deps to IPA | 19:36 |
| JayF | if we need program X for IPA, we'd have to get that dep merged into DIB before the IPA patch could merge | 19:37 |
| JayF | and we've been burned a lot lately by cross-project review deps | 19:37 |
| TheJulia | yeah, we need to boot to make sure it actually works | 19:37 |
| greghaynes | So ya, my thinking is that an ipa dib test is really only going to be helpful if you actually go through with a boot. What would also be hugely helpful though is more coverage in dib tets for dib bugs users hit, but I dont think that needs to be IPA specific | 19:37 |
| TheJulia | As silly as that may seem, but in CI it feels rare for our deployment to actually fail | 19:37 |
| greghaynes | JayF: I dont think that should ben issue | 19:38 |
| greghaynes | something I have wondering for a while is - why wouldnt the ipa element live in an ironic repo | 19:38 |
| jroll | greghaynes: ++ | 19:38 |
| JayF | greghaynes: <3 | 19:38 |
| TheJulia | JayF: which is kind of why I'm leaving the job in bifrost non-voting, I get an idea if its broken via visability | 19:38 |
| greghaynes | :) | 19:38 |
| TheJulia | that is a good point, we could then gate on it really easilly | 19:38 |
| TheJulia | still have the issue with package consumption possibly breaking the build | 19:39 |
| JayF | TheJulia: would you make sure a summary of some of this chatter ends up in that rfe bug then? | 19:39 |
| jroll | we have that with all build-from-source jobs, though | 19:39 |
| greghaynes | dib is also fine with having two elements with the same name in the element's search path, there are some caveats but as long as youre aware you should be able to do it backwards-compat just fine | 19:39 |
| TheJulia | JayF: absolutely | 19:40 |
| greghaynes | so you could add the element in to one repo and tell users the other is deprecated in dib tree while leaving it there for a cycle or two | 19:40 |
| TheJulia | jroll: true, although the CI images have almost everything pre-cached for git repos | 19:40 |
| jroll | TheJulia: yeah, but coreos build depends on coreos and debian hosting, tinyipa depends on something else | 19:41 |
| TheJulia | greghaynes: that seems rather viable | 19:41 |
| TheJulia | true | 19:41 |
| *** anush_ has joined #openstack-ironic | 19:41 | |
| TheJulia | so if we move it, then that would make things less headachy | 19:41 |
| TheJulia | jroll: absolutely, a little later though | 19:41 |
| greghaynes | oh, unrelated mostly - but I got a PoC working where dib can make a tinyipa image | 19:42 |
| greghaynes | fyi :p | 19:42 |
| cinerama | greghaynes, nice | 19:42 |
| jroll | TheJulia: "later"? | 19:42 |
| TheJulia | \o/ | 19:42 |
| jroll | greghaynes: ORLY | 19:42 |
| jroll | nice :) | 19:42 |
| TheJulia | jroll: in 30-45 minutes, have only so many threads that can concurrently execute | 19:42 |
| JayF | TheJulia: we also have a little bit of experience limiting those failures, like we switched from httpredir to a specific debian mirror and reduced IPA gate breakage significantly | 19:43 |
| jroll | TheJulia: sorry, I was still going on about the tests on building DIB images, thought we were still on that :P | 19:44 |
| jroll | TheJulia: oh I think you were talking about jay's request to update the bug, oops | 19:44 |
| TheJulia | jroll: heh :) | 19:45 |
| * jroll no longer confused | 19:45 | |
| TheJulia | \o/ | 19:45 |
| jroll | \o/ | 19:45 |
| jroll | magnum midcycle might be at CERN, complete with tours, someone find me a business case to go pls | 19:46 |
| TheJulia | uhhhhh | 19:47 |
| *** ijw has quit IRC | 19:47 | |
| JayF | Ironic and Magnum are clearly on a collision course | 19:47 |
| TheJulia | ++ | 19:47 |
| JayF | like two fast protons looking for subatomic particles | 19:47 |
| *** wajdi_ has joined #openstack-ironic | 19:47 | |
