| *** rpioso has quit IRC | 00:00 | |
| *** rama_y_ has quit IRC | 00:03 | |
| *** nicodemos has quit IRC | 00:04 | |
| *** mtanin___ has joined #openstack-ironic | 00:26 | |
| *** mtanino has quit IRC | 00:28 | |
| *** lindycoder has joined #openstack-ironic | 00:43 | |
| *** hw_wutianwei has joined #openstack-ironic | 00:44 | |
| openstackgerrit | Merged openstack/python-ironicclient: Updated from global requirements https://review.openstack.org/405887 | 00:46 |
|---|---|---|
| *** ijw has quit IRC | 00:49 | |
| *** baoli has joined #openstack-ironic | 00:50 | |
| *** ijw_ has joined #openstack-ironic | 00:51 | |
| *** SerenaFeng has joined #openstack-ironic | 00:53 | |
| *** zhs__ has joined #openstack-ironic | 00:53 | |
| *** baoli has quit IRC | 00:55 | |
| *** hoangcx_ has joined #openstack-ironic | 00:55 | |
| *** ijw_ has quit IRC | 00:55 | |
| *** vsaienko has quit IRC | 00:56 | |
| *** zhs_ has quit IRC | 00:57 | |
| *** rbudden has quit IRC | 01:00 | |
| *** yufei_ has quit IRC | 01:04 | |
| *** aNuposic has quit IRC | 01:04 | |
| *** ijw has joined #openstack-ironic | 01:06 | |
| *** rama_y has joined #openstack-ironic | 01:08 | |
| *** zhangjl has joined #openstack-ironic | 01:10 | |
| *** yuanying has quit IRC | 01:11 | |
| *** ijw has quit IRC | 01:11 | |
| *** yuanying has joined #openstack-ironic | 01:11 | |
| *** fragatin_ has joined #openstack-ironic | 01:12 | |
| *** tuanluong has joined #openstack-ironic | 01:12 | |
| *** fragatin_ has quit IRC | 01:14 | |
| *** fragatina has quit IRC | 01:15 | |
| *** fragatina has joined #openstack-ironic | 01:15 | |
| *** jcoufal has quit IRC | 01:15 | |
| *** yufei_ has joined #openstack-ironic | 01:17 | |
| *** ijw has joined #openstack-ironic | 01:22 | |
| *** ijw has quit IRC | 01:27 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ironic-ui: Updated from global requirements https://review.openstack.org/409972 | 01:28 |
| *** srobert has joined #openstack-ironic | 01:33 | |
| *** srobert has quit IRC | 01:33 | |
| *** srobert has joined #openstack-ironic | 01:33 | |
| *** rajinir has quit IRC | 01:36 | |
| *** baoli has joined #openstack-ironic | 01:36 | |
| *** Sukhdev has joined #openstack-ironic | 01:38 | |
| *** ijw has joined #openstack-ironic | 01:38 | |
| *** Sukhdev has quit IRC | 01:38 | |
| *** zhangjl has quit IRC | 01:43 | |
| *** ijw has quit IRC | 01:43 | |
| *** aNuposic has joined #openstack-ironic | 01:46 | |
| *** srobert has quit IRC | 01:48 | |
| *** hoangcx has quit IRC | 01:48 | |
| *** ijw has joined #openstack-ironic | 01:54 | |
| *** dsneddon has quit IRC | 01:55 | |
| *** rama_y has quit IRC | 01:58 | |
| *** ijw has quit IRC | 01:59 | |
| *** mtanino has joined #openstack-ironic | 02:03 | |
| *** ijw has joined #openstack-ironic | 02:03 | |
| *** mtanin___ has quit IRC | 02:04 | |
| *** fragatina has quit IRC | 02:06 | |
| *** fragatina has joined #openstack-ironic | 02:07 | |
| *** zhangjl has joined #openstack-ironic | 02:09 | |
| *** baoli has quit IRC | 02:10 | |
| *** yufei has joined #openstack-ironic | 02:11 | |
| *** yuanying has quit IRC | 02:11 | |
| *** Syed__ has quit IRC | 02:15 | |
| *** jkilpatr has quit IRC | 02:21 | |
| *** rloo has quit IRC | 02:22 | |
| *** gcb has joined #openstack-ironic | 02:25 | |
| *** aNuposic has quit IRC | 02:34 | |
| *** mtanino has quit IRC | 02:39 | |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic: Generic power interface for soft reboot and soft power off https://review.openstack.org/216730 | 02:44 |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic: Ipmitool power driver for soft reboot and soft power off https://review.openstack.org/216738 | 02:45 |
| *** chlong has quit IRC | 02:45 | |
| *** ijw has quit IRC | 02:46 | |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic: iRMC power driver for soft reboot and soft power off https://review.openstack.org/216743 | 02:47 |
| openstackgerrit | Naohiro Tamura proposed openstack/ironic: Update the existing APIs due to adding get_supported_power_states https://review.openstack.org/382194 | 02:48 |
| *** gcb has quit IRC | 02:56 | |
| *** chlong has joined #openstack-ironic | 02:59 | |
| *** zhangjl1 has joined #openstack-ironic | 03:09 | |
| *** zhangjl has quit IRC | 03:10 | |
| *** gcb has joined #openstack-ironic | 03:13 | |
| *** pmannidi_ has joined #openstack-ironic | 03:18 | |
| *** pmannidi has quit IRC | 03:19 | |
| *** nicodemos has joined #openstack-ironic | 03:26 | |
| *** raginbajin has quit IRC | 03:27 | |
| *** raginbajin has joined #openstack-ironic | 03:32 | |
| *** vikrant has joined #openstack-ironic | 03:35 | |
| *** harlowja has quit IRC | 03:43 | |
| *** nicodemos has quit IRC | 03:53 | |
| *** nicodemos has joined #openstack-ironic | 03:54 | |
| *** lindycoder has quit IRC | 04:00 | |
| tuanluong | Morning Ironic | 04:04 |
| *** adreznec has quit IRC | 04:18 | |
| *** adreznec has joined #openstack-ironic | 04:21 | |
| *** harlowja has joined #openstack-ironic | 04:21 | |
| *** SerenaFeng has quit IRC | 04:22 | |
| *** links has joined #openstack-ironic | 04:22 | |
| *** ijw has joined #openstack-ironic | 04:25 | |
| *** ijw has quit IRC | 04:30 | |
| *** yuanying has joined #openstack-ironic | 04:33 | |
| *** hoangcx_ is now known as hoangcx | 04:42 | |
| *** deray has joined #openstack-ironic | 04:47 | |
| *** amotoki has joined #openstack-ironic | 04:55 | |
| *** harlowja has quit IRC | 05:08 | |
| *** moshele has joined #openstack-ironic | 05:18 | |
| *** krtaylor has quit IRC | 05:19 | |
| *** fragatina has quit IRC | 05:26 | |
| *** fragatina has joined #openstack-ironic | 05:26 | |
| *** SerenaFeng has joined #openstack-ironic | 05:39 | |
| *** harlowja has joined #openstack-ironic | 05:56 | |
| *** vsaienko has joined #openstack-ironic | 05:57 | |
| *** nmathew has joined #openstack-ironic | 05:57 | |
| *** vsaienko has quit IRC | 05:59 | |
| *** tuanluong has quit IRC | 06:01 | |
| *** hoangcx has quit IRC | 06:01 | |
| *** hoangcx has joined #openstack-ironic | 06:01 | |
| *** tuanluong has joined #openstack-ironic | 06:01 | |
| *** SerenaFeng has quit IRC | 06:03 | |
| *** gcb has quit IRC | 06:04 | |
| *** yuanying has quit IRC | 06:12 | |
| *** yuanying has joined #openstack-ironic | 06:12 | |
| *** yuanying has quit IRC | 06:12 | |
| *** yuanying has joined #openstack-ironic | 06:13 | |
| *** SerenaFeng has joined #openstack-ironic | 06:13 | |
| *** yuanying has quit IRC | 06:17 | |
| openstackgerrit | zhangyanying proposed openstack/ironic: Modify the "qemu-kvm" excutable path for RedHat/CentOS. https://review.openstack.org/410052 | 06:18 |
| *** lindycoder has joined #openstack-ironic | 06:20 | |
| *** gcb has joined #openstack-ironic | 06:28 | |
| *** lindycoder has quit IRC | 06:29 | |
| *** vsaienko has joined #openstack-ironic | 06:32 | |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: DON NOT REVIEW https://review.openstack.org/408195 | 06:33 |
| *** links has quit IRC | 06:34 | |
| *** links has joined #openstack-ironic | 06:39 | |
| *** vsaienko has quit IRC | 06:42 | |
| *** amotoki has quit IRC | 06:52 | |
| *** amotoki has joined #openstack-ironic | 06:52 | |
| *** fragatin_ has joined #openstack-ironic | 07:05 | |
| *** fragatin_ has quit IRC | 07:06 | |
| *** fragatin_ has joined #openstack-ironic | 07:07 | |
| *** fragatina has quit IRC | 07:08 | |
| *** fxpester has joined #openstack-ironic | 07:09 | |
| *** fragatin_ has quit IRC | 07:11 | |
| openstackgerrit | Merged openstack/ironic-ui: Imported Translations from Zanata https://review.openstack.org/403336 | 07:15 |
| *** gcb has quit IRC | 07:17 | |
| *** dsneddon has joined #openstack-ironic | 07:19 | |
| *** zhs__ has quit IRC | 07:22 | |
| *** zhs has joined #openstack-ironic | 07:23 | |
| *** harlowja has quit IRC | 07:25 | |
| pas-ha | morning all :) | 07:27 |
| *** gcb has joined #openstack-ironic | 07:28 | |
| *** vsaienko has joined #openstack-ironic | 07:30 | |
| tuanluong | morning ironic, pas-ha | 07:33 |
| pas-ha | morning tuanluong | 07:33 |
| tuanluong | pas-ha, I have a problem when deploy an instance follow http://docs.openstack.org/developer/ironic/dev/dev-quickstart.html#deploying-ironic-with-devstack | 07:34 |
| pas-ha | tuanluong: what are the symptoms? | 07:34 |
| tuanluong | When i deploy cirrros image. it succesfull, but when i create ubuntu image using disk-image-buider | 07:35 |
| tuanluong | i have "active" but in VM | 07:35 |
| tuanluong | not bootable | 07:36 |
| openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic: Remove agent vendor passthru from OneView drivers https://review.openstack.org/397846 | 07:36 |
| openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic: Remove iBoot, WoL and AMT drivers https://review.openstack.org/397847 | 07:36 |
| openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic: Remove agent vendor passthru completely https://review.openstack.org/397848 | 07:36 |
| tuanluong | Do you have similar experiences? | 07:36 |
| *** Romanenko_K has joined #openstack-ironic | 07:40 | |
| *** sacharya has quit IRC | 07:42 | |
| *** vsaienko has quit IRC | 07:44 | |
| tuanluong | I'm following http://docs.openstack.org/developer/ironic/kilo/deploy/install-guide.html#image-requirements to create ubuntu image | 07:44 |
| tuanluong | Hello pas-ha | 07:45 |
| *** vsaienko has joined #openstack-ironic | 07:45 | |
| *** ralonsoh has joined #openstack-ironic | 07:45 | |
| tuanluong | hello vsaienk0, | 07:46 |
| pas-ha | tuanluong: are you following the netboot procedure? with kernel/initrd also created and uploaded to glance? | 07:48 |
| *** moshele has quit IRC | 07:48 | |
| tuanluong | pas-ha, I use whole disk-image | 07:48 |
| tuanluong | i just use image.qcow2 | 07:48 |
| pas-ha | it seems the above will actually create an image suitable for netboot only | 07:49 |
| tuanluong | thus, how cirros-image successfull | 07:49 |
| tuanluong | I will try with partitions image later | 07:49 |
| pas-ha | the image you've created might be missing the bootloader installed into boot sector | 07:50 |
| pas-ha | cirros whole-disk has it | 07:50 |
| tuanluong | pas-ha, I thinks the demo using cirros is whole-disk-iamge.? | 07:51 |
| pas-ha | yes, but the image you are building with those instructions is not | 07:51 |
| *** yuanying has joined #openstack-ironic | 07:51 | |
| pas-ha | try adding a 'bootloader' dib element when building | 07:51 |
| pas-ha | the default images in devstack with ironic are both whole-disk (the one ending with -disk) and partition/netboot (the one ending with -uec) | 07:52 |
| tuanluong | pas-ha Thanks. disk-image-create ubuntu baremetal dhcp-all-interfaces bootloader -o my-image | 07:52 |
| pas-ha | yep | 07:52 |
| tuanluong | is that correct. thanks you | 07:53 |
| *** vsaienko has quit IRC | 07:55 | |
| pas-ha | I think it should work | 07:57 |
| openstackgerrit | Moshe Levi proposed openstack/ironic-inspector: Adding InfiniBand Support https://review.openstack.org/264257 | 07:57 |
| *** ccamacho has joined #openstack-ironic | 07:59 | |
| *** jaosorior has joined #openstack-ironic | 07:59 | |
| *** sacharya has joined #openstack-ironic | 08:00 | |
| *** rbartal has joined #openstack-ironic | 08:00 | |
| *** rcernin has joined #openstack-ironic | 08:05 | |
| *** sacharya has quit IRC | 08:05 | |
| *** pcaruana has joined #openstack-ironic | 08:10 | |
| *** e0ne has joined #openstack-ironic | 08:11 | |
| *** jaosorior has quit IRC | 08:12 | |
| *** jaosorior has joined #openstack-ironic | 08:12 | |
| *** milan has joined #openstack-ironic | 08:16 | |
| *** rcernin has quit IRC | 08:22 | |
| *** rcernin has joined #openstack-ironic | 08:25 | |
| openstackgerrit | Galyna Zholtkevych proposed openstack/ironic: Make _get_sensors_data concurrent https://review.openstack.org/407429 | 08:26 |
| *** amotoki_ has joined #openstack-ironic | 08:29 | |
| *** kodokuuu has joined #openstack-ironic | 08:29 | |
| *** amotoki has quit IRC | 08:32 | |
| *** milan has quit IRC | 08:36 | |
| openstackgerrit | Vasyl Saienko proposed openstack/python-ironicclient: Add python API and CLI for port groups https://review.openstack.org/335964 | 08:42 |
| *** mjura has joined #openstack-ironic | 08:45 | |
| *** priteau has joined #openstack-ironic | 08:47 | |
| openstackgerrit | Vasyl Saienko proposed openstack/python-ironicclient: Add more tests to node_shell https://review.openstack.org/410100 | 08:49 |
| *** kodokuuu has quit IRC | 08:54 | |
| *** jpich has joined #openstack-ironic | 08:57 | |
| *** zzzeek has quit IRC | 09:00 | |
| *** ohamada has joined #openstack-ironic | 09:00 | |
| *** zzzeek has joined #openstack-ironic | 09:01 | |
| *** athomas has joined #openstack-ironic | 09:02 | |
| openstackgerrit | Vasyl Saienko proposed openstack/python-ironicclient: Remove 'id' from API representation of resources https://review.openstack.org/410112 | 09:11 |
| *** moshele has joined #openstack-ironic | 09:11 | |
| *** fxpester has quit IRC | 09:13 | |
| *** fxpester has joined #openstack-ironic | 09:14 | |
| openstackgerrit | Vasyl Saienko proposed openstack/python-ironicclient: Fix API object representation in unittests https://review.openstack.org/410112 | 09:23 |
| lucas-afk | morning all | 09:32 |
| *** lucas-afk is now known as lucasagomes | 09:32 | |
| *** amotoki has joined #openstack-ironic | 09:36 | |
| *** derekh has joined #openstack-ironic | 09:41 | |
| *** yuanying has quit IRC | 09:47 | |
| *** dtantsur|afk is now known as dtantsur | 09:47 | |
| dtantsur | morning Ironic | 09:47 |
| tuanluong | morning dtantsur, lucasagomes | 09:50 |
