| janders | TheJulia thank you for your comments on https://review.opendev.org/#/c/744117/, I will work on fixing it up | 00:22 |
|---|---|---|
| patchbot | patch 744117 - ironic - Generic way to configure clean step priorites - 11 patch sets | 00:22 |
| *** Goneri has quit IRC | 00:22 | |
| janders | I need to have a small problem with my phone fixed, back in an hour | 00:28 |
| janders | and I should probably order some Yubikeys... | 00:28 |
| *** chihhsin has joined #openstack-ironic | 00:40 | |
| openstackgerrit | gugug proposed openstack/tenks master: Use importlib to take place of imp module https://review.opendev.org/748818 | 01:10 |
| *** gyee has quit IRC | 01:25 | |
| *** ricolin_ has joined #openstack-ironic | 01:31 | |
| openstackgerrit | Armstrong Liu proposed openstack/ironic master: clean up mac address with the pxe configuration files https://review.opendev.org/750545 | 01:49 |
| *** rloo has quit IRC | 01:50 | |
| *** Qianbiao has joined #openstack-ironic | 02:00 | |
| Qianbiao | Morning Ironic. | 02:05 |
| *** zaneb has quit IRC | 02:09 | |
| *** zaneb has joined #openstack-ironic | 02:10 | |
| *** rcernin has quit IRC | 02:29 | |
| *** mkrai has joined #openstack-ironic | 03:02 | |
| *** stevebaker has quit IRC | 03:14 | |
| *** zzzeek has quit IRC | 03:18 | |
| *** zzzeek has joined #openstack-ironic | 03:19 | |
| *** rcernin has joined #openstack-ironic | 03:29 | |
| *** rcernin has quit IRC | 03:29 | |
| *** rcernin has joined #openstack-ironic | 03:29 | |
| *** Qianbiao has quit IRC | 03:32 | |
| *** Qianbiao has joined #openstack-ironic | 03:32 | |
| *** rh-jelabarre has quit IRC | 03:51 | |
| *** zzzeek has quit IRC | 03:55 | |
| *** zzzeek has joined #openstack-ironic | 03:58 | |
| openstackgerrit | Merged openstack/bifrost master: Remove transform_boot_image https://review.opendev.org/750324 | 04:18 |
| *** belmoreira has joined #openstack-ironic | 04:23 | |
| *** Qianbiao has quit IRC | 04:39 | |
| *** zzzeek has quit IRC | 04:45 | |
| *** zzzeek has joined #openstack-ironic | 04:48 | |
| *** tzumainn has quit IRC | 05:09 | |
| *** zzzeek has quit IRC | 05:17 | |
| *** zzzeek has joined #openstack-ironic | 05:18 | |
| openstackgerrit | Merged openstack/ironic master: Change [agent]image_download_source=http https://review.opendev.org/750203 | 05:26 |
| *** Qianbiao has joined #openstack-ironic | 05:57 | |
| *** abdysn has joined #openstack-ironic | 06:47 | |
| openstackgerrit | Yushiro FURUKAWA proposed openstack/ironic master: Fix handling OctetString for pysnmp https://review.opendev.org/750401 | 06:52 |
| arne_wiebalck | Good morning, ironic! | 06:58 |
| *** jtomasek has joined #openstack-ironic | 07:01 | |
| janders | good morning arne_wiebalck o/ | 07:02 |
| arne_wiebalck | Hey janders o/ | 07:03 |
| *** mkrai has quit IRC | 07:04 | |
| iurygregory | good morning janders arne_wiebalck and Ironic! | 07:24 |
| openstackgerrit | Jacob Anders proposed openstack/ironic master: Generic way to configure clean step priorites https://review.opendev.org/744117 | 07:30 |
| janders | hey iurygregory o/ | 07:30 |
| janders | how are things? | 07:30 |
| *** dougsz has joined #openstack-ironic | 07:31 | |
| iurygregory | doing good and you? =) | 07:32 |
| janders | That's good to hear! | 07:32 |
| janders | It's been raining here all day so I haven't managed to do any exercise - I will pop out for a half an hour run, back soon. | 07:33 |
| iurygregory | enjoy! | 07:33 |
| janders | thank you! :) | 07:33 |