| jroll | ha | 19:47 |
| TheJulia | and we need to measure integration points at the subatomic level | 19:47 |
| * jroll wonders what he has started here | 19:48 | |
| TheJulia | the universe | 19:48 |
| jroll | lol | 19:49 |
| *** harlowja has joined #openstack-ironic | 19:49 | |
| JayF | you mean | 19:49 |
| JayF | The universe, maaaannnn | 19:49 |
| *** e0ne has quit IRC | 19:49 | |
| jroll | okay, I think my brain is sufficiently toasted for the day, I'll see you all tomorrow \o | 19:49 |
| *** aarefiev_ has quit IRC | 19:49 | |
| jroll | so much so that I couldn't find a cern-related way to say that | 19:49 |
| TheJulia | JayF: ++ | 19:49 |
| *** wajdi has quit IRC | 19:51 | |
| *** wajdi_ has quit IRC | 19:52 | |
| *** wajdi has joined #openstack-ironic | 19:53 | |
| *** Jabadia has joined #openstack-ironic | 19:55 | |
| Jabadia | I'm working on a bug and someone assigned it to himself. what's the procedure ? | 19:56 |
| jroll | Jabadia: just assign it back to you, maybe leave a note that says you're already working on it | 19:56 |
| rloo | jroll, Jabadia: well, is it clear from the bug that you were working on it? I mean, when did the person assign it to themself and do they have a patch up already? | 19:57 |
| Jabadia | rloo: I already issue a PR which is being reviewed ( ~3 hours ago ) | 19:59 |
| Jabadia | the person assigned it to himself 5 minutes ago | 19:59 |
| Jabadia | anyway, re-assigned to self with a note . | 19:59 |
| jroll | cool, thanks | 19:59 |
| rloo | Jabadia: yeah, that is recent, so should be ok to reassign. | 20:00 |
| Jabadia | Tnx guys | 20:00 |
| rloo | Jabadia: you could also send them email directly in case they don't see your comment. | 20:00 |
| Jabadia | will do. | 20:00 |
| rloo | Jabadia: they should see your comment but one never knows... | 20:00 |
| *** wajdi_ has joined #openstack-ironic | 20:01 | |
| *** alexpilotti has quit IRC | 20:01 | |
| *** alexpilotti has joined #openstack-ironic | 20:02 | |
| *** wajdi has quit IRC | 20:04 | |
| *** dprince has quit IRC | 20:08 | |
| openstackgerrit | Stephanie Miller proposed openstack/bifrost: Unify test playbooks https://review.openstack.org/327912 | 20:10 |
| JayF | Jabadia: that is me | 20:10 |
| JayF | Jabadia: and I have no idea why launchpad assigned that bug to me | 20:10 |
| Jabadia | No problems :) | 20:11 |
| openstackgerrit | Ruby Loo proposed openstack/ironic: Replace dict.get(key) with dict[key] in unit tests https://review.openstack.org/327726 | 20:12 |
| *** Sukhdev has quit IRC | 20:15 | |
| TheJulia | greghaynes: any chance those caveats to the same element name being in the path multiple time being in document someplace? | 20:16 |
| *** jcoufal has quit IRC | 20:20 | |
| TheJulia | JayF: done, and tossed in another cent of context :) | 20:22 |
| greghaynes | TheJulia: hrm, good question | 20:23 |
| TheJulia | I think that is something we would want to likely plaster over the element until we are able to remove the second one | 20:23 |
| TheJulia | but we need to fully understand those caveats | 20:23 |
| TheJulia | :) | 20:23 |
| greghaynes | its pretty easy for me to explain, just seeing if we actually have it written down | 20:24 |
| greghaynes | yea, I don't think it is | 20:24 |
| greghaynes | basically what goes on is that as long as your files are named the same thing then the later element in the element's search path overwrites the files of the earilier element | 20:25 |
| *** jaybeale has joined #openstack-ironic | 20:25 | |
| greghaynes | and really it has nothing to do with the name of the element, any element with a install.d/50-myscript will overwrite an existing install.d/50-myscript that was before it in the elements search path | 20:26 |