| *** milan has joined #openstack-ironic | 09:50 | |
| lucasagomes | o/ | 09:55 |
| milan | morning Ironic! :) | 09:57 |
| milan | morning lucasagomes! :) | 09:57 |
| lucasagomes | hey there | 09:59 |
| aarefiev | morning lucasagomes, dtantsur, milan | 10:00 |
| *** amotoki has quit IRC | 10:00 | |
| milan | morning aarefiev! :) | 10:00 |
| milan | morning dtantsur! :) | 10:00 |
| dtantsur | o/ | 10:08 |
| *** e0ne has quit IRC | 10:09 | |
| *** SerenaFeng has quit IRC | 10:09 | |
| *** e0ne has joined #openstack-ironic | 10:10 | |
| *** links has quit IRC | 10:16 | |
| joanna | morning :) | 10:17 |
| *** hoangcx has quit IRC | 10:21 | |
| dtantsur | aarefiev, is inspector CI still down? | 10:21 |
| aarefiev | dtantsur: fix is merged, let see | 10:22 |
| dtantsur | Inspector has received supports-upgrade and follows-standard-deprecation tags \o/ | 10:23 |
| aarefiev | \o/ | 10:23 |
| openstackgerrit | Yuriy Zveryanskyy proposed openstack/ironic: Add ironic resources CRUD notifications https://review.openstack.org/356541 | 10:24 |
| aarefiev | dtantsur: btw what is the process of updating repository tags | 10:25 |
| aarefiev | we have merged | 10:25 |
| dtantsur | aarefiev, you mean https://review.openstack.org/406100 ? | 10:26 |
| patchbot | patch 406100 - governance - Claim supports-upgrade and follows-standard-deprec... (MERGED) | 10:26 |
| openstackgerrit | Yuriy Zveryanskyy proposed openstack/ironic: Add node maintenance notifications https://review.openstack.org/396239 | 10:27 |
| aarefiev | dtantsur: sorry I mean this badges | 10:28 |
| aarefiev | graphicals | 10:28 |
| dtantsur | they're pulled from this governance repo, so I guess we should see new ones soon | 10:28 |
| aarefiev | ahh, cool | 10:29 |
| dtantsur | dunno, I see different things on github and in the docs.. | 10:29 |
| *** links has joined #openstack-ironic | 10:30 | |
| openstackgerrit | Yuriy Zveryanskyy proposed openstack/ironic: Add node console notifications https://review.openstack.org/397812 | 10:30 |
| vdrok | morning ironic, pas-ha tuanluong lucasagomes dtantsur aarefiev and milan ! | 10:41 |
| lucasagomes | o/ | 10:41 |
| milan | morning joana, vdrok! :) | 10:42 |
| vdrok | morning joanna | 10:42 |
| aarefiev | morning joanna, vdrok | 10:43 |
| openstackgerrit | Aparna proposed openstack/ironic-specs: In-band hpsum firmware update for iLO drivers https://review.openstack.org/410161 | 10:47 |
| openstackgerrit | Aparna proposed openstack/ironic-specs: In-band hpsum firmware update for iLO drivers https://review.openstack.org/410161 | 10:50 |
| openstackgerrit | Galyna Zholtkevych proposed openstack/ironic: Make _get_sensors_data concurrent https://review.openstack.org/407429 | 10:56 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic: Introduce generic hardware types https://review.openstack.org/400678 | 10:56 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic: Support defining and loading hardware types https://review.openstack.org/336626 | 10:56 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic: [WIP] Load hardware types in the conductor https://review.openstack.org/409812 | 10:58 |
| dtantsur | rebase time \o/ | 10:58 |
| *** amotoki_ has quit IRC | 11:01 | |
| *** mgould|afk is now known as mgould | 11:01 | |
| mgould | morning Ironic, pas-ha tuanluong lucasagomes dtantsur aarefiev milan vdrok joanna | 11:01 |
| milan | morning mgould! :) | 11:02 |
| *** daemontool has joined #openstack-ironic | 11:02 | |
| aarefiev | o/ | 11:02 |
| *** MattMan has quit IRC | 11:04 | |
| *** MattMan has joined #openstack-ironic | 11:04 | |
| *** pester has joined #openstack-ironic | 11:05 | |
| *** zhangjl1 has quit IRC | 11:07 | |
| *** fxpester has quit IRC | 11:08 | |
| *** nmathew has quit IRC | 11:08 | |
| dtantsur | hey mgould | 11:09 |
| mgould | hi dtantsur | 11:09 |
| openstackgerrit | Joanna Taryma proposed openstack/ironic: Fail ironic startup if no protocol prefix in ironic api address https://review.openstack.org/404975 | 11:11 |
| *** zhugaoxiao has joined #openstack-ironic | 11:13 | |
| *** SerenaFeng has joined #openstack-ironic | 11:15 | |
| *** links has quit IRC | 11:15 | |
| openstackgerrit | zhangyanying proposed openstack/ironic: Add a "qemu-kvm" excutable path for RedHat/CentOS. https://review.openstack.org/410052 | 11:16 |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Make CONF.debug also reflect on IPA https://review.openstack.org/410168 | 11:16 |
| *** deray has quit IRC | 11:17 | |
| *** SerenaFeng has quit IRC | 11:18 | |
| *** SerenaFeng has joined #openstack-ironic | 11:19 | |
| *** jkilpatr has joined #openstack-ironic | 11:22 | |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Do not wait Neutron by default for SSH power https://review.openstack.org/410170 | 11:25 |
| *** amoralej|off is now known as amoralej | 11:25 | |
| vdrok | morning mgould :) | 11:26 |
| *** SerenaFeng has quit IRC | 11:32 | |
| *** links has joined #openstack-ironic | 11:32 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic-inspector: Use the device hints matching mechanism from ironic-lib https://review.openstack.org/408552 | 11:51 |
| *** jkilpatr has quit IRC | 11:51 | |
| *** zhugaoxiao has quit IRC | 11:56 | |
| dtantsur | inspector CI is back up, good! | 11:56 |
| dtantsur | milan, btw https://review.openstack.org/#/c/409789/ was by your request looong time ago | 11:56 |
| patchbot | patch 409789 - python-ironic-inspector-client - Clarify that node names can be used in addition to... | 11:56 |
| *** zhugaoxiao has joined #openstack-ironic | 11:56 | |
| milan | dtantsur, I don't remember O:-) /me looks | 11:57 |
| dtantsur | it was months ago, really :) I found a TODO item in my personal task list yesterday | 11:57 |
| milan | dtantsur, aaah :D | 11:58 |
| milan | OK | 11:58 |
| * milan adds himself to review | 11:58 | |
| * milan right now updates the list statuses patch according to aarefiev 's review ;) | 11:58 | |
| dtantsur | yeah, no hurry with this one | 12:00 |
| robcresswell | Hey guys, do you have a microversion history like http://docs.openstack.org/developer/nova/api_microversion_history.html ? | 12:05 |
| robcresswell | ah, found it, nvm | 12:06 |
| lucasagomes | :-) | 12:07 |
| robcresswell | Mucking around with the standalone ironic UI :) | 12:07 |
| *** tuanluong has quit IRC | 12:07 | |
| *** nmathew has joined #openstack-ironic | 12:07 | |
| openstackgerrit | Vasyl Saienko proposed openstack/ironic: Do not wait Neutron by default with *_ssh driver https://review.openstack.org/410170 | 12:10 |
| *** jkilpatr has joined #openstack-ironic | 12:10 | |
| *** jaosorior is now known as jaosorior_brb | 12:10 | |
| *** SerenaFeng has joined #openstack-ironic | 12:12 | |
| *** dprince has joined #openstack-ironic | 12:23 | |
| *** SerenaFeng has quit IRC | 12:25 | |
| *** yufei has quit IRC | 12:35 | |
| dtantsur | "CoreOS Linux is Now Container Linux" https://coreos.com/blog/tectonic-self-driving.html | 12:41 |
| dtantsur | soon we'll have to rename some jobs :D | 12:42 |
| *** krtaylor has joined #openstack-ironic | 12:44 | |
| *** hw_wutianwei has quit IRC | 12:46 | |
| *** vikrant has quit IRC | 12:52 | |
| *** dprince has quit IRC | 12:52 | |
| *** lucasagomes is now known as lucas-hungry | 12:58 | |
| *** eroux has joined #openstack-ironic | 13:05 | |
| nicodemos | good morning! | 13:08 |
| *** bfournie has quit IRC | 13:08 | |
| openstackgerrit | Galyna Zholtkevych proposed openstack/ironic: [WIP] ETAG supporting to enhance API evolution https://review.openstack.org/392213 | 13:10 |
| *** lindycoder has joined #openstack-ironic | 13:17 | |
| *** amotoki has joined #openstack-ironic | 13:18 | |
| openstackgerrit | Szymon Borkowski proposed openstack/ironic: Add RPC and object version pinning https://review.openstack.org/407491 | 13:20 |
| *** lindycoder has quit IRC | 13:21 | |
| *** lindycoder has joined #openstack-ironic | 13:23 | |
| TheJulia | Good morning! | 13:29 |
| TheJulia | dtantsur: to have minimal OS references I hope :) | 13:29 |
| milan | morning nicodemos, TheJulia! :) | 13:29 |
| *** lindycoder has quit IRC | 13:30 | |
| dtantsur | TheJulia, morning :) | 13:30 |
| * milan not sure how that would work; giving them numbers the cattle way I guess ;) | 13:31 | |
| milan | anyone knows how to use https://github.com/openstack/osc-lib/blob/master/osc_lib/utils.py#L248 properly? | 13:31 |
| * milan getting frustrated w/ http://paste.openstack.org/show/592217/ | 13:32 | |
| milan | where ('U', 't', 'i', 'r') comes from I don't follow | 13:33 |
| milan | the fields I feed the function with are: ('UUID', 'Started at', 'Finished at', 'Error') | 13:33 |
| * milan starts to think about blaming his Mac for it | 13:34 | |
| milan | dtantsur, ^ ;) | 13:34 |
| TheJulia | milan: change the test order maybe? | 13:34 |
| dtantsur | I highly suspect it, yes | 13:34 |
| *** trown|outtypewww is now known as trown | 13:35 | |
| * milan tries changing test order, then blaming the Mac :D | 13:35 | |
| dtantsur | I think input is mismatched somewhere | 13:35 |
| dtantsur | hard to tell without looking at the code | 13:35 |
| milan | yeah most likely | 13:37 |
| milan | I'll push it I guess | 13:37 |
| milan | actually, TheJulia was right, just changed the order of the tests and it works O.o | 13:37 |
| * milan tries changing it back and forth | 13:38 | |
| milan | yup, depends on the order :-/ | 13:39 |
| dtantsur | UGH | 13:40 |
| milan | coffee nr. 5 | 13:41 |
| vdrok | morning nicodemos and TheJulia | 13:41 |
| vdrok | woah milan, you know the max for the day is 50? :) | 13:42 |
| milan | meh, I still have 45 to go then, vdrok :D | 13:43 |
| openstackgerrit | Szymon Borkowski proposed openstack/ironic: Add RPC and object version pinning https://review.openstack.org/407491 | 13:43 |
| vdrok | milan: also futurama says that after 100 you'll be able to slow down the time | 13:43 |
| vdrok | if you survive | 13:44 |
| milan | I'm definitely trying that one day :D | 13:44 |
| *** bfournie has joined #openstack-ironic | 13:44 | |
| *** amoralej is now known as amoralej|lunch | 13:44 | |
| milan | actually, slowing down the time to see the race condition in the test case might be helpful now :) | 13:45 |
| milan | but I've got a strong feeling this actually going to be one of those between the keyboard and chair issues ;) | 13:45 |
| *** amotoki has quit IRC | 13:45 | |
| *** srobert has joined #openstack-ironic | 13:46 | |
| mgould | morning nicodemos TheJulia | 13:46 |
| mgould | milan, what you actually need is some cake: https://www.youtube.com/watch?v=Xbq3kc29Tmg | 13:47 |
| nicodemos | hey, vdrok, milan, TheJulia, mgould | 13:47 |
| mgould | but beware of Czech Neck | 13:48 |
| *** jkilpatr has quit IRC | 13:49 | |
| milan | :D lol | 13:49 |
| *** hoangcx has joined #openstack-ironic | 13:50 | |
| *** srobert has quit IRC | 13:50 | |
| *** jheroux has joined #openstack-ironic | 13:51 | |
| *** hoangcx has quit IRC | 13:51 | |
| *** hoangcx has joined #openstack-ironic | 13:52 | |
| xavierr | good morning Ironic'ers o/ | 13:55 |
| mgould | morning xavierr | 13:55 |
| milan | morning xavierr! :) | 13:55 |
| *** glonlas has joined #openstack-ironic | 13:56 | |
| openstackgerrit | Hugo Nicodemos proposed openstack/ironic: Onetime boot when set_boot_device isn't persistent https://review.openstack.org/340596 | 13:57 |
| *** jcoufal has joined #openstack-ironic | 13:57 | |
| *** glonlas has quit IRC | 13:57 | |
| *** glonlas has joined #openstack-ironic | 13:59 | |
| *** glonlas has quit IRC | 13:59 | |
| vdrok | morning xavierr | 14:00 |
| *** glonlas has joined #openstack-ironic | 14:00 | |
| *** dprince has joined #openstack-ironic | 14:02 | |
| *** sborkows has joined #openstack-ironic | 14:05 | |
| openstackgerrit | Kyrylo Romanenko proposed openstack/python-ironicclient: Add basic tests for OSC plugin baremetal driver commands https://review.openstack.org/367359 | 14:05 |
| *** jkilpatr has joined #openstack-ironic | 14:06 | |
| *** jkilpatr_ has joined #openstack-ironic | 14:08 | |
| *** lucas-hungry is now known as lucasagomes | 14:08 | |
| *** jaosorior_brb is now known as jaosorior | 14:09 | |
| * milan asks around for some cake | 14:09 | |
| xavierr | milan: lol | 14:09 |
| openstackgerrit | Hugo Nicodemos proposed openstack/ironic: Onetime boot when set_boot_device isn't persistent https://review.openstack.org/340596 | 14:10 |
| *** rbudden has joined #openstack-ironic | 14:10 | |
| *** jkilpatr has quit IRC | 14:10 | |
| mrtenio | Morning Ironic. | 14:11 |
| mgould | morning mrtenio | 14:12 |
| milan | morning mrtenio! :) | 14:12 |
| mgould | milan: the guy in that clip who said "should we ask a question about this in Parliament?" was an actual sitting MP, known for his hard-on-drugs stances | 14:12 |
| xavierr | hey mrtenio :) | 14:13 |
| milan | mgould, isn't it that one who actually thought that was a real thing? | 14:13 |
| *** Goneri has joined #openstack-ironic | 14:15 | |
| TheJulia | nicodemos: slight nits on the release note, if any make sense to you, you may want to put up a follow-up patch. | 14:16 |
| TheJulia | or, you could revise it if you want, would be easy for us to toss new +2's on it :) | 14:17 |
| vdrok | dtantsur: lucasagomes TheJulia http://logs.openstack.org/70/410170/2/check/gate-tempest-dsvm-ironic-ipa-wholedisk-agent_ipmitool-tinyipa-ubuntu-xenial/8267b81/console.html I've started seeing this recently, the reason is - set console mode is async, and seems it slowed down a bit recently, so that consequent get still gets old value | 14:17 |
| TheJulia | I spotted that on a job already once this morning as well :( | 14:18 |