| janders | maybe doing a half-run-half-swim will help me understand Redfish better :) will be handy for getting back to the Redfish/MAC work later today | 07:34 |
| Qianbiao | Good morning, Ironic. | 07:39 |
| iurygregory | good morning Qianbiao | 07:40 |
| Qianbiao | Hey iurygregory | 07:40 |
| Qianbiao | Hey janders and arne_wiebalck o/ | 07:41 |
| Qianbiao | hi iurygregory, which one should i use, dib or ipab to build deploy ramdisk | 07:41 |
| iurygregory | Qianbiao, both should work =) | 07:42 |
| Qianbiao | ok, should the ipa version exactly matched with ironic version? | 07:43 |
| iurygregory | better if it does =) | 07:43 |
| Qianbiao | sure, and it seems ipab version does not map to ironic. latest the best? | 07:44 |
| iurygregory | you want to know the versions of ipa for a given release? | 07:45 |
| Qianbiao | ipab | 07:45 |
| Qianbiao | ironic-python-agent-builder version for like ironic-train | 07:45 |
| iurygregory | oh, ipa-builder you can use master | 07:45 |
| Qianbiao | ok. | 07:45 |
| iurygregory | you only need to worry if you want something compatible with py2 I would say | 07:45 |
| iurygregory | you can use the latest | 07:45 |
| Qianbiao | yes, i need to use python2 for now. | 07:46 |
| Qianbiao | because devstack is running on python2 in my ENV. | 07:46 |
| iurygregory | what?! | 07:46 |
| iurygregory | =O | 07:46 |
| Qianbiao | =( | 07:46 |
| iurygregory | what versions you trying to use? (just wondering) | 07:46 |
| Qianbiao | I am running a devstack@train | 07:47 |
| Qianbiao | on Python 2.7.17 | 07:47 |
| Qianbiao | Now, i am try to find a ipa-builder + ipa version for this env. | 07:48 |
| iurygregory | yeah train has support for python2.7 | 07:48 |
| iurygregory | so don't go with ipa-builder | 07:48 |
| iurygregory | I would say it's better to do with the diskimage-builder | 07:48 |
| Qianbiao | ok, will reading doc now. | 07:48 |
| iurygregory | and you need a version that works for py27 | 07:48 |
| Qianbiao | which one? ipa? | 07:48 |
| Qianbiao | ipa@stable/train will work? | 07:49 |
| iurygregory | diskimage-builder need to be version 2.38.0 (this one still has support for py2) | 07:50 |
| iurygregory | yes ipa stable/train will work | 07:50 |
| Qianbiao | ok thanks iurygregory | 08:00 |
| *** lucasagomes has joined #openstack-ironic | 08:10 | |
| iurygregory | np =) | 08:13 |
| *** rcernin has quit IRC | 08:20 | |
| janders | back.. the rain politely stopped for the majority of my run | 08:28 |
| Qianbiao | iurygregory i still need to install ipa-builder when using diskimage-builder, right? | 08:29 |
| Qianbiao | i am reading this doc: https://docs.openstack.org/ironic-python-agent-builder/latest/admin/dib.html | 08:29 |
| *** Lucas_Gray has joined #openstack-ironic | 08:30 | |
| janders | iurygregory I have a quick question if you have time | 08:31 |
| iurygregory | Qianbiao, depends if you need elements that are only on diskimage-builder | 08:32 |
| janders | I'd like to address Julia's comment in https://review.opendev.org/#/c/744117/11 and add some usage examples for clean_steps_overrides. Where in the repo is the best place to put these? | 08:32 |
| patchbot | patch 744117 - ironic - Generic way to configure clean step priorites - 12 patch sets | 08:32 |
| iurygregory | janders, sure | 08:32 |
| *** derekh has joined #openstack-ironic | 08:32 | |
| * iurygregory looks | 08:32 | |
| openstackgerrit | Xinliang Liu proposed openstack/ironic master: Add uefi pxe iscsi boot https://review.opendev.org/746628 | 08:32 |