| greghaynes | so basically it 'just works' but dont rename files in your element during the deprecation period or you'll get two copies of it | 20:26 |
| TheJulia | Okay, in that case, I'll just copy/paste that into the bug so we have that visibility | 20:26 |
| greghaynes | SGTM | 20:27 |
| *** mbound has joined #openstack-ironic | 20:27 | |
| *** aNupoisc has quit IRC | 20:27 | |
| *** aginwala has joined #openstack-ironic | 20:28 | |
| *** rama_y has quit IRC | 20:28 | |
| *** rama_y has joined #openstack-ironic | 20:29 | |
| TheJulia | greghaynes: thank you! | 20:29 |
| greghaynes | TheJulia: I already had a big blurb typed out so I just sent it | 20:29 |
| greghaynes | I think we raced with a lot of that info | 20:30 |
| *** aNupoisc has joined #openstack-ironic | 20:30 | |
| greghaynes | JayF: You all have some kind of artifact build and then post on tarballs.o.o process, right? | 20:31 |
| greghaynes | JayF: for coreos images? | 20:31 |
| JayF | the post jobs for ironic-python-agent | 20:31 |
| JayF | for coreos and tinyipa images | 20:31 |
| JayF | those binary images are what Ironic tests against | 20:31 |
| greghaynes | do other test jobs then consume those artifacts? | 20:31 |
| greghaynes | awesome | 20:31 |
| greghaynes | just making sure I was remembering right :) | 20:32 |
| greghaynes | so ya, the same could be done for dib | 20:32 |
| JayF | so basically IPA tests from source, and sorta "enforces" the "IPA works regardless of ramdisk" contract | 20:32 |
| JayF | then Ironic tests with TinyIPA (because it's tiny) to make sure the "Ironic works with IPA" contract is fulfilled | 20:32 |
| JayF | I don't think we want there to be any implication that TinyIPA is the 'default' for anything but devstack/ci | 20:32 |
| greghaynes | gotcha | 20:33 |
| greghaynes | the big benefit though is that youre not calling out to the wider internet to build the image during other tests which need tinyipa | 20:33 |
| greghaynes | because theres really no way to make that resilian | 20:33 |
| greghaynes | re, resiliant | 20:33 |
| JayF | exactly | 20:33 |
| JayF | we're hiding all that risk in IPA | 20:33 |
| JayF | which honestly, over time is going to be changing less | 20:33 |
| *** baoli has joined #openstack-ironic | 20:34 | |
| JayF | because ironic-lib is going to get a lot of the disk manipulation code | 20:34 |
| greghaynes | Yea. Theres a few projects who want a similar thing for dib | 20:34 |
| greghaynes | because that pattern is really a must if youre CIing with some image | 20:34 |
| JayF | and honestly those builds on t.o.o are really the only way ipa 'release's at all | 20:35 |
| JayF | like it's not on pypi, etc | 20:35 |
| openstackgerrit | Moshe Levi proposed openstack/ironic-python-agent: Add vendor id and vendor name to interface information https://review.openstack.org/279940 | 20:35 |
| openstackgerrit | Moshe Levi proposed openstack/ironic-python-agent: tinyipa: Fix git clone tags when 1.6.5 < git < 1.7.10 https://review.openstack.org/327825 | 20:35 |
| greghaynes | oh, huh | 20:36 |
| greghaynes | right, we grab it from git in the dib element | 20:36 |
| *** rcernin has joined #openstack-ironic | 20:36 | |
| *** joprovost has quit IRC | 20:40 | |
| *** jaybeale has quit IRC | 20:43 | |
| *** moshele has quit IRC | 20:44 | |
| *** thiagop-planning is now known as thiagop | 20:49 | |
| *** Goneri has quit IRC | 20:51 | |
| *** ametts has quit IRC | 20:52 | |
| *** aginwala has quit IRC | 20:56 | |
| *** aginwala has joined #openstack-ironic | 21:00 | |
| *** aginwala has quit IRC | 21:01 | |
| *** jjohnson2 has quit IRC | 21:06 | |
| *** Jabadia has quit IRC | 21:06 | |
| *** nicodemos is now known as nico-afk | 21:07 | |
| *** aginwala has joined #openstack-ironic | 21:07 | |
| *** PollyZ has quit IRC | 21:09 | |
| *** PollyZ has joined #openstack-ironic | 21:10 | |