| vdrok | should we just add polling there? remove the test? wait for notifications to merge and use them? | 14:18 |
| nicodemos | TheJulia, I'll fix that. Thanks. | 14:18 |
| *** lets has joined #openstack-ironic | 14:18 | |
| TheJulia | vdrok: polling most likely, it is what we do in the tests for other async things. | 14:18 |
| TheJulia | vdrok: like state changes | 14:18 |
| TheJulia | That would keep it consistent at least | 14:19 |
| *** baoli has joined #openstack-ironic | 14:19 | |
| *** baoli has quit IRC | 14:19 | |
| *** ricardoas has left #openstack-ironic | 14:20 | |
| *** ricardoas has joined #openstack-ironic | 14:20 | |
| *** baoli has joined #openstack-ironic | 14:20 | |
| *** rloo has joined #openstack-ironic | 14:21 | |
| vdrok | ok, let's go that way for now :) | 14:21 |
| openstackgerrit | Hugo Nicodemos proposed openstack/ironic: Onetime boot when set_boot_device isn't persistent https://review.openstack.org/340596 | 14:21 |
| TheJulia | vdrok: and there are helper methods already for waiting until something returns the desired result :) | 14:23 |
| openstackgerrit | Kyrylo Romanenko proposed openstack/python-ironicclient: Add basic tests for OSC plugin baremetal port commands https://review.openstack.org/365692 | 14:24 |
| TheJulia | nicodemos: Thank you! I guess now we just need to wait until the oneview ci reports in | 14:25 |
| vdrok | TheJulia: not in api tests tho :) that's in scenario manager | 14:25 |
| *** dsneddon has quit IRC | 14:25 | |
| TheJulia | oh weird.... because I thought I spotted it in tempest test output earlier | 14:25 |
| TheJulia | oh, I get what your saying | 14:26 |
| *** pester has quit IRC | 14:26 | |
| TheJulia | I thought they were in the base tools and that was imported into the api tests | 14:26 |
| *** pester has joined #openstack-ironic | 14:27 | |
| vdrok | TheJulia: aha, ok, there is another waiters.py in common | 14:27 |
| vdrok | strange that we have basically the same thing in two places | 14:28 |
| TheJulia | fun! | 14:29 |
| nicodemos | TheJulia: \o/ yeah. Thanks. and vdrok too. --] | 14:30 |
| patchbot | Error: Spurious "]". You may want to quote your arguments with double quotes in order to prevent extra brackets from being evaluated as nested commands. | 14:30 |
| milan | btw my test issue was indeed my fault. Twice. Facepalm. | 14:30 |
| mgould | milan: d'oh! | 14:32 |
| *** vsaienko has joined #openstack-ironic | 14:33 | |
| * milan could use some cake now | 14:33 | |
| mgould | and I think they *all* thought it was real apart from the presenter :-) | 14:33 |
| *** daemontool has quit IRC | 14:33 | |
| *** amoralej|lunch is now known as amoralej | 14:36 | |
| *** spartacloud has joined #openstack-ironic | 14:38 | |
| clif_h | anyone know if there's a way in gerrit to unbold reviews that are updated, but you've looked at and have nothing to add? | 14:39 |
| *** srobert has joined #openstack-ironic | 14:39 | |
| openstackgerrit | Milan Kováčik proposed openstack/python-ironic-inspector-client: List introspection statuses support https://review.openstack.org/408116 | 14:41 |
| openstackgerrit | Andrey Shestakov proposed openstack/bifrost: Add support of remote logging https://review.openstack.org/410247 | 14:43 |
| vdrok | clif_h: don't think it's possible, unless you add a vote/comment | 14:43 |
| *** zzzeek has quit IRC | 14:43 | |
| *** zzzeek has joined #openstack-ironic | 14:43 | |
| *** mkrai_ has joined #openstack-ironic | 14:44 | |
| clif_h | vdrok: thanks, see related conversation just now in #openstack-infra if you're interested | 14:44 |
| openstackgerrit | Kyrylo Romanenko proposed openstack/python-ironicclient: Add basic tests for OSC plugin baremetal chassis commands https://review.openstack.org/366158 | 14:44 |
| clif_h | apparently gertty supports this feature and is console-based https://pypi.python.org/pypi/gertty | 14:46 |
| *** hoangcx has quit IRC | 14:46 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic: Make CONF.debug also reflect on IPA https://review.openstack.org/410168 | 14:53 |
| openstackgerrit | Kyrylo Romanenko proposed openstack/python-ironicclient: Describe possible exception in docstring https://review.openstack.org/410253 | 14:55 |
| *** glonlas_ has joined #openstack-ironic | 14:56 | |
| *** rbartal has quit IRC | 14:56 | |
| *** glonlas_ has quit IRC | 14:57 | |
| *** glonlas_ has joined #openstack-ironic | 14:57 | |
| *** glonlas has quit IRC | 14:59 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic: Support defining and loading hardware types https://review.openstack.org/336626 | 15:05 |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic: Use polling in set_console_mode tempest test https://review.openstack.org/410254 | 15:05 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic: Introduce generic hardware types https://review.openstack.org/400678 | 15:05 |
| *** mjura has quit IRC | 15:07 | |
| *** nicodemos has quit IRC | 15:08 | |
| *** nicodemos has joined #openstack-ironic | 15:09 | |
| *** nmathew- has joined #openstack-ironic | 15:10 | |
| openstackgerrit | Lucas Alvares Gomes proposed openstack/ironic-lib: Correct reraising of exception https://review.openstack.org/405734 | 15:11 |
| *** nmathew has quit IRC | 15:11 | |
| *** chlong has quit IRC | 15:13 | |
| *** vsaienko has quit IRC | 15:13 | |
| *** glonlas_ has quit IRC | 15:14 | |
| *** mtanino has joined #openstack-ironic | 15:15 | |
| *** glonlas has joined #openstack-ironic | 15:15 | |
| *** vsaienko has joined #openstack-ironic | 15:16 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic: [WIP] Load hardware types in the conductor https://review.openstack.org/409812 | 15:17 |
| *** rama_y has joined #openstack-ironic | 15:26 | |
| *** rama_y_ has joined #openstack-ironic | 15:26 | |
| *** yufei has joined #openstack-ironic | 15:27 | |
| NobodyCam | Good Morning Ironic'ers | 15:28 |
| *** vsaienko has quit IRC | 15:29 | |
| nicodemos | morning, NobodyCam | 15:29 |
| vdrok | morning NobodyCam | 15:29 |
| dtantsur | morning NobodyCam | 15:30 |
| NobodyCam | morning nicodemos vdrok dtantsur :) | 15:30 |
| rloo | hello and a Gooooood Morning NobodyCam, nicodemos, vdrok, dtantsur :) | 15:31 |
| vdrok | morning rloo :) | 15:31 |
| dtantsur | morning rloo | 15:31 |
| NobodyCam | Good Morning rloo :) | 15:31 |
| *** vsaienko has joined #openstack-ironic | 15:31 | |
| nicodemos | morning, rloo | 15:31 |
| milan | morning NobodyCam rloo! :) | 15:32 |
| NobodyCam | morning milan :) | 15:32 |
| rloo | hiya milan | 15:32 |
| openstackgerrit | Kyrylo Romanenko proposed openstack/python-ironicclient: Add basic tests for OSC plugin baremetal chassis commands https://review.openstack.org/366158 | 15:35 |
| lucasagomes | NobodyCam, morning | 15:39 |
| lucasagomes | rloo, morning too :-) | 15:39 |
| NobodyCam | Good Morning lucasagomes :) | 15:39 |
| rloo | hi lucasagomes! | 15:39 |
| NobodyCam | hey hey lucasagomes i had a question RE staging-drivers | 15:41 |
| lucasagomes | NobodyCam, shoot | 15:41 |
| *** links has quit IRC | 15:42 | |
| NobodyCam | several of the drivers have a python-requirements.txt file, one of them (amt driver) has other-requirements.txt is it ok if I rename other- to python- so we are standardized across all the drivers? | 15:42 |
| vdrok | NobodyCam: there is a difference -- https://github.com/openstack/ironic-staging-drivers/blob/master/devstack/plugin.sh#L38 | 15:43 |
| NobodyCam | ahh | 15:44 |
| nicodemos | Hey, TheJulia. If you dont mind: https://review.openstack.org/#/c/358041 , the CI its passing now. =D | 15:44 |
| patchbot | patch 358041 - ironic - Reusing oneview_client when possible | 15:44 |
| NobodyCam | okay, i will need to handle both then | 15:44 |
| NobodyCam | thank you lucasagomes and vdrok :) | 15:44 |
| lucasagomes | NobodyCam, yeah, so others is just a plan bash script to install the dependencies | 15:44 |
| lucasagomes | NobodyCam, when they are not on pip | 15:45 |
| vdrok | np :) | 15:45 |
| *** moshele has quit IRC | 15:45 | |
| rloo | dtantsur, mariojv, lucasagomes, vdrok if you understand objects -- do you know what the very first version of class NodeCRUDPayload should be? https://review.openstack.org/#/c/356541/28/ironic/objects/node.py | 15:47 |
| patchbot | patch 356541 - ironic - Add ironic resources CRUD notifications | 15:47 |
| * dtantsur does not understand them | 15:47 | |
| rloo | i don't understand them either, but it seems to me that the very first time we have an object, it is version 1.0. regardless of the version of its parent class/object. | 15:48 |
| vdrok | rloo: in a meeting, will take a look in 15 minutes :) | 15:48 |
| rloo | vdrok: thx. | 15:48 |
| * lucasagomes looks | 15:48 | |
| TheJulia | nicodemos: of course, the logs address does not resolve :( | 15:49 |
| mariojv | looking rloo | 15:49 |
| *** sborkows has quit IRC | 15:49 | |
| mariojv | rloo: i think if it was only NodePayload, then it might have to be 1.1, but since it's adding fields, 1.0 is fine | 15:50 |
| mariojv | rloo: i'd print out the full payload and double check first though | 15:50 |
| lucasagomes | rloo, apparently all payloads inheriting from nepayload is 1.1 | 15:51 |
| lucasagomes | because the parent (nodepayload) was updated once | 15:51 |
| rloo | lucasagomes: but don't you just update the ones inheriting cuz they existed as version 1.0 using version 1.0 of the parent class? | 15:51 |
| lucasagomes | rloo, true that | 15:51 |
| rloo | this is a new class. i don't see why it should be version 1.1 | 15:52 |
| rloo | eg what if NodePayload was version 1.14 and we added a new class that inherited. would it start off as version 1.14? | 15:52 |
| lucasagomes | rloo, yeah might worth asking yuriyz about it. But it sounds like it should be '1.0' because it's a new payload | 15:53 |
| nicodemos | TheJulia, thanks for the report. ricardoas ^ | 15:53 |
| rloo | i did ask yuriyz. he doesn't think it should be 1.0 but i haven't gotten a reason that makes sense to me | 15:53 |
| rloo | lucasagomes: i don't want to approve that w/o knowing. cuz it has ramifications for future objects, etc. and i guess we really should understand it. | 15:54 |
| yuriyz | hi we should bump version of inherited class if parent changed | 15:54 |
| mariojv | yuriyz: why? | 15:54 |
| rloo | yuriyz: what do you mean by 'bump'. the parent changed BEFORE this class was created. | 15:54 |
| yuriyz | but new payload should start from 1.0 | 15:54 |
| mariojv | oh | 15:54 |
| mariojv | yes, i agree | 15:54 |
| lucasagomes | yuriyz, I agree with that, but you should bump it when the payload that inherits it already exist, right ? | 15:54 |
| mariojv | yuriyz: so L592 here https://review.openstack.org/#/c/356541/28/ironic/objects/node.py should change to 1.0 | 15:55 |
| *** glonlas_ has joined #openstack-ironic | 15:55 | |
| patchbot | patch 356541 - ironic - Add ironic resources CRUD notifications | 15:55 |
| *** Goneri has quit IRC | 15:55 | |
| lucasagomes | in this case, we are adding a new one. So the initial version (1.0) is already based on NodePayload 1.1 | 15:55 |
| mariojv | 593, i mean | 15:55 |
| ricardoas | Hey, nicodemos! I'll check this out... it should have been resolving to 150.165.85.31 :( | 15:56 |
| mariojv | we should certainly update the docs to resolve this ambiguity, once there's consensus on what to do in that patch | 15:56 |
| ricardoas | TheJulia can you please check if this ip is accessible? | 15:56 |
| mariojv | i can do that | 15:56 |
| yuriyz | mariojv please look at NodeCorrectedPowerStatePayload and NodeSetPowerStatePayload they have 1.1 already | 15:57 |
| *** glonlas_ has quit IRC | 15:57 | |
| mariojv | yuriyz: ah, you're right | 15:58 |
| *** glonlas_ has joined #openstack-ironic | 15:58 | |
| mariojv | :| | 15:58 |
| *** glonlas has quit IRC | 15:58 | |
| mariojv | yuriyz: however, didn't those first merge with 1.0? | 15:58 |
| *** rajinir has joined #openstack-ironic | 15:59 | |
| rloo | yuriyz, mariojv: exactly. they existed BEFORE NodePayload got updated to v1.1 | 15:59 |
| mariojv | yuriyz: https://github.com/openstack/ironic/commit/ff32b51bbffc566d339dbb1f7f19b3b7429a91d1#diff-11867de23f1b786138f931d9736841ddR482 | 15:59 |
| mariojv | ctrl+f NodeSetPowerStatePayload | 15:59 |
| yuriyz | yes https://review.openstack.org/#/c/401311/12/ironic/objects/node.py | 15:59 |
| patchbot | patch 401311 - ironic - Move interface validation from API to conductor side (MERGED) | 15:59 |
| mariojv | yuriyz: so given that this is a new notification, and that was just a payload update, i think the payload for a new notification should be 1.0 | 16:00 |
| JayF | Does anyone know who Szymon Borkowski is on IRC? | 16:00 |
| yufei | ping | 16:00 |
| mariojv | yuriyz: unless there's a compelling reason to make it 1.0 at first | 16:00 |
| mariojv | *not make it 1.0 | 16:00 |
| yuriyz | mariojv ok agree | 16:00 |
| mariojv | thanks yuriyz | 16:00 |
| mariojv | i'll update docs later today to point this out for any new notifications that get added in the future | 16:01 |
| *** mkrai_ has quit IRC | 16:01 | |
| rloo | thx mariojv, yuriyz. | 16:01 |
| lucasagomes | yuriyz, so the order they got in the code matters right ? For the CRUDPayload, it's initial version (== 1.0) is based on NodePayload 1.1 | 16:01 |
| mariojv | np, thanks for bringing it up rloo | 16:01 |
| rloo | mariojv: i think it is an 'object, ovo' thing, the versioning. | 16:01 |
| mariojv | rloo: for the docs you mean? i agree, but it's ironic policy that determines how to use the versions in notifications | 16:02 |