| janders | https://opendev.org/openstack/ironic/src/branch/master/doc/source/admin/cleaning.rst would be my guess but I thought I would ask first | 08:33 |
| iurygregory | janders, https://docs.openstack.org/ironic/latest/admin/cleaning.html | 08:33 |
| iurygregory | yeah | 08:33 |
| iurygregory | that would be my guess also =) | 08:33 |
| janders | if we both thought that it is the way to go :) | 08:33 |
| janders | is it best to add these to the same commit? | 08:33 |
| iurygregory | that would make a lot of sense I would say | 08:33 |
| janders | ok! will do | 08:34 |
| janders | thank you iurygregory! | 08:34 |
| iurygregory | it would probably help reviewers =) | 08:34 |
| janders | first I will have to re-create my config for this, as it got lost while reshuffling servers to make space for dev-scripts install :) | 08:34 |
| janders | but that shouldn't be hard | 08:34 |
| iurygregory | ouch =( | 08:35 |
| iurygregory | sorry to hear that | 08:35 |
| janders | it's funny :) I kept everything except /etc | 08:37 |
| Qianbiao | <iurygregory> thanks, i find the ironic-agent element on dib | 08:37 |
| janders | my bifrost uninstall script was backing up ironic.conf (how clever!) | 08:37 |
| janders | *but* | 08:38 |
| janders | also to /etc (how dumb!) :) | 08:38 |
| iurygregory | ops =P | 08:38 |
| janders | FAIL! | 08:38 |
| *** ianychoi__ has quit IRC | 08:40 | |
| openstackgerrit | Jacob Anders proposed openstack/ironic master: Generic way to configure clean step priorites https://review.opendev.org/744117 | 08:42 |
| janders | iurygregory regarding docs, there is an existing section in the admin / cleaning guide: "How do I change the priority of a cleaning step?" which is using the "old" method. Do you think it's okay if I move the existing way to "option 1" and add my implementation as "option 2"? | 08:45 |
| janders | It's a new feature so the devops in me is thinking we shouldn't remove the old one till the new one gets more testing in user environments. What do you think? | 08:45 |
| * iurygregory reads the section | 08:46 | |
| iurygregory | yeah it makes totally sense to keep and add the details on how you can use the generic way | 08:47 |
| janders | great, thank you iurygregory | 08:48 |
| iurygregory | janders, yw | 08:49 |
| *** mkrai has joined #openstack-ironic | 08:56 | |
| openstackgerrit | Vishal Manchanda proposed openstack/ironic-ui master: [DNM] Testing CI https://review.opendev.org/750599 | 08:57 |
| openstackgerrit | Ameya Raut proposed openstack/ironic stable/train: DRAC: Fix RAID create_config clean step https://review.opendev.org/750605 | 09:13 |
| *** Lucas_Gray has quit IRC | 09:25 | |
| *** mkrai has quit IRC | 09:31 | |
| *** mkrai_ has joined #openstack-ironic | 09:31 | |
| *** Lucas_Gray has joined #openstack-ironic | 09:34 | |
| Qianbiao | iurygregory i am wondering how dib knows which version of ironic-python-agent to use? | 09:34 |
| *** dtantsur|afk is now known as dtantsur | 09:35 | |
| dtantsur | Qianbiao: I think dib uses master by default | 09:35 |
| dtantsur | morning ironic | 09:35 |
| iurygregory | Qianbiao, you need to set the branch you want | 09:35 |
| Qianbiao | morning dtantsur. | 09:36 |
| iurygregory | version* | 09:36 |
| iurygregory | morning dtantsur | 09:36 |
| Qianbiao | any doc tells how? | 09:36 |
| * iurygregory cheks | 09:37 | |
| iurygregory | -b or --branch arg should do | 09:38 |