| *** ijw has joined #openstack-ironic | 21:11 | |
| mrda | Morning Ironic | 21:11 |
| NobodyCam | morning mrda | 21:12 |
| *** baoli has quit IRC | 21:13 | |
| mrda | hey NobodyCam | 21:13 |
| *** baoli has joined #openstack-ironic | 21:13 | |
| openstackgerrit | Jay Faulkner proposed openstack/ironic-python-agent: Use latest CoreOS stable when building https://review.openstack.org/327891 | 21:14 |
| *** rcernin has quit IRC | 21:15 | |
| NobodyCam | :) | 21:15 |
| openstackgerrit | Jay Faulkner proposed openstack/ironic-python-agent: Use latest CoreOS stable when building https://review.openstack.org/327891 | 21:15 |
| JayF | Ugh | 21:16 |
| JayF | if I pushed a change with the wrong bug number initially | 21:16 |
| JayF | is there a way to get launchpad to stop assigning that (wrong) bug back to me with each patchset? | 21:16 |
| JayF | yes I've fixed the commit message now | 21:16 |
| JayF | this is why jabadia's bug got assigned to me. | 21:17 |
| *** Sukhdev has joined #openstack-ironic | 21:17 | |
| NobodyCam | JayF: seems like Mr. Jenkins is not letting you live it down :p | 21:20 |
| * NobodyCam *ducks* | 21:20 | |
| JayF | I think I got it to simmah down | 21:20 |
| *** wajdi_ has quit IRC | 21:23 | |
| *** trown is now known as trown|outtypewww | 21:24 | |
| openstackgerrit | Ruby Loo proposed openstack/ironic: fix sed strings in developer doc https://review.openstack.org/327968 | 21:26 |
| *** PollyZ has quit IRC | 21:28 | |
| *** ayoung has quit IRC | 21:34 | |
| *** piet has joined #openstack-ironic | 21:34 | |
| *** ayoung has joined #openstack-ironic | 21:35 | |
| *** aginwala has quit IRC | 21:38 | |
| *** mbound has quit IRC | 21:47 | |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [inspector] https://review.openstack.org/301955 | 21:51 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [iscsi] https://review.openstack.org/309615 | 21:51 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [pxe] https://review.openstack.org/309612 | 21:51 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [seamicro] https://review.openstack.org/303113 | 21:51 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [agent] https://review.openstack.org/309603 | 21:51 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [ssh] https://review.openstack.org/303649 | 21:51 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [deploy] https://review.openstack.org/309206 | 21:51 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [ipmi] https://review.openstack.org/301967 | 21:51 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [dhcp] https://review.openstack.org/301830 | 21:51 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [swift] https://review.openstack.org/303653 | 21:51 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [console] https://review.openstack.org/295946 | 21:51 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [virtualbox] https://review.openstack.org/303655 | 21:51 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [glance] https://review.openstack.org/304829 | 21:51 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [DEFAULT] https://review.openstack.org/309070 | 21:51 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [api] https://review.openstack.org/309186 | 21:51 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [database] https://review.openstack.org/301815 | 21:51 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [irmc] https://review.openstack.org/302476 | 21:51 |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic: Centralize config options - [iboot] https://review.openstack.org/301856 | 21:51 |
| *** jaybeale has joined #openstack-ironic | 21:53 | |
| *** wajdi has joined #openstack-ironic | 21:57 | |
| *** amotoki has joined #openstack-ironic | 21:58 | |