| mariojv | unless there's a cross-project notification spec i don't know about | 16:02 |
| rloo | mariojv: ah, ok, that makes sense. | 16:02 |
| *** rcernin has quit IRC | 16:04 | |
| *** Syed__ has joined #openstack-ironic | 16:06 | |
| *** daemontool has joined #openstack-ironic | 16:07 | |
| *** Goneri has joined #openstack-ironic | 16:08 | |
| *** pcaruana has quit IRC | 16:09 | |
| openstackgerrit | Merged openstack/ironic-ui: Updated from global requirements https://review.openstack.org/409972 | 16:11 |
| openstackgerrit | Merged openstack/ironic-ui: Consolidate node last_error processing https://review.openstack.org/406184 | 16:11 |
| *** jaosorior has quit IRC | 16:12 | |
| *** jaosorior has joined #openstack-ironic | 16:13 | |
| *** bfournie has left #openstack-ironic | 16:16 | |
| *** Nisha_Agarwal has joined #openstack-ironic | 16:18 | |
| Nisha_Agarwal | hi | 16:19 |
| *** bfournie has joined #openstack-ironic | 16:19 | |
| Nisha_Agarwal | I am trying to login to gerrit code review, it opens the OpenID authentication page....anyone knows what we need to enter there? anyhelp is appreciated | 16:19 |
| mrtenio | JayF, I don't know if you found it. Did you try sborkows? | 16:21 |
| JayF | mrtenio: I don't see anyone by that name in here, but it's fine, I just emailed the list about the issue more generally | 16:21 |
| mrtenio | np | 16:22 |
| TheJulia | ricardoas: The IP is accessible | 16:23 |
| openstackgerrit | Yuriy Zveryanskyy proposed openstack/ironic: Add ironic resources CRUD notifications https://review.openstack.org/356541 | 16:23 |
| -openstackstatus- NOTICE: Launchpad SSO is not currently working, so logins to our services like review.openstack.org and wiki.openstack.org are failing; the admins at Canonical are looking into the issue but there is no estimated time for a fix yet. | 16:24 | |
| *** ChanServ changes topic to "Launchpad SSO is not currently working, so logins to our services like review.openstack.org and wiki.openstack.org are failing; the admins at Canonical are looking into the issue but there is no estimated time for a fix yet." | 16:24 | |
| openstackgerrit | Yuriy Zveryanskyy proposed openstack/ironic: Add node maintenance notifications https://review.openstack.org/396239 | 16:27 |
| *** Romanenko_K has quit IRC | 16:27 | |
| *** jaosorior has quit IRC | 16:28 | |
| *** e0ne has quit IRC | 16:28 | |
| openstackgerrit | Yuriy Zveryanskyy proposed openstack/ironic: Add node console notifications https://review.openstack.org/397812 | 16:30 |
| *** baoli has quit IRC | 16:30 | |
| rloo | Nisha_Agarwal: did you see the notice above ^^? | 16:30 |
| Nisha_Agarwal | rloo, thanks. so currently no solution? | 16:31 |
| lucasagomes | vdrok, are you going to fix the test_set_console_mode() from tempest to poll ? | 16:32 |
| rloo | Nisha_Agarwal: dunno, i only know what i read ^^ | 16:32 |
| *** glonlas_ has quit IRC | 16:32 | |
| *** aNuposic has joined #openstack-ironic | 16:33 | |
| Nisha_Agarwal | rloo, :( | 16:33 |
| vdrok | Lucasagomes yup it's on review | 16:33 |
| lucasagomes | vdrok, oh will take a look. Thanks! | 16:33 |
| vdrok | Out of office currently, lights gone off :( | 16:33 |
| Nisha_Agarwal | vdrok, lucasagomes are u guys able to access review.openstack.org? | 16:33 |
| vdrok | Will continue at home | 16:34 |
| lucasagomes | Nisha_Agarwal, it's working for me | 16:34 |
| *** glonlas has joined #openstack-ironic | 16:34 | |
| *** glonlas has quit IRC | 16:34 | |
| *** Goneri has quit IRC | 16:34 | |
| vdrok | Wfm Nisha_Agarwal | 16:34 |
| *** glonlas has joined #openstack-ironic | 16:34 | |
| Nisha_Agarwal | today i just rebooted my laptop, i think it upgraded something, and its not working after that :( | 16:34 |
| vdrok | Or you are not able only to log in? | 16:34 |
| vdrok | If so, notice is ^^ | 16:35 |
| Nisha_Agarwal | vdrok, i dont know what shall be OpenID for my account | 16:35 |
| JayF | Nisha_Agarwal: As the status indicates, it's an issue on the server. I'd wait and expect the status bot to update when it's fixed. | 16:35 |
| Nisha_Agarwal | JayF, hmmm | 16:36 |
| vsaienk0 | morning rloo, once you have a time please have look at: https://review.openstack.org/#/c/327046/29/ironic/drivers/modules/network/common.py@123 We didn't add VIF RPC objects so we sending VIF via RPC as json, it makes impossible to extend VIF object in future. I think that we should add VIF RPC object (inherited from oslo versioned objects) to have possibility to extend it. wdyt? | 16:37 |
| patchbot | patch 327046 - ironic - Add Virtual Network Interface Driver APIs | 16:37 |
| rloo | vsaienk0: looking... | 16:38 |
| *** sacharya has joined #openstack-ironic | 16:38 | |
| rloo | vsaienk0: yeah, i was going to look at the other patches to see what that vif 'object' was. | 16:38 |
| rloo | vsaienk0: so you're saying there is no ovo for vif? | 16:39 |
| rloo | vsaienk0: i think part of the problem might be that this vif 'object' is vague. | 16:39 |
| vsaienk0 | rloo: yes, there is no ovo for vif | 16:39 |
| rloo | vsaienk0: all it is now is an ID (some string). i think the intent is that it could be something more, depending on the usecase. | 16:40 |
| rloo | vsaienk0: and then the actual interface/driver, will know what to do with it. | 16:40 |
| rloo | vsaienk0: so we could create an ovo for it, and then up the version after it changes. to something like an id, and a dict :) | 16:41 |
| rloo | vsaienk0: or we leave it as a json -- and then have to update the rcp api etc if we change it to an object later. not sure which approach to take. what do you think? | 16:42 |
| *** moshele has joined #openstack-ironic | 16:42 | |
| vsaienk0 | rloo: fair enough, if we decide to extend it, we may convert to ovo, and then extend. | 16:43 |
| vsaienk0 | rloo: thanks! | 16:43 |
| rloo | vsaienk0: the *only* problem. is that vif_attach() is an interface API. | 16:43 |
| rloo | vsaienk0: if we change the type of vif (json to object), we can't break out-of-tree interfaces later. | 16:44 |
| vsaienk0 | rloo: we can pass both json and ovo, and then deprecate json | 16:44 |
| *** eroux has quit IRC | 16:45 | |
| *** vsaienko has quit IRC | 16:45 | |
| *** aNuposic has quit IRC | 16:45 | |
| *** chlong has joined #openstack-ironic | 16:46 | |
| *** aNuposic has joined #openstack-ironic | 16:46 | |
| rloo | vsaienk0: give me a few minutes, i want to look at all the patches together. | 16:47 |
| *** rbartal has joined #openstack-ironic | 16:48 | |
| *** rama_y_ has quit IRC | 16:51 | |
| *** rama_y has quit IRC | 16:51 | |
| lucasagomes | hi all, can someone please take a look at https://review.openstack.org/#/c/409755/ ? This prevents VBMC from hiding errors from IPMI (can lead to Ironic being fooled if a request fail and not retry) | 16:52 |
| patchbot | patch 409755 - virtualbmc - Return proper errors on BMC action failures | 16:52 |
| JayF | lucasagomes: no, nobody can :) | 16:53 |
| JayF | lucasagomes: mainly because gerrit is busted, lol | 16:53 |
| ricardoas | TheJulia thanks for checking... we'll take a look at our dns asap :) | 16:53 |
| lucasagomes | JayF, :-( | 16:54 |
| * TheJulia guesses this is the perfect time to prep dinner to cook for the afternoon, and then run to the store. | 16:55 | |
| openstackgerrit | Anup Navare proposed openstack/ironic: [WIP] Allow logical name along with UUID in port creation https://review.openstack.org/410319 | 16:55 |
| TheJulia | uhh... did we ever decide to make names unique? | 16:56 |
| TheJulia | hmm, looks like we did | 16:58 |
| rloo | vsaienk0: so there are a few things that bother me about this VIF stuff. the new interface is meant to be generic, to allow something other than neutron to be used. looking at the spec, we don't save the vif in the db as a first class citizen. we save it as internal info to some node or port. | 16:59 |
| dtantsur | TheJulia, tbh I never understood openstack's tradition of non-unique names | 16:59 |
| rloo | vsaienk0: the spec sez 'These API endpoints will take via a POST body, a JSON representation of a generic VIF object. Making it generic allows for non-neutron based implementations to use this API' | 17:00 |
| *** rcernin has joined #openstack-ironic | 17:00 | |
| *** sacharya_ has joined #openstack-ironic | 17:00 | |
| *** ChanServ changes topic to "Bare Metal Provisioning | Status: http://bit.ly/ironic-whiteboard | Docs: http://docs.openstack.org/developer/ironic/ | Bugs: https://bugs.launchpad.net/ironic" | 17:01 | |
| -openstackstatus- NOTICE: Canonical admins have resolved the issue with login.launchpad.net, so authentication should be restored now. | 17:01 | |
| TheJulia | dtantsur: names are for humans and are only for our mental connecting memories :) | 17:01 |
| rloo | Nisha_Agarwal: ^^ | 17:01 |
| Nisha_Agarwal | rloo, ok let me try | 17:01 |
| Nisha_Agarwal | i tried few min back it didnt work | 17:01 |
| NobodyCam | I just logged in | 17:01 |
| *** sacharya has quit IRC | 17:02 | |
| Nisha_Agarwal | rloo, yes it worked | 17:02 |
| Nisha_Agarwal | :) thanks | 17:02 |
| rloo | Nisha_Agarwal: now get to work :D | 17:02 |
| Nisha_Agarwal | rloo, :) yes | 17:03 |
| mgould | dtantsur: unique per-tenant or globally? | 17:04 |
| mgould | unique globally is going to be suboptimal in most cases for a cloud service | 17:04 |
| Nisha_Agarwal | rloo, no it didnt login...it just open the gerrit without logi | 17:04 |
| Nisha_Agarwal | :( | 17:04 |
| Nisha_Agarwal | i have to yet resolve it | 17:04 |
| lucasagomes | mgould, I think Ironic does it globally | 17:05 |
| lucasagomes | but yeah, per tenant would make more sense now (before ironic was an admin only thing) | 17:06 |
| * mgould nods | 17:06 | |
| openstackgerrit | Merged openstack/ironic-inspector: Remove upgrade from non-ironic setup https://review.openstack.org/408569 | 17:06 |
| mgould | but yeah, for most of openstack "unique per tenant" is the most that makes sense | 17:07 |
| TheJulia | we're unique globally, we would have do store tenant and such in the table to make a unique relationship... but are names even unique for nova? | 17:07 |
| *** strigazi is now known as strigazi_AFK | 17:07 | |
| *** glonlas_ has joined #openstack-ironic | 17:07 | |
| dtantsur | mgould, I guess this is why it was not done | 17:08 |
| lucasagomes | TheJulia, no, for nova they are not | 17:08 |
| lucasagomes | and neither is for neutron AFAICT | 17:08 |
| TheJulia | so... we really can't logically support name use for ports associations | 17:09 |
| * TheJulia thinks this discussion has been had at least twice before | 17:09 | |
| lucasagomes | you can have X instances called "foo" for the same tenant if you want, but it will prevent a "nova delete <name>" if there are two or more | 17:09 |
| lucasagomes | TheJulia, I think the idea with ports was to not store the name in the port object at all ? | 17:09 |
| lucasagomes | TheJulia, but instead just use it to find the right node UUID and store the UUID instead | 17:10 |
| * lucasagomes understood that from the last conversations | 17:10 | |
| rloo | lucasagomes, TheJulia: sorry, haven't followed the entire conversation. is this a question of using the node's name instead of UUID, related to port stuff? | 17:11 |
| *** glonlas has quit IRC | 17:11 | |
| rloo | lucasagomes, TheJulia: i think lucasagomes is right; we use node name to get the node uuid and the node id is associated with the port. | 17:11 |
| lucasagomes | rloo, yeah, for port creation | 17:11 |
| *** glonlas_ has quit IRC | 17:12 | |
| openstackgerrit | Chris Krelle proposed openstack/bifrost: Adding staging driver support https://review.openstack.org/406401 | 17:12 |
| lucasagomes | NobodyCam, w00t! | 17:12 |
| rloo | lucasagomes: should be fine. i think/thought the bug described it but it was awhile ago. | 17:12 |
| *** dsneddon has joined #openstack-ironic | 17:12 | |
| *** moshele has quit IRC | 17:13 | |
| lucasagomes | rloo, cool, I will take a look at that patch soon | 17:13 |
| lucasagomes | thanks for clarifying btw | 17:13 |
| rloo | lucasagomes: :) that's my understanding anyway but i could be wrong, it's been awhile. | 17:13 |
| * lucasagomes same | 17:13 | |
| TheJulia | Yeah, I seem to remember disagreement regarding port updates supporting names since we really shouldn't force the unique constraint on the names as we do, but *shrug* | 17:14 |
| NobodyCam | lucasagomes: hehehehe | 17:14 |
| *** rpioso has joined #openstack-ironic | 17:18 | |
| openstackgerrit | Peter Piela proposed openstack/ironic-ui: Extend support for the Ironic state machine https://review.openstack.org/410326 | 17:18 |
| *** jpich has quit IRC | 17:19 | |
| *** fragatina has joined #openstack-ironic | 17:22 | |
| *** fragatina has quit IRC | 17:23 | |
| *** fragatina has joined #openstack-ironic | 17:23 | |
| *** rcernin has quit IRC | 17:24 | |
| yufei | hello, ironicers, I’m trying to deploy a whole-disk image to my baremetal machine today, but after ironic-conductor sucessfuly write the os to the machine disk, it throws an error “Error: Could not determine a suitable URL for the plugin”，and say deploy failed, but in fact it succeed, and it seems like something about keystone is wrong in conductor.conf, but when I try to delploy partition image to this node, I don’t meet this problem. | 17:24 |
| yufei | Does someone has meet similar problem before? | 17:25 |
| *** fragatina has quit IRC | 17:26 | |
| *** fragatina has joined #openstack-ironic | 17:26 | |
| * mgould hasn't | 17:28 | |
| mgould | yufei: could you post the relevant section of the logs to paste.openstack.org? | 17:28 |