| *** ricolin_ has quit IRC | 09:38 | |
| iurygregory | https://docs.openstack.org/ironic-python-agent-builder/latest/admin/dib.html explains | 09:39 |
| iurygregory | export DIB_REPOREF_ironic_python_agent... export DIB_REPOREF_requirements | 09:39 |
| Qianbiao | urygregory I am using dib without ipa-builder.. | 09:39 |
| Qianbiao | you said ipa-builder is not for python2 now. | 09:39 |
| iurygregory | this is for disk-image... | 09:40 |
| Qianbiao | hmm, it requires a path to ipa-builder | 09:41 |
| iurygregory | oh wait | 09:41 |
| Qianbiao | export ELEMENTS_PATH=$HOME/.local/share/ironic-python-agent-builder/dib | 09:41 |
| Qianbiao | So, i thought it requries ipa-builder. | 09:41 |
| iurygregory | ipa-builder would work for py2 but I'm not sure the version you need to use for that | 09:42 |
| iurygregory | I don't remember when we dropped and what version you will need to specific install | 09:42 |
| Qianbiao | I can build on python3 env. | 09:42 |
| Qianbiao | Not sure if i build it on python3, will it work with ironic run in python2? | 09:43 |
| Qianbiao | I guess they are two different system, one run in baremetal node. | 09:43 |
| iurygregory | dtantsur, please refresh my memory we can build things with ipa builder master for this case? | 09:44 |
| *** Abdallahyas has joined #openstack-ironic | 09:44 | |
| Qianbiao | And whether the ipa version must be the same as ironic version. dtantsur. | 09:45 |
| dtantsur | IPA version should be the same as ironic version, but sometimes people use versions that differ | 09:45 |
| Qianbiao | if it's not required, i can simple download from openstack tarballs ：） | 09:45 |
| *** abdysn has quit IRC | 09:45 | |
| dtantsur | which versions are we talking about? given that you mention Python 2, train or older? | 09:45 |
| iurygregory | train | 09:45 |
| Qianbiao | yes train. | 09:46 |
| Qianbiao | what I want is building a deploy ramdisk to work with ironic train run with python27. | 09:46 |
| dtantsur | I think you should be able to build IPA with IPA-builder on CentOS 7 and the right argument | 09:46 |
| dtantsur | (or any other distribution using python 2) | 09:46 |
| Qianbiao | <dtantsur> is ipa-builder version required | 09:47 |
| Qianbiao | or i can just install with pip latest | 09:47 |
| dtantsur | ipa-builder has no branches | 09:47 |
| Qianbiao | You mean i need to specify a ipa version, but not ipa-builder version. | 09:48 |
| Qianbiao | ok, will try. | 09:48 |
| dtantsur | yep, just take the latest ipa-builder | 09:48 |
| dtantsur | speaking of the builder, I could use a review on https://review.opendev.org/#/c/749975/ | 09:51 |
| patchbot | patch 749975 - ironic-python-agent-builder - Install libssl-dev / openssl-devel when building IPA - 1 patch set | 09:51 |
| iurygregory | I will add to my list =) | 09:51 |
| iurygregory | dtantsur, if possible can you look later at https://review.opendev.org/#/c/750403/ | 09:51 |
| patchbot | patch 750403 - ironic-prometheus-exporter - Update ram to avoid Kernel Panic - 2 patch sets | 09:51 |
| iurygregory | to fix the IPE CI | 09:51 |
| dtantsur | heh, topic:dnm probably made it disappear from my dashboard | 09:59 |
| dtantsur | approved | 09:59 |
| iurygregory | heheh sorry | 10:00 |
| iurygregory | first patch I pushed with dnm | 10:00 |
| *** k_mouza has joined #openstack-ironic | 10:01 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/bifrost master: [PoC] A naive attempt at creating an upgrade job https://review.opendev.org/750376 | 10:05 |