| *** wajdi_ has joined #openstack-ironic | 21:58 | |
| *** amotoki has quit IRC | 21:59 | |
| *** amotoki has joined #openstack-ironic | 21:59 | |
| JayF | jlvillal: or anyone else who worked on the grenade stuff around/ | 22:00 |
| *** wajdi has quit IRC | 22:01 | |
| *** cdearborn has quit IRC | 22:03 | |
| *** ayoung has quit IRC | 22:05 | |
| *** alexpilotti has quit IRC | 22:05 | |
| JayF | https://review.openstack.org/#/c/316662 is landing | 22:10 |
| JayF | someone get some of those red solo cups, some chips, and a pinata | 22:11 |
| JayF | grenade party time \o/ | 22:11 |
| thiagop | \o/\o/o\/\o/o\/o\/o\/o\o/\o/ | 22:11 |
| JayF | vsaienko: going to push a commit to your project-config thing to split the grenade and ironic adds into separate patches | 22:12 |
| JayF | vsaienko: so we can get in -nv on ironic asap and let the chatter about grenade voting job not hold up the one we care the most about | 22:12 |
| jlvillal | JayF: Woo hoo! | 22:14 |
| JayF | https://review.openstack.org/319336 Make grenade-dsvm-ironic non voting | 22:15 |
| JayF | https://review.openstack.org/327985 Make grenade-dsvm-ironic -nv on grenade | 22:15 |
| *** piet has quit IRC | 22:16 | |
| *** SiRiuS__ has quit IRC | 22:25 | |
| *** ijw has quit IRC | 22:33 | |
| *** ijw has joined #openstack-ironic | 22:34 | |
| *** Sukhdev has quit IRC | 22:40 | |
| *** mbound has joined #openstack-ironic | 22:48 | |
| JayF | https://review.openstack.org/#/c/327891/ (use CoreOS stable) is running out of room in the ramdisk trying to run in 1024mb (that means the unextracted size of the ramdisk must be >512mb, since by default you get half the ram for tmpfs) | 22:50 |
| JayF | that's pretty crazy | 22:51 |
| thiagop | I'm calling it a day | 22:52 |
| thiagop | see ya! | 22:52 |
| JayF | I wonder what the largest possible VM size is for the gate. Given that's not where the Ironic-proper jobs run I am curious if we could just bump it up beyond 1024mb | 22:52 |
| JayF | thiagop: o/ | 22:52 |
| *** mbound has quit IRC | 22:53 | |
| *** thiagop has quit IRC | 22:53 | |
| *** mtanino has quit IRC | 22:58 | |
| *** ijw has quit IRC | 23:00 | |
| *** baoli has quit IRC | 23:03 | |
| *** davideagnello has joined #openstack-ironic | 23:03 | |
| *** wajdi_ has quit IRC | 23:03 | |
| *** rbudden has quit IRC | 23:04 | |
| *** wajdi has joined #openstack-ironic | 23:04 | |
| JayF | sambetts: https://review.openstack.org/#/c/327994 is relevant to your interests | 23:04 |
| *** davideagnello has quit IRC | 23:04 | |
| openstackgerrit | Jay Faulkner proposed openstack/ironic: DONOTMERGE: Make CoreOS root partition bigger https://review.openstack.org/327995 | 23:11 |
| openstackgerrit | Jay Faulkner proposed openstack/ironic-python-agent: Use latest CoreOS stable when building https://review.openstack.org/327891 | 23:12 |
| * JayF sciencing some coreos | 23:12 | |
| JayF | and with that calling it a day o/ | 23:12 |
| *** jaybeale has quit IRC | 23:21 | |
| * jlvillal does a happy dance to see the last make Grenade patch work get merged | 23:37 | |
| * jlvillal still waits for the make it a non-voting job patch to get merged. | 23:38 | |
| openstackgerrit | Devananda van der Veen proposed openstack/ironic: [WIP] Add keystone policy support to Ironic https://review.openstack.org/325599 | 23:38 |
| * devananda also calls it a day | 23:38 | |
| mrda | night devananda! | 23:40 |
| *** wrouesnel1 has joined #openstack-ironic | 23:41 | |
| *** ijw has joined #openstack-ironic | 23:42 | |
| *** wrouesnel1 has quit IRC | 23:45 | |
| *** chlong has quit IRC | 23:45 | |
| openstackgerrit | Merged openstack/ironic: Document packing and unpacking the deploy ramdisk https://review.openstack.org/326995 | 23:49 |
| *** rbudden has joined #openstack-ironic | 23:56 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!