| *** baoli has joined #openstack-ironic | 17:29 | |
| *** vsaienko has joined #openstack-ironic | 17:32 | |
| *** ohamada has quit IRC | 17:33 | |
| *** athomas has quit IRC | 17:33 | |
| *** ralonsoh_ has joined #openstack-ironic | 17:34 | |
| *** ccamacho has quit IRC | 17:35 | |
| *** rcernin has joined #openstack-ironic | 17:35 | |
| *** rcernin has quit IRC | 17:35 | |
| *** fragatina has quit IRC | 17:35 | |
| *** ralonsoh has quit IRC | 17:37 | |
| milan | TheJulia, have got a minute for a boot-from-volume question? | 17:37 |
| TheJulia | milan: sure | 17:37 |
| milan | TheJulia, https://review.openstack.org/#/c/326620/3..4/releasenotes/notes/optional-root-disk-9b972f504b2e6262.yaml | 17:38 |
| patchbot | patch 326620 - ironic-inspector - Add an option to not fail when root device is not ... | 17:38 |
| milan | aarefiev asks about local_gb having to be set manually | 17:38 |
| milan | dtantsur wanted that in that time ^ | 17:38 |
| * dtantsur is not to blame, it's all your mac | 17:38 | |
| milan | TheJulia, isn't that information now provided by the connector in Ironic? | 17:39 |
| milan | dtantsur :D | 17:39 |
| milan | I mean the local_gb for the scheduler | 17:39 |
| * milan searches firefox tabs for the specs.... | 17:40 | |
| yufei | hi, mgould，I paste it here http://paste.openstack.org/show/592249/ | 17:40 |
| *** fragatina has joined #openstack-ironic | 17:41 | |
| *** fragatina has quit IRC | 17:42 | |
| TheJulia | milan: I would think local_gb should be set to "0" since no storage was found, but i don't know the intricacies of flavor matching in that case. The local_gb is not provided by the connector, nova has to schedule on to the node, provide the connector info to ironic then | 17:42 |
| *** fragatina has joined #openstack-ironic | 17:42 | |
| *** nmathew- has quit IRC | 17:43 | |
| TheJulia | yufei: are the glance and/or swift credentials/settings correct? | 17:43 |
| *** derekh has quit IRC | 17:43 | |
| *** vsaienko has quit IRC | 17:44 | |
| yufei | It should be ok, because I can see it download the images to /var/lib/ironic successfuly | 17:45 |
| milan | TheJulia, ah, OK thx, I'll update the patch that way, local_gb == 0, I'm afraid I have no other option than to test the flavour matching ;) | 17:45 |
| TheJulia | yufei: interesting, agent_base_vendor... hmmm | 17:47 |
| *** rama_y has joined #openstack-ironic | 17:47 | |
| *** fragatina has quit IRC | 17:48 | |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic: Use polling in set_console_mode tempest test https://review.openstack.org/410254 | 17:48 |
| vdrok | lucasagomes: thanks for review, done | 17:49 |
| lucasagomes | vdrok, o/ | 17:49 |
| lucasagomes | +2 | 17:49 |
| lucasagomes | and folks, I'm calling it a day | 17:50 |
| lucasagomes | have a great evening all! | 17:50 |
| vdrok | TheJulia: rloo dtantsur ^^ easy patch to fix the race in the api test | 17:50 |
| vdrok | lucasagomes: good night :) | 17:50 |
| NobodyCam | night lucasagomes | 17:50 |
| *** lucasagomes is now known as lucas-afk | 17:50 | |
| milan | good night lucasagomes! :) | 17:50 |
| lucas-afk | o/ | 17:50 |
| TheJulia | yufei: so there are some lines before that we are likely missing that might provide a clue, but needless to say setting the boot loader seemed to fail, we just don't know why | 17:50 |
| * milan relocates | 17:51 | |
| TheJulia | vdrok: thank you | 17:51 |
| TheJulia | goodnight lucas-afk | 17:51 |
| *** moshele has joined #openstack-ironic | 17:52 | |
| * mgould -> home; good night! | 17:53 | |
| *** mgould is now known as mgould|afk | 17:53 | |
| yufei | hi, TheJulia, I’m using newton, I’m not sure whether this exits in devstack master, I will test it tomorrow | 17:53 |
| nicodemos | mgould|afk, lucas-afk night | 17:53 |
| openstackgerrit | Vladyslav Drok proposed openstack/ironic: Enhance wait_for_bm_node_status waiter https://review.openstack.org/410343 | 17:53 |
| openstackgerrit | Andre Aranha proposed openstack/python-oneviewclient: Added validation for local_link_connection https://review.openstack.org/377103 | 17:54 |
| *** moshele has quit IRC | 17:54 | |
| openstackgerrit | Andrey Shestakov proposed openstack/bifrost: Add support of remote logging https://review.openstack.org/410247 | 17:55 |
| yufei | hi, TheJulia, please help take a look at this patch when you have time. https://review.openstack.org/#/c/397517/9/, hope to get your suggestions. | 17:56 |
| patchbot | patch 397517 - ironic - Update multitenancy docs | 17:56 |
| *** milan has quit IRC | 17:57 | |
| rloo | vdrok: thx. TheJulia, does it need another review or are you good with +A'ing https://review.openstack.org/#/c/410254/? | 17:59 |
| patchbot | patch 410254 - ironic - Use polling in set_console_mode tempest test | 17:59 |
| *** yufei has left #openstack-ironic | 18:00 | |
| vdrok | rloo: re vif objects, do you think it's necessary to create a db table if we switch to ovo? | 18:01 |
| TheJulia | rloo: I'm fine with +A'ing | 18:02 |
| rloo | vdrok: nope, not necessary. my only concern is that people may assume if we have ovo, that it is also being saved in db | 18:02 |
| JayF | I just landed it, in that case | 18:02 |
| vdrok | rloo: well, it's not the case for notifications already :) | 18:02 |
| JayF | I was reviewing it already so when julia said to +A I added that with my +2 :D | 18:02 |
| rloo | vdrok: either way, json/dict, or ovo, isn't perfect in this case | 18:02 |
| rloo | vdrok: oh yeah, you're right. | 18:02 |
| rloo | vdrok: i'm fine if we go with ovo. was worried it'd be more work. | 18:03 |
| TheJulia | yuriyz: Sorry I didn't review it yesterday. I was pondering the wording. I put some suggestions in for clarity, sorry to drag this along :( | 18:03 |
| rloo | vdrok: and also, not sure what it buys us if this vif object is generic. | 18:03 |
| *** ijw has joined #openstack-ironic | 18:04 | |
| rloo | vdrok: all we know is that it has an ID. the rest will have to be in a dict for it to be generic. | 18:04 |
| *** ralonsoh_ has quit IRC | 18:04 | |
| rloo | vdrok: but having said that, fine if we have an ovo for it. | 18:04 |
| openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic: Move heartbeat processing to separate mixin class https://review.openstack.org/404364 | 18:04 |
| rloo | JayF, TheJulia: thx! | 18:05 |
| JayF | I guess I can pull that "itnermittant console failures" from the whiteboard now | 18:06 |
| TheJulia | :) | 18:06 |
| rloo | JayF: *after* it lands? :) | 18:06 |
| * TheJulia runs to the store, bbiab | 18:07 | |
| *** daemontool has quit IRC | 18:08 | |
| vdrok | rloo: will think on that with vasyl. otoh, to have proper ovo, we have to have a separate entity vif, with functions on it available, backportable if needed etc. | 18:09 |
| *** krtaylor has quit IRC | 18:09 | |
| dtantsur | going now, see you tomorrow | 18:10 |
| rloo | bye dtantsur | 18:10 |
| * dtantsur will present a lightning talk about ironic deployed by tripleo tomorrow in the office :) | 18:10 | |
| JayF | dtantsur: planning on recording it or otherwise for upstream folks to consume? I guess that also assumes it's in english :P | 18:10 |
| *** dtantsur is now known as dtantsur|afk | 18:10 | |
| dtantsur|afk | JayF, I suspect the recording will be for internal consumption only :( | 18:11 |
| rloo | dtantsur: cool! and ^^ although i was going to ask if there will be a wiki or doc about it | 18:11 |
| *** spartacloud is now known as zackf | 18:11 | |
| dtantsur|afk | well, if you want serious docs, we have them http://tripleo.org/advanced_deployment/baremetal_overcloud.html | 18:11 |
| dtantsur|afk | this is mostly like "look folks what we've got!" in 5-10 minutes kind of talk | 18:11 |
| * rloo sorry she asked :) | 18:11 | |
| dtantsur|afk | haha | 18:12 |
| openstackgerrit | Anup Navare proposed openstack/ironic: Allow logical name along with UUID in port creation https://review.openstack.org/405586 | 18:13 |
| rloo | vdrok: the other thing about the vifs stuff. you can attach a vif to a node, that vif can have id and other info. if you do a vif-list though, you only get the IDs, the other stuff isn't saved. if we do ovo, it would be odd not to save the rest of the vif somewhere, and then to return it. | 18:14 |
| *** lindycoder has joined #openstack-ironic | 18:16 | |
| vdrok | rloo: yeah, 'other stuff' def needs to be stored somewhere | 18:17 |
| *** pester has quit IRC | 18:20 | |
| rloo | vdrok: i added another comment, about vif-list should be a list of whatever thing that vif was in vif_attach. | 18:22 |
| rloo | vdrok: thx for fixing it :) | 18:23 |
| *** zackf has quit IRC | 18:23 | |
| vdrok | rloo: vifs patch? that was vsaienk0 :) | 18:24 |
| rloo | vdrok: i know, but he isn't around and you are :D | 18:24 |
| vdrok | hah | 18:24 |
| vdrok | I'll forward it to him tomorrow :) | 18:25 |
| *** zackf has joined #openstack-ironic | 18:28 | |
| *** fragatina has joined #openstack-ironic | 18:29 | |
| vdrok | xavierr: around? | 18:36 |
| xavierr | vdrok: yeap | 18:38 |
| vdrok | xavierr: I'm looking at https://review.openstack.org/397846 currently, and can not get one thing. | 18:39 |
| patchbot | patch 397846 - ironic - Remove agent vendor passthru from OneView drivers | 18:39 |
| vdrok | xavierr: the two functions that are moved to oneview deploy mixin were not present previously in the oneview deploy, there were in agent base vendor | 18:40 |
| vdrok | xavierr: err, oneview vendor interface | 18:40 |
| vdrok | xavierr: so when the /heartbeat as a separate interface was implemented, deploy.heartbeat was called | 18:41 |
| vdrok | xavierr: which in turn was calling self.reboot_to_instance where self was still a deploy interface. | 18:41 |
| vdrok | xavierr: since those methods were not in deploy interface, the ones from the base agent/agent_base_vendor mixins were used | 18:42 |
| vdrok | xavierr: so /heartbeat requests were not working before? | 18:42 |
| vdrok | xavierr: only node vendor passthru heartbeats were OK? | 18:43 |
| *** glonlas has joined #openstack-ironic | 18:44 | |
| mrtenio | JayF, could you take a look here https://review.openstack.org/#/c/377073/? Thank you | 18:45 |
| patchbot | patch 377073 - ironic - Adds another validation step when using dynamic al... | 18:45 |
| JayF | mrtenio: fwiw I have a +1 on that :) | 18:45 |
| JayF | mrtenio: but I'm also wondering; that patch conflicts with one that's significantly larger | 18:45 |
| JayF | mrtenio: that also has a single +2, I'm thinking it might be better to land that one first? | 18:46 |
| JayF | mrtenio: https://review.openstack.org/#/c/358041/ | 18:46 |
| patchbot | patch 358041 - ironic - Reusing oneview_client when possible | 18:46 |
| xavierr | vdrok: they were not working | 18:48 |
| xavierr | vdrok: they are working for this patch because methods were moved to there and are being called | 18:49 |
| vdrok | xavierr: aha, so in fact it fixes another bug | 18:49 |
| vdrok | apart from the removal of passthru | 18:50 |
| vdrok | xavierr: do you mind creating it and following up with a reno then? | 18:50 |
| xavierr | vdrok: what we are doing now is drop the need for this OneViewAgentDeployMixin in favor of this: https://review.openstack.org/#/c/408298/ | 18:50 |
| patchbot | patch 408298 - ironic - Shutdown server before change boot order for agent... | 18:50 |
| xavierr | if this new patch get merged we will not need this OneViewAgentDeployMixin anymore | 18:51 |
| vdrok | xavierr: I mean, filing a bug that /heartbeat does not work, and adding a reno that it is fixed after https://review.openstack.org/397846 ? | 18:51 |
| patchbot | patch 397846 - ironic - Remove agent vendor passthru from OneView drivers | 18:51 |
| vdrok | xavierr: yup, gotcha | 18:51 |
| vdrok | xavierr: still would be good to track that for some period, that endpoint did not work | 18:52 |
| xavierr | would be better adding a reno or working in 408298 directly? | 18:53 |
| *** dprince has quit IRC | 18:54 | |
| vdrok | xavierr: I'd say, just amend the 397846 reno, add another -fixes point. and then 408298 is just a refactor to drop unnecessary code | 18:54 |
| xavierr | vdrok: our vendor interface was necessary because of this bug, look: https://bugs.launchpad.net/ironic/+bug/1503855 | 18:54 |
| openstack | Launchpad bug 1503855 in Ironic "Set boot device while server is on" [High,In progress] - Assigned to Xavier (marcusrafael) | 18:54 |
| vdrok | xavierr: sure, I get that. the question was about - /heartbeat was not working as those two methods were called only when vendor passthru was used. the vendor interface removal fixes /heartbeat. would be good to mention it | 18:55 |
| vdrok | xavierr: it can be done as a followup to 397846 reno | 18:56 |
| xavierr | vdrok: indeed. so I will add a reno in <fixes> session to 397846, ok? | 18:56 |
| vdrok | xavierr: yup. in a followup please :) to make it merge faster :P | 18:57 |
| vdrok | ok, merging 397846 then, thanks for clarifying xavierr | 18:57 |
| xavierr | vdrok: thank you! | 18:58 |
| xavierr | vdrok: press the red button | 18:58 |
| *** Sukhdev has joined #openstack-ironic | 18:59 | |
| vdrok | done :) | 18:59 |
| *** fragatina has quit IRC | 19:01 | |
| *** krtaylor has joined #openstack-ironic | 19:01 | |
| *** fragatina has joined #openstack-ironic | 19:01 | |
| *** fragatina has quit IRC | 19:02 | |
| mrtenio | JayF, I see it, and also it would be nice your opinion on that, if you could take a look on that patch. We from OneView driver think it would be better if the patch https://review.openstack.org/#/c/358041/ gets in first. :) | 19:02 |
| patchbot | patch 358041 - ironic - Reusing oneview_client when possible | 19:02 |