| Qianbiao | hi dtantsur, if i build ramdisk based on centos8, did i need to run on centos? | 10:06 |
| Qianbiao | it alerts me with on ubuntu: Failed to download initial packages: basesystem filesystem setup centos-release centos-repos centos-gpg-keys | 10:07 |
| dtantsur | Qianbiao: if you need python 2, don't build a ramdisk on centos 8 | 10:12 |
| dtantsur | you should be able to build it on ubuntu though, the error must have something to do with repository mirrors | 10:12 |
| Qianbiao | dtantsur I am wondering why python version relevant | 10:13 |
| dtantsur | Qianbiao: centos 8 does not have python 2 | 10:13 |
| Qianbiao | IPA can run on python3 in this centos 8 ramdisk? | 10:14 |
| Qianbiao | It should have none connect with build os python env | 10:14 |
| *** vmud213 has joined #openstack-ironic | 10:18 | |
| dtantsur | this is correct. I think you have problems reaching the centos mirrors. | 10:19 |
| openstackgerrit | Dmitry Tantsur proposed openstack/bifrost master: [PoC] A naive attempt at creating an upgrade job https://review.opendev.org/750376 | 10:20 |
| openstackgerrit | Iury Gregory Melo Ferreira proposed openstack/ironic-prometheus-exporter master: Fallback to `node_uuid` if`node_name` is not present https://review.opendev.org/723176 | 10:20 |
| Qianbiao | ok, willl check. thanks, dtantsur | 10:21 |
| *** vmud213 has quit IRC | 10:25 | |
| *** vmud213 has joined #openstack-ironic | 10:27 | |
| *** vmud213 has quit IRC | 10:39 | |
| *** vmud213 has joined #openstack-ironic | 10:41 | |
| *** priteau has joined #openstack-ironic | 10:42 | |
| openstackgerrit | Merged openstack/ironic-prometheus-exporter master: Update ram to avoid Kernel Panic https://review.opendev.org/750403 | 10:45 |
| *** rcernin has joined #openstack-ironic | 10:52 | |
| *** zbitter has joined #openstack-ironic | 10:54 | |
| *** zaneb has quit IRC | 10:55 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/bifrost master: Force updating repos with copy_from_local_path https://review.opendev.org/750617 | 10:58 |
| openstackgerrit | Dmitry Tantsur proposed openstack/bifrost master: Force updating repos with copy_from_local_path https://review.opendev.org/750617 | 11:01 |
| openstackgerrit | Dmitry Tantsur proposed openstack/bifrost master: [PoC] A naive attempt at creating an upgrade job https://review.opendev.org/750376 | 11:02 |
| *** thiagop has joined #openstack-ironic | 11:10 | |
| *** Abdallahyas has quit IRC | 11:19 | |
| *** abdysn has joined #openstack-ironic | 11:23 | |
| *** mkrai_ has quit IRC | 11:27 | |
| *** rcernin has quit IRC | 11:28 | |
| openstackgerrit | paresh sao proposed openstack/ironic master: Follow-up patch for One Button Secure Erase clean step https://review.opendev.org/750624 | 11:30 |
| *** lucasagomes has quit IRC | 11:32 | |
| *** lucasagomes has joined #openstack-ironic | 11:32 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/bifrost master: [PoC] A naive attempt at creating an upgrade job https://review.opendev.org/750376 | 11:33 |
| *** mkrai has joined #openstack-ironic | 11:37 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/bifrost master: Force updating repos with copy_from_local_path https://review.opendev.org/750617 | 11:39 |
| *** rcernin has joined #openstack-ironic | 11:45 | |
| *** Qianbiao has quit IRC | 11:46 | |
| *** Qianbiao has joined #openstack-ironic | 11:47 | |
| *** thiagop has quit IRC | 11:54 | |
| *** thiagop has joined #openstack-ironic | 11:55 | |