| *** fragatina has joined #openstack-ironic | 19:02 | |
| JayF | mrtenio: that's what I suspected. MAybe make the other patch follow on behind that one then? sicne you're going to need a rebase anyway? | 19:03 |
| JayF | mrtenio: just landed that one, if you wanna rebase the other I can review it again | 19:05 |
| mrtenio | Doing it now | 19:08 |
| *** dprince has joined #openstack-ironic | 19:08 | |
| nicodemos | wow, thanks. JayF. :) | 19:09 |
| *** cloudnull has left #openstack-ironic | 19:10 | |
| *** rcernin has joined #openstack-ironic | 19:14 | |
| openstackgerrit | Merged openstack/ironic-lib: Correct reraising of exception https://review.openstack.org/405734 | 19:14 |
| *** madhu_ak has joined #openstack-ironic | 19:16 | |
| vdrok | rloo: added a question on the crud notifications patch. I might be missing some context and maybe it's too late ... | 19:19 |
| rloo | vdrok: how can it be too late? | 19:19 |
| rloo | vdrok: looking.... | 19:19 |
| rloo | vdrok: that's a discussion for the spec. | 19:19 |
| JayF | rloo: because I landed it :x | 19:20 |
| vdrok | yeah I know :( | 19:20 |
| rloo | vdrok, JayF; even though we land specs, we are also good at recognizing when we are wrong/can do better :D | 19:20 |
| JayF | vdrok: yeah, that question was handled in the spec, I was one of the advocates for that | 19:20 |
| rloo | vdrok: i can't recall now, if initially it was just 'done'. might have been me or JayF ^^ that wanted start, end ... | 19:20 |
| vdrok | JayF: ok, lemme look through the history | 19:21 |
| rloo | vdrok: i know that i wanted something similar to what nova was doing | 19:21 |
| JayF | vdrok: mainly because a deployer/operator, who doesn't have in depth knowledge of the system, doesn't have the context to realize for those few sync notifications that they didn't "miss" part of the notifications | 19:21 |
| rloo | vdrok: you don't know what/why someone wants notifications. | 19:21 |
| JayF | basically, the idea it's sync vs async is something mostly only a developer cares about | 19:21 |
| JayF | and it creates a notification chain that's harder to follow without that contextual knowledge | 19:21 |
| vdrok | JayF: but we still have the success-only notifications in chassis case, and in port.create | 19:22 |
| JayF | mariojv: ^ you might find this chat interesting | 19:22 |
| JayF | vdrok: well that's inconsistent and silly, we should've done the start-error-end style pattern for all of them :( | 19:22 |
| JayF | vdrok: you're exactly right, and I missed that in review | 19:23 |
| openstackgerrit | Stenio Araujo proposed openstack/ironic: Adds another validation step when using dynamic allocation https://review.openstack.org/377073 | 19:23 |
| openstackgerrit | Stenio Araujo proposed openstack/ironic: Reusing oneview_client when possible https://review.openstack.org/358041 | 19:23 |
| *** milan has joined #openstack-ironic | 19:23 | |
| rloo | JayF: i saw that 'success' thing but i think by that time, i was happy that i got <something else> agreed to and didn't want to pick on more stuff. | 19:24 |
| JayF | okay, so vdrok is right | 19:24 |
| JayF | the implementation doesn't match the spec | 19:24 |
| JayF | the spec shows a single, node.create.success for creating a node | 19:24 |
| mrtenio | JayF, 'did it. :) | 19:24 |
| JayF | the code implements a start-error-end | 19:24 |
| mariojv | hm | 19:24 |
| mariojv | i think .success makes sense for non-async operations | 19:24 |
| rloo | JayF: you must have an old version, we changed node create cuz the creating moved to the conductor | 19:24 |
| rloo | JayF: creation of node is a start/end/error, whereas chassis/port creation is a success i think. | 19:25 |
| JayF | rloo: aha, you're right. I was reading the spec via the review.openstack link, and there was also an update patch landed after that | 19:26 |
| vdrok | rloo: JayF if the whatever system is used filters out the representation of resources that has updated_at value earlier than what it has in cache, it should be ignoring that. If it is newer then that question still holds: if we do update start, then update error, we'll send the exact same representation of node we did already have in cache. and in case of | 19:26 |
| vdrok | crud, I can't think of a case when this system is not a caching representations system | 19:26 |
| rloo | mariojv: i think it is consistent to have start/end w/i the API level for port/chassis, but i don't care that much | 19:26 |
| mariojv | correctness is more important than consistency to me | 19:27 |
| mariojv | if it's async or long running, should be start/end, otherwise if it's quick and sync, .success | 19:27 |
| rloo | vdrok: so i honestly have no idea how notifications are going to be used. seems like we should design it for general cases, whatever that might be. | 19:27 |
| mariojv | i don't remember that part of code exactly though | 19:27 |
| JayF | mrtenio: now to wait for oneview ci to pass :) | 19:27 |
| mariojv | i think they're mainly going to be used for searchlight. i was going to use them for an automated resolution system (downstream feature), but now we're going to have specific fault support to cover that | 19:28 |
| rloo | vdrok: honestly, except for the notifications that have already landed, it isn't too late to change anything. | 19:28 |
| mariojv | it isn't too late for those, either, would just need version bumps | 19:28 |
| mariojv | imo | 19:28 |
| vdrok | rloo: the ones that landed are async, that should be OK | 19:28 |
| rloo | mariojv: we know that they are being used for searchlight, but i would like to believe we're doing this for other consumers (in the future) besides searchlight. | 19:28 |
| mariojv | ++ | 19:28 |
| JayF | FWIW, I didn't land the cRUD notifications, only voted +2. I think I intended to land it and just didn't. | 19:29 |
| JayF | rloo: ++ one case I would have for it if I were still operating onmetal, is using the notification ordering to track a node | 19:29 |
| rloo | JayF: doesn't matter who landed it or who reviewed it. we're a community. | 19:29 |
| rloo | JayF: i'm going to blame it on vdrok for not reviewing it in the first place :D | 19:29 |
| vdrok | I'm trying to find are there definitions for start/end/success notification statuses anywhere outside of ironic | 19:29 |
| JayF | rloo: OK. I just still get a little nervous about landing stuff of that size :) | 19:29 |
| *** fragatina has quit IRC | 19:30 | |
| mariojv | vdrok: start/end yes, i doubt it for success, we came up with that term i believe | 19:30 |
| rloo | JayF: i'm nervous about a lot of stuff. but gotta land them at some point and then iterate/fix later. | 19:30 |
| vdrok | mariojv: aha, ok | 19:30 |
| *** fragatina has joined #openstack-ironic | 19:30 | |
| rloo | JayF: some stuff isn't noticed and/or harder to argue about until the developer has coded. | 19:30 |
| rloo | vdrok: nova's: http://docs.openstack.org/developer/nova/notifications.html | 19:32 |
| vdrok | yeah I'm reading through that currently | 19:33 |
| rloo | i think mariojv is right, we came up with 'success'. | 19:33 |
| vdrok | it does not define the meanings tho | 19:33 |
| *** ijw has quit IRC | 19:33 | |
| vdrok | as they appear to be self-explanatory :D | 19:33 |
| vdrok | JayF: rloo: mariojv on that nova page, it seems that instance.update is kind of close to our case | 19:35 |
| rloo | vdrok: honestly, i think notification event types should be a xproject thing so there is consistency etc | 19:35 |
| vdrok | of synchronous update | 19:35 |
| *** alexpilo_ has joined #openstack-ironic | 19:36 | |
| vdrok | will look into code | 19:36 |
| vdrok | ugh, nova is so slow even for clone | 19:37 |
| *** alexpilotti has quit IRC | 19:39 | |
| *** ijw has joined #openstack-ironic | 19:39 | |
| *** Sukhdev has quit IRC | 19:41 | |
| *** Nisha_Agarwal has quit IRC | 19:44 | |
| vdrok | rloo: JayF mariojv yeah, so in nova, compute.instance.update is a notification sent from objects.instance.save | 19:44 |
| vdrok | no start/end | 19:45 |
| JayF | mariojv: ^ | 19:45 |
| JayF | vdrok: I'm realy curious what mariojv has to say about that design; I know before he worked on upstream openstack he did a *lot* of work parsing nova notifications for Rackspace | 19:45 |
| *** e0ne has joined #openstack-ironic | 19:45 | |
| JayF | and I think a lot of those lessons learned were reflected in the spec | 19:45 |
| vdrok | I guess it does not hurt much if we append success to that, even if it's different from nova tho | 19:46 |
| vdrok | JayF: sure | 19:46 |
| JayF | I would rather not roll back all that design work in the spec on a whim, without fully getting into the details again | 19:46 |
| vdrok | well, I can not really do that myself :D | 19:46 |
| rloo | vdrok: i think yuriyz may have initially had the CRUD stuff emit notifications in the .save() but for whatever reasons (which i must have commented on), i wanted him to move them to the api level. | 19:46 |
| vdrok | rloo: yeah, IIRC there were some special casing magic there | 19:47 |
| rloo | JayF: well, i wouldn't call it a whim. vdrok needs to be able to support his case and convince folks. | 19:47 |
| mariojv | nova's notifications weren't as structured as ours | 19:47 |
| vdrok | rloo: but looking at the nova's instance.save, I doubt we'll be able to get any close to that :) | 19:48 |
| mariojv | note that it's compute.instance.update, not compute.instance.update.<phase> | 19:48 |
| JayF | rloo: sure; I get that. We just already went through a few rounds of that in the spec :P | 19:48 |
| vdrok | I'm not opposed to doing it on the api side tho | 19:48 |
| rloo | JayF: yeah, I know :) | 19:48 |
| rloo | vdrok: after all that ^^, what are you proposing? | 19:48 |
| vdrok | mariojv: yes, exactly | 19:48 |
| mariojv | update is describing the actionbeing taken. it's not an indicator that the action is sync | 19:48 |
| vdrok | for now, I have in mind - replacing all the sync actions notifications to have only .success event | 19:49 |
| rloo | vdrok: but what if it failed? | 19:49 |
| mariojv | then .error is sent | 19:49 |
| vdrok | that includes all {node,port,chassis,portgroup}.{create,update,delete} | 19:50 |
| *** e0ne has quit IRC | 19:50 | |
| JayF | I didn't think all of those were sync anymore though? | 19:50 |
| JayF | Hasn't node create been moved to the conductor? | 19:50 |
| JayF | Or are you talking about, things done over RPC but return sync from API client perspective | 19:50 |
| rloo | JayF: define 'sync'. Some of those do not go through the conductor, some do. | 19:50 |
| JayF | yeah, that's what I'm trying to nail down, what's mean by "sync" in this case | 19:51 |
| mariojv | https://specs.openstack.org/openstack/ironic-specs/specs/not-implemented/notifications.html#proposed-change | 19:51 |
| mariojv | success means "immediate" | 19:51 |
| patchbot | The operation succeeded. means immediate | 19:51 |
| mariojv | ??? ^ | 19:51 |
| vdrok | rloo: JayF mariojv: if the node is not created, why send any notification? user will get 500 error code or something. same for failed update/delete, they will get error code from the api. and the representation won't change | 19:51 |
| mariojv | start/end are not immediate | 19:51 |
| mariojv | so not necessarily sync vs. async | 19:51 |
| rloo | vdrok: because we don't know what the notification system is being used for. it is not the user. | 19:51 |
| mariojv | vdrok: this is for a service that may be external to the client to check | 19:51 |
| mariojv | searchlight won't know, for example, if you fail to create a node | 19:52 |
| JayF | vdrok: use case: I have a metric based on how many failures. If I get too many .error notifications, I don't meet my sla of "running a stable ironic cloud" or something. | 19:52 |
| rloo | vdrok: what if the notification system is used for eg, checking that the ironic service isn't flaky, that 'starts' have 'ends' or 'errors' and dont' just disappear. | 19:52 |
| JayF | vdrok: this is almost identical to how Rackspace uses nova notifications today, fwiw | 19:52 |
| *** wajdi has joined #openstack-ironic | 19:52 | |
| mariojv | yes | 19:52 |
| vdrok | mariojv: well, if you do that, then you know it failed. if someone else tells you that he's updated the node when he's got 500, he's lying :) | 19:53 |
| mariojv | and it's an extremely large pain | 19:53 |
| mariojv | because some errors do have notifications, and some don't | 19:53 |
| mariojv | *i* know | 19:53 |
| JayF | notifications, logs, and immediate api response codes | 19:53 |
| mariojv | but my external collection service does not | 19:53 |
| JayF | all have different use cases and potentially different consumers | 19:53 |
| vdrok | JayF: OK, that justifies error notification (but not its body sending the whole node) | 19:53 |
| *** ijw has quit IRC | 19:54 | |
| vdrok | but start notification is kind of useless to me | 19:54 |
| mariojv | well, if i want to troubleshoot why my node create calls are failing | 19:54 |
| mariojv | it's useful to see what was in the node | 19:54 |
| JayF | and it's not a bad thing; it's a good thing to duplicate information between them | 19:54 |
| JayF | logs are more single-instance troubleshooting targeted, notifications can be used to troubleshoot items on a larger scale (i.e. find patterns, troubleshoot a node over a long period of time), immediate response codes are for the user right then | 19:54 |
| mariojv | maybe all nodes that have a particular field error out | 19:54 |
| mariojv | also, the consumer can just discard info it doesn't need | 19:54 |
| JayF | vdrok: ^ I think troubleshooting a node over time, and making patterns when the errors occur, justify storing the node | 19:55 |
| mariojv | that's what searchlight does, only tracks changes | 19:55 |
| rloo | vdrok: i can see that .start may be useless to you. but for mariojv and JayF, it is useful. | 19:55 |
| vdrok | mariojv: but you already have the representation of the node sent in previous successful notifications? | 19:55 |
| JayF | we explicitly /want/ to be able to go back in time and see the exact state of the node when it errored vs when it didn't error | 19:55 |
| rloo | vdrok: given that .start is useful to others, are you opposed to having .start. you can just ignore it. | 19:55 |
| mariojv | vdrok: also, if you just want error notifications, there's a config option for that | 19:55 |
| JayF | when troubleshooting these things in the real world, the one biggest lesson I learned is never underestimate how useful more informatin is :) | 19:56 |
| JayF | which is why I was especially pleased we put the node in all relevant notifications | 19:56 |
| rloo | mariojv, vdrok: would it be useful to have .start at a diff config level from .end? | 19:56 |
| JayF | rloo: it would have to be orthoganal to the level | 19:56 |
| mariojv | i think it makes things too complicated | 19:56 |
| JayF | rloo: i.e. level configured, aand like what phases to notify on | 19:57 |
| JayF | but I agree with mariojv about it adding complexity | 19:57 |
| JayF | especially when a consumer could easily be configured to ignore the notifications it doesn't want | 19:57 |
| mariojv | it's easy for operators to understand priority levels, like log levels | 19:57 |
| rloo | JayF, mariojv: adds complexity if you want to change the default behavior. am trying to figure out something that works for you three :) | 19:57 |
| mariojv | configuring at the other granularity would probably require familiarity with the lifecycle of an operation in ironic | 19:57 |
| vdrok | rloo: JayF mariojv OK, i get the error case now. in case of start - 3 possibilities : create (start of create only useful if you want to see the body caused the failure), update( start not needed as you have the representation of the node from previous notifications), delete (same as update) | 19:57 |
| mariojv | i don't understand what the issue is though | 19:57 |
| rloo | mariojv: the issue is vdrok doesn't care about .start notifications. | 19:58 |
| vdrok | in case of create failure, I'd still stick with the ironic logs, as delivery of notifications is not guaranteed | 19:58 |
| mariojv | for debugging sure | 19:59 |
| mariojv | not for alerting, unless you have real time log analytics | 19:59 |
| JayF | I mean, even for debugging: at a certain larger scale you're going to be looking at aggreggate statistics | 20:00 |
| JayF | and logs are /awful/ for tracking the life cycle of a single node | 20:00 |
| JayF | and notifications can be good for that | 20:00 |
| mariojv | yup | 20:00 |
| vdrok | mariojv: JayF rloo you'll still get sucess/error right? without start | 20:01 |
| vdrok | would it be enough for statistics? | 20:01 |
| *** ijw has joined #openstack-ironic | 20:01 | |
| JayF | Going back to the real-world usage of notifications from nova that rackspace does today | 20:01 |
| JayF | we explicitly, for some actions, time the difference between the start/end notifications | 20:02 |
| JayF | in order to determine system performance | 20:02 |
| mariojv | no | 20:02 |
| JayF | mariojv: no? | 20:02 |
| mariojv | no to vdrok's question | 20:02 |
| JayF | ah, okay | 20:02 |
| JayF | yeah b/c we use start/end for instance-boot-timings in nova, right? | 20:02 |
| mariojv | maybe i'm getting DDoSed, and conductor is held up before everything finishes | 20:02 |
| mariojv | i see 10K .start events, alert an engineer | 20:02 |
| mariojv | yes, and timing | 20:02 |
| mariojv | that can be solved separately with the @METRICS decorators thoguh | 20:03 |
| vdrok | hmmm | 20:03 |
| vdrok | OK then in theory you need start/end/error everywhere, no matter sync operation or not. just in case of any operation going outside of ironic boundaries (like write to db, sending over rpc)? | 20:04 |
| openstackgerrit | Milan Kováčik proposed openstack/python-ironic-inspector-client: List introspection statuses support https://review.openstack.org/408116 | 20:05 |
| JayF | vdrok: that's exactly why I hrm'd at the idea of only doing success/error for stuff that hits rpc | 20:05 |
| mariojv | hmm, maybe | 20:05 |
| mariojv | i guess it depends on the definition of "immediate" | 20:05 |
| vdrok | the only thing that would be unnecessary to have start/end/error would be eg modifying in-memory object then :) | 20:06 |
| mariojv | and sometimes we might notify on that | 20:06 |
| mariojv | for example, conductor startup | 20:06 |
| JayF | Well, I think the DB operations can be pretty much presumed to be immediate, but I can see an argument for that | 20:06 |
| mariojv | i disagree, long-running queries happen | 20:06 |
| vdrok | I see your logic now. but that will end up very extensive on the messaging bus, like 3 notifications for every node action at least, very close to each other in time | 20:07 |
| mariojv | vdrok: that's what priority level config is for | 20:07 |
| mariojv | vdrok: you can ignore all notifications with priority level < ERROR if you want | 20:07 |
| vdrok | mariojv: JayF also for the aggregate statistics thing, we don't really need the actual object representation right? | 20:08 |
| mariojv | vdrok: also, nova emits *tons* more notifications than we ever will and is fine on some beefy rabbitmq nodes | 20:08 |
| mariojv | vdrok: yes, you do | 20:08 |
| JayF | vdrok: remember that multiple use cases exist; the object representation is useful for troubleshooting case, even if not useful in all cases | 20:08 |
| mariojv | vdrok: "show me all attempts at creating nodes with agent_ipmitool driver in the past 24 hours" | 20:08 |
| vdrok | right. that's hard :( | 20:09 |
| mariojv | vdrok: "show me all fields changed on all nodes with .error events for the past hour" | 20:09 |
| vdrok | to predict everything | 20:09 |
| openstackgerrit | Merged openstack/ironic: Remove agent vendor passthru from OneView drivers https://review.openstack.org/397846 | 20:10 |
| vdrok | mariojv: JayF rloo OK, after that discussed, and believing you :D that a whole lot more nova notifications can be handled without much trouble, I'm good with that, thanks for clarifying! | 20:11 |
| mariojv | thank you vdrok! always good to check my assumptions | 20:11 |
| JayF | These discussions are super fun | 20:11 |
| rloo | thx vdrok, mariojv, JayF :) | 20:11 |
| JayF | honestly | 20:11 |
| JayF | :D | 20:11 |
| vdrok | that still leaves the problem of the patch having success things. yuriyz won't be happy about the outcome :D | 20:11 |
| rloo | JayF: wish we had time to document/update spec with some of this info but anyway | 20:12 |
| rloo | vdrok: you mean, you don't like 'success'? | 20:12 |
| JayF | vdrok: Can you summarize what we think the diff is from what already exists in the spec? | 20:12 |
| mariojv | i don't think there is a difference, lol | 20:12 |
| mariojv | maybe more clarity on the word "immediate" | 20:12 |
| * rloo doesn't like success but didn't think it was worth arguing over it | 20:12 | |
| vdrok | mariojv: hm, if i got you correctly, even port/chassis create would have to emit start/end/error right? | 20:13 |
| mariojv | vdrok: i don't really care that much | 20:13 |
| vdrok | JayF: rloo will do that first thing in the morning tomorrow | 20:13 |
| mariojv | vdrok: i'd have to read the code, but imo (the way i was envisioning in the spec) | 20:13 |
| mariojv | vdrok: was send .start when api call is received, .end when done | 20:14 |
| mariojv | .error if something bad happens in between that ironic can't handle | 20:14 |
| mariojv | dunno about client error | 20:14 |
| JayF | mariojv: the spec explicitly ssays success/error style for chassis/port create | 20:14 |
| mariojv | i'm fine with the way yuriyz was doing it for now though | 20:14 |
| mariojv | ohhh | 20:14 |
| JayF | mariojv: so if we're doing start/error/end, that's a change from the spec and the preexisting code | 20:15 |
| mariojv | well, like i said, i don't really care that much | 20:15 |
| mariojv | it can always be changed with version bumps later | 20:15 |
| rloo | i'd be very happy if we got rid of success and replaced with start/end/error | 20:15 |
| mariojv | i cared more about start/end with nodes | 20:15 |
| vdrok | if we are saying about trying to catch all cases, then it seems to me 'start/end/error' everywhere and no success | 20:16 |
| rloo | i think it'd make it easier for notification system, not to have to deal with 'success'. | 20:16 |
| mariojv | i don't think it would make things easier | 20:16 |
| mariojv | it's just parsing a string | 20:16 |
| rloo | mariojv: but two diff cases instead of just one. | 20:16 |
| mariojv | if you want one case, just write a catch-all | 20:17 |
| rloo | mariojv: and if we ever end up moving the chassis/port creates to the conductor, the notifications will change. | 20:17 |
| mariojv | only look at .end/.success | 20:17 |
| JayF | ^ that to me is the best argument for making port/chassis creates start/error/end | 20:17 |
| JayF | what rloo just said | 20:17 |
| *** amoralej is now known as amoralej|off | 20:17 | |
| mariojv | so, i'm +0.5 to making all the API calls start/end | 20:17 |
| mariojv | i'm -1 to removing .success notifications altogether | 20:17 |
| rloo | mariojv: do we already have some .success? | 20:18 |
| vdrok | mariojv: I just don't know how to define it then | 20:18 |
| mariojv | i was just checking that | 20:18 |
| mariojv | thought maybe some of the power notifications might be | 20:18 |
| vdrok | success is not immediate | 20:18 |
| patchbot | The operation succeeded. is not immediate | 20:18 |
| vdrok | wow :D | 20:18 |
| mariojv | be quiet patchbot | 20:18 |
| JayF | success I am still a stupid patchbot. | 20:18 |
| patchbot | The operation succeeded. I am still a stupid patchbot. | 20:18 |
| rloo | mariojv: yup baremetal.node.power_state_corrected.success | 20:18 |
| rloo | baremetal.node.provision_set.success | 20:19 |
| vdrok | these are async right? | 20:19 |
| vdrok | crud are all sync | 20:19 |
| mariojv | no | 20:19 |
| rloo | ahh, those are special.the ones with .success are not initiated by any API call. | 20:20 |
| mariojv | https://github.com/openstack/ironic/blob/master/ironic/conductor/manager.py#L973 | 20:20 |
| rloo | baremetal.node.provision_set.success is emitted when ironic-conductor successfully changes provision state instantly, without any intermediate work required (example is AVAILABLE to MANAGEABLE). It has notification level INFO. | 20:20 |
| mariojv | yup rloo | 20:20 |
| rloo | baremetal.node.power_state_corrected.success is emitted by ironic-conductor when the power state on the baremetal hardware is different from the previous known power state of the node and the database is corrected to reflect this new power state. It has notification level “info”. | 20:20 |
| rloo | mariojv: i love documentation :) | 20:20 |
| mariojv | \o/ | 20:20 |
| rloo | mariojv: and i think we went back/forth on those ^^ but it didn't make sense to ahve start/end on them. | 20:21 |
| mariojv | exactly | 20:21 |
| mariojv | you either correct it or you don't | 20:21 |
| vdrok | mariojv: mariojv rloo but that is still db write | 20:21 |
| rloo | mariojv: so yeah, i'm good with them being .success. | 20:21 |
| mariojv | yeah | 20:21 |
| mariojv | implicitly | 20:22 |
| vdrok | that's why I don't know how to define success properly | 20:22 |
| mariojv | also, one thing | 20:22 |
| vdrok | only db write can be success, but db write + rpc can not> | 20:22 |
| mariojv | well, never mind | 20:22 |
| vdrok | ? | 20:22 |
| openstackgerrit | Merged openstack/ironic: Use polling in set_console_mode tempest test https://review.openstack.org/410254 | 20:23 |
| mariojv | i guess the start/end/success distinction isn't very specific | 20:23 |
| mariojv | you're right, db writes are considered "immediate" from that perspective | 20:24 |
| vdrok | rloo: JayF are you ok with that? ^^ | 20:24 |
| mariojv | vdrok: i have an idea | 20:24 |
| mariojv | vdrok: what if we have a DB error notification? | 20:24 |
| mariojv | that's just a ".error" | 20:24 |
| mariojv | no .success for db writes because that'd be wild | 20:25 |
| mariojv | vdrok: presumably, ironic will give up after a while if a db call hangs? | 20:25 |
| rloo | vdrok: i honestly didn't/don't think of it in terms of db writes. almost any change is going to involve a db write. any change worth notifying i think. | 20:25 |
| mariojv | yeah, RPC vs. db is a distinction worth making i think | 20:26 |
| mariojv | message queues are lossier | 20:26 |
| vdrok | mariojv: success emitted in the api and and error in dbapi? | 20:26 |
| mariojv | vdrok: just .error in dbapi | 20:26 |
| mariojv | so i have baremetal.node.power_state_corrected.success | 20:26 |
| mariojv | if that task.process_event call right above the line i linked fails | 20:27 |
| mariojv | it's because of a db error | 20:27 |
| vdrok | mariojv: in case of update, that'd be hard to distinguish api-based update from ironic itself updating | 20:27 |
| mariojv | send a notification, something like baremetal.db.write.error | 20:27 |
| mariojv | the error is unrelated to whatever you're notifying on, it's a DB problem | 20:27 |