| *** rh-jelabarre has joined #openstack-ironic | 11:58 | |
| *** rh-jelabarre has joined #openstack-ironic | 11:59 | |
| openstackgerrit | Merged openstack/ironic-python-agent-builder master: Install libssl-dev / openssl-devel when building IPA https://review.opendev.org/749975 | 12:00 |
| *** Goneri has joined #openstack-ironic | 12:07 | |
| *** jamesdenton has quit IRC | 12:08 | |
| *** jamesdenton has joined #openstack-ironic | 12:09 | |
| *** jhesketh has quit IRC | 12:17 | |
| *** jhesketh has joined #openstack-ironic | 12:17 | |
| openstackgerrit | Radosław Piliszek proposed openstack/bifrost master: [WIP] [CI] Check with Kolla Ansible https://review.opendev.org/750637 | 12:19 |
| Qianbiao | Quite today. :) | 12:22 |
| Qianbiao | * Quiet | 12:23 |
| openstackgerrit | Radosław Piliszek proposed openstack/bifrost master: [WIP] [CI] Check with Kolla Ansible https://review.opendev.org/750637 | 12:23 |
| yoctozepto | morning ironic | 12:24 |
| Qianbiao | hi yoctozepto | 12:24 |
| yoctozepto | don't mind me too much on bifrost for now, I am testing the testing framework :-) | 12:24 |
| yoctozepto | Qianbiao: hi! | 12:24 |
| *** Abdallahyas has joined #openstack-ironic | 12:27 | |
| *** abdysn has quit IRC | 12:30 | |
| *** derekh has quit IRC | 12:31 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/bifrost master: Force updating repos with copy_from_local_path https://review.opendev.org/750617 | 12:33 |
| openstackgerrit | Dmitry Tantsur proposed openstack/bifrost master: [PoC] A naive attempt at creating an upgrade job https://review.opendev.org/750376 | 12:34 |
| openstackgerrit | Radosław Piliszek proposed openstack/bifrost master: [DNM] Dummy https://review.opendev.org/750195 | 12:37 |
| *** rloo has joined #openstack-ironic | 12:39 | |
| *** rcernin has quit IRC | 12:41 | |
| openstackgerrit | Jacob Anders proposed openstack/ironic master: Generic way to configure clean step priorites https://review.opendev.org/744117 | 12:43 |
| janders | see you tomorrow Ironic o/ | 12:46 |
| iurygregory | bye janders o/ | 12:47 |
| openstackgerrit | Christopher Dearborn proposed openstack/ironic master: WIP: Redfish driver firmware update https://review.opendev.org/749619 | 13:05 |
| *** derekh has joined #openstack-ironic | 13:06 | |
| *** rcernin has joined #openstack-ironic | 13:14 | |
| openstackgerrit | Christopher Dearborn proposed openstack/ironic master: Redfish driver firmware update https://review.opendev.org/749619 | 13:20 |
| *** dougsz has quit IRC | 13:22 | |
| openstackgerrit | Christopher Dearborn proposed openstack/ironic master: Redfish driver firmware update https://review.opendev.org/749619 | 13:25 |
| *** jhesketh has quit IRC | 13:31 | |
| *** mkrai has quit IRC | 13:32 | |
| *** mkrai_ has joined #openstack-ironic | 13:33 | |
| *** dougsz has joined #openstack-ironic | 13:35 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/bifrost master: test-bifrost: use existing playbooks https://review.opendev.org/750656 | 13:39 |
| *** zbitter is now known as zaneb | 13:40 | |
| *** tzumainn has joined #openstack-ironic | 13:40 | |
| vmud213 | Hello Ironic | 13:41 |
| *** rcernin has quit IRC | 13:41 | |
| openstackgerrit | Merged openstack/ironic-python-agent-builder master: Add element to configure IPA with TLS, use configdir https://review.opendev.org/747309 | 13:41 |
| openstackgerrit | Julia Kreger proposed openstack/metalsmith master: Rework jobs so failed devstack