| vdrok | hm | 20:27 |
| JayF | mariojv: I like that, as an operator, but I think it's kinda orthoganal to the conversation, right? | 20:28 |
| mariojv | JayF: i think it's good because it keeps our assumption that db writes are immediate in place | 20:28 |
| vdrok | mariojv: OK, I'll need to discuss it with yuriyz, because it may require significant changes from the current impl | 20:28 |
| mariojv | allowing us to distinguish things like rpc calls from things that should be immediate | 20:28 |
| mariojv | sounds good | 20:28 |
| mariojv | vdrok: i wasn't saying that we should do the db error notification *now* | 20:29 |
| mariojv | just that it would solve the distinction between .success and .start/.end | 20:29 |
| vdrok | otherwise the patch is good to go as is :) | 20:29 |
| mariojv | so, what needs to change about the patch? | 20:29 |
| mariojv | oh, just make it .success? to match spec? | 20:30 |
| JayF | port and chassis create need to go from success/error -> start/error/end | 20:30 |
| vdrok | but I'm not ready to push +A now, will do it tomorrow, along with spec update. As it will be mostly clarifying the use cases/our intents on why we have done it this way | 20:30 |
| JayF | right? | 20:30 |
| vdrok | JayF: oh, right, errors | 20:30 |
| vdrok | JayF: so I think that ^^ discussions means, success/error where it is success currently | 20:31 |
| vdrok | and start/end/error everywhere else | 20:31 |
| mariojv | fine by me | 20:32 |
| mariojv | i prefer start/end everywhere for api stuff | 20:32 |
| JayF | Someone should probably link to this IRC discussion | 20:32 |
| JayF | with a -1 on the existing patch | 20:32 |
| mariojv | but i can propose that as a follow up at some point | 20:32 |
| JayF | so that yuriy know's what happened to his patch :x | 20:32 |
| vdrok | mariojv: so end or error instead of success or error?\ | 20:32 |
| mariojv | vdrok: don't worry about it, let's go with what you said | 20:33 |
| vdrok | OK phew :) | 20:33 |
| mariojv | the sppec update for clarifyign | 20:33 |
| mariojv | *clarifying | 20:33 |
| *** mrtenio is now known as mrtenio-afk | 20:33 | |
| vdrok | I'll add a link and a comment | 20:33 |
| mariojv | my follow up == 6 months from now :) | 20:34 |
| *** ijw has quit IRC | 20:34 | |
| rloo | mariojv: i think if you just update dev docs, would be sufficient? | 20:35 |
| JayF | https://etherpad.openstack.org/p/iFoG829Xig is some info ttx pulled about participation | 20:35 |
| JayF | it was mentioned at the TC meeting | 20:35 |
| JayF | Ironic has less changes proposed but more merged | 20:36 |
| rloo | JayF: participation in what? | 20:36 |
| JayF | rloo: changes proposed / changes merged | 20:36 |
| JayF | rloo: i.e. vs the first few weeks of newton | 20:36 |
| mariojv | that seems good | 20:36 |
| *** ijw has joined #openstack-ironic | 20:36 | |
| JayF | rloo: ironic merged more commits but had less proposed, which I viewed as a good sign -- we're all very well aligned on the stuff that matters, and are staying focused on the priorities | 20:36 |
| rloo | JayF: didn't realize i was working on a 'protocore' project. glad to know that our core team 'operates correctly'... | 20:37 |
| rloo | JayF: i'm not sure what to think of these numbers. it should be the numbers. quality. makes me a bit sad. anyway, maybe it is useful... | 20:37 |
| JayF | rloo: *shrug* I ignored the headers because those are more or less a function of time | 20:37 |
| rloo | JayF: s/should/should not/ | 20:37 |
| JayF | It's just an interesting thing to read, and I think a credit to our ability to stay on target and land code | 20:38 |
| mariojv | this seems similar to measuring lines of code as impact though, lol | 20:38 |
| JayF | in a time when other projects are seeing a drop | 20:38 |
| mariojv | except for the contributor metric perhaps | 20:38 |
| JayF | I tend to try to find the things to be happy about in numbers, and just use them as a tool for motivation :) | 20:38 |
| rloo | JayF: dunno about that. i wouldn't pat ourselves in our back, or whatever the expression is. | 20:38 |
| rloo | JayF: good attitude! | 20:38 |
| *** ijw has quit IRC | 20:38 | |
| JayF | I'm pessimistic by default so I'm always fighting the negativity | 20:39 |
| JayF | lol | 20:39 |
| rloo | JayF: really, thought the opposite of you and *I* am pessimistic :D | 20:39 |
| *** ijw has joined #openstack-ironic | 20:40 | |
| TheJulia | rloo: I've known JayF for a very long time. He is absolutely a pessimist by default :) | 20:40 |
| rloo | TheJulia: thx for confirming :) | 20:41 |
| JayF | That's part of why I try to be ... expressive in a positive way | 20:41 |
| JayF | I'm not trying to cheer others up, I'm trying to keep myself cheery | 20:42 |
| openstackgerrit | Julia Kreger proposed openstack/ironic: Add storage_interface to base driver class https://review.openstack.org/348006 | 20:42 |
| vdrok | JayF: interesting stuff. nova people are who should be pessimistic, not us :) | 20:44 |
| * vdrok leaves now | 20:44 | |
| vdrok | good night! | 20:44 |
| rloo | night vdrok, thx for the discussion! | 20:45 |
| mariojv | good night | 20:45 |
| *** moshele has joined #openstack-ironic | 20:57 | |
| *** ijw has quit IRC | 20:59 | |
| TheJulia | dtantsur|afk: I finally got some brain cycles to begin work again on 355625, and I replied to your comments regarding auth parameters. I'll try to get up early my time tomorrow to discuss if necessary. | 20:59 |
| *** [1]rpioso has joined #openstack-ironic | 21:01 | |
| *** ijw has joined #openstack-ironic | 21:02 | |
| *** causten has joined #openstack-ironic | 21:03 | |
| *** rpioso has quit IRC | 21:04 | |
| *** jkilpatr_ has quit IRC | 21:05 | |
| *** moshele has quit IRC | 21:10 | |
| *** srobert has quit IRC | 21:14 | |
| *** moshele has joined #openstack-ironic | 21:16 | |
| *** baoli_ has joined #openstack-ironic | 21:19 | |
| *** baoli__ has joined #openstack-ironic | 21:19 | |
| *** baoli__ has quit IRC | 21:21 | |
| *** jkilpatr_ has joined #openstack-ironic | 21:21 | |
| *** baoli__ has joined #openstack-ironic | 21:21 | |
| *** baoli has quit IRC | 21:22 | |
| *** baoli_ has quit IRC | 21:24 | |
| *** glonlas_ has joined #openstack-ironic | 21:25 | |
| *** rajinir has quit IRC | 21:26 | |
| *** glonlas has quit IRC | 21:28 | |
| *** madhu_ak has quit IRC | 21:28 | |
| openstackgerrit | Pavlo Shchelokovskyy proposed openstack/ironic-python-agent: Probe for TC mirror during tinyipa build https://review.openstack.org/410404 | 21:29 |
| *** glonlas_ has quit IRC | 21:30 | |
| *** chlong has quit IRC | 21:30 | |
| *** dsneddon has quit IRC | 21:35 | |
| *** baoli has joined #openstack-ironic | 21:36 | |
| *** amoralej|off is now known as amoralej | 21:36 | |
| *** baoli__ has quit IRC | 21:37 | |
| *** ijw has quit IRC | 21:37 | |
| *** baoli_ has joined #openstack-ironic | 21:37 | |
| *** baoli_ has quit IRC | 21:38 | |
| *** baoli_ has joined #openstack-ironic | 21:38 | |
| *** dprince has quit IRC | 21:39 | |
| *** dprince has joined #openstack-ironic | 21:39 | |
| *** baoli has quit IRC | 21:41 | |
| *** dprince has quit IRC | 21:50 | |
| JayF | pas-ha is going to win most ingenious IPA patch award as soon as https://review.openstack.org/#/c/410404 passes the gate so I can vote on it | 21:56 |
| patchbot | patch 410404 - ironic-python-agent - Probe for TC mirror during tinyipa build | 21:56 |
| JayF | lol. | 21:56 |
| *** jcoufal has quit IRC | 21:57 | |
| pas-ha | Hehe | 21:58 |
| pas-ha | I'll post another patch set tomorrow, one thing missing | 21:58 |
| *** moshele has quit IRC | 21:59 | |
| *** trown is now known as trown|outtypewww | 21:59 | |
| *** led_ has joined #openstack-ironic | 21:59 | |
| *** krtaylor has quit IRC | 22:00 | |
| *** amoralej is now known as amoralej|off | 22:01 | |
| *** priteau has quit IRC | 22:04 | |
| *** rajinir has joined #openstack-ironic | 22:08 | |
| pas-ha | thou shall find mirrors :) http://logs.openstack.org/04/410404/1/check/gate-tempest-dsvm-ironic-ipa-partition-pxe_ipmitool-tinyipa-src-ubuntu-xenial/7bb02d8/logs/devstacklog.txt.gz#_2016-12-13_21_48_18_171 | 22:08 |
| * pas-ha signing off, have a good day all | 22:10 | |
| *** milan has quit IRC | 22:10 | |
| *** sacharya_ has quit IRC | 22:10 | |
| *** sacharya has joined #openstack-ironic | 22:11 | |
| *** [1]rpioso has quit IRC | 22:12 | |
| *** bfournie has quit IRC | 22:19 | |
| *** jkilpatr_ has quit IRC | 22:25 | |
| *** zackf has quit IRC | 22:28 | |
| *** zackf has joined #openstack-ironic | 22:29 | |
| *** jrcloud has joined #openstack-ironic | 22:31 | |
| *** spartacloud has joined #openstack-ironic | 22:32 | |
| *** ijw has joined #openstack-ironic | 22:33 | |
| *** zackf has quit IRC | 22:34 | |
| *** jrcloud has quit IRC | 22:35 | |
| *** zackf has joined #openstack-ironic | 22:45 | |
| *** jheroux has quit IRC | 22:45 | |
| *** spartacloud has quit IRC | 22:47 | |
| *** wajdi_ has joined #openstack-ironic | 22:51 | |
| *** wajdi_ has quit IRC | 22:52 | |
| *** wajdi_ has joined #openstack-ironic | 22:53 | |
| *** wajdi has quit IRC | 22:55 | |
| *** bfournie has joined #openstack-ironic | 23:02 | |
| *** wajdi_ has quit IRC | 23:04 | |
| *** jkilpatr_ has joined #openstack-ironic | 23:05 | |
| openstackgerrit | Ramamani Yeleswarapu proposed openstack/ironic-inspector: Update documentation to deploy Ironic Inspector with DevStack https://review.openstack.org/410456 | 23:24 |
| openstackgerrit | Hugo Nicodemos proposed openstack/ironic: Reusing oneview_client when possible https://review.openstack.org/358041 | 23:26 |
| *** swatson has joined #openstack-ironic | 23:29 | |
| *** swatson has left #openstack-ironic | 23:29 | |
| *** swatson has joined #openstack-ironic | 23:29 | |
| swatson | Anyone in the channel I could ask a quick question regarding your DB migration tests? | 23:30 |
| swatson | "your" being the ones in ironic | 23:30 |
| JayF | You're welcome to ask :) This is one of the quieter times w/r/t timezones and such | 23:32 |
| JayF | and while I'm here, I doubt I know what you're asking | 23:32 |
| JayF | lol | 23:32 |
| swatson | JayF: Haha well thanks :) And I kinda figured the time wasn't great anyway, but worth a shot | 23:33 |
| swatson | Anyway, I was reading test_migration.py in the DB unit tests and it mentions openstack_citest, but I couldn't find any mention of it in code | 23:33 |
| swatson | Is that something set up in a config file that gets read by oslo_db stuff? | 23:33 |
| swatson | I should probably add I'm working on adding migration testing for magnum, and was pointed to the ironic tests | 23:35 |
| JayF | oh; so I think what that's talking about | 23:35 |
| JayF | is to run those unit tests you have to have a DB setup with access for it to use | 23:35 |
| swatson | Right, but I couldn't find anywhere where that string is actually used, at least in code | 23:36 |
| JayF | yeah; I'm on a hunt for it now too | 23:36 |
| swatson | :) | 23:36 |
| swatson | It's definitely getting used, since I can run the migration tests but 14 get skipped (the other 4 all pass) | 23:38 |
| swatson | I'm 99% certain it's because I haven't set up that DB yet. | 23:38 |
| JayF | I'm checking codesearch.openstack.org | 23:38 |
| JayF | I'd expect that to get setup in CI or something | 23:38 |
| JayF | but it's not in ironic's devstack anywhere | 23:38 |
| JayF | http://git.openstack.org/cgit/openstack-infra/storyboard/tree/doc/source/contributing.rst references setting it up for storyboard ci too | 23:39 |
| JayF | so it's a common openstack pattern | 23:39 |
| JayF | swatson: aha! It appears to be pre-setup before the job even runs | 23:40 |
| JayF | swatson: via project-config | 23:40 |
| JayF | http://git.openstack.org/cgit/openstack-infra/project-config/tree/jenkins/jobs/macros.yaml search that file for openstack_citest | 23:40 |
| JayF | swatson: I'm curious, it was a bit of an XY question, what are you really trying to do? | 23:41 |
| swatson | JayF: End goal is to set up migration testing for magnum, as we currently have none | 23:41 |
| JayF | aha | 23:41 |
| JayF | well it looks like project-config does the magic db setting up for you | 23:41 |
| swatson | Thanks for the find :D | 23:41 |
| JayF | yeah | 23:41 |
| JayF | and it looks like the code is directly in our repo, for test_migrations | 23:42 |
| JayF | to se if it can connect, if not it just happily skips it | 23:42 |
| swatson | Yeah, I was turning my wheels trying to figure out how the test base got the info | 23:42 |
| swatson | This at least explains how Jenkins starts the DB but it looks to me like there's something between Jenkins and tox that tells the test the DB name | 23:43 |
| JayF | swatson: https://github.com/openstack/ironic/blob/master/ironic/tests/unit/db/sqlalchemy/test_migrations.py#L232 | 23:44 |
| swatson | D'oh | 23:45 |
| JayF | swatson: I'm going to garner a guess that FIXTURES is configured via env variables | 23:45 |
| JayF | swatson: after tracking the code back | 23:46 |
| swatson | JayF: I saw that call and must have just glossed over it after not seeing my string | 23:46 |
| JayF | I have the advantage of knowing I miss things, so I always use ^f to find the thing | 23:46 |
| JayF | lol | 23:46 |
| swatson | Smart :D | 23:47 |
| swatson | I' | 23:47 |
| swatson | I'll poke around with this a bit more and see what I dig up. Thanks for the help! | 23:48 |
| JayF | no problem. Good luck! | 23:48 |
| *** sacharya has quit IRC | 23:57 | |
| *** baoli_ has quit IRC | 23:59 | |
| *** baoli has joined #openstack-ironic | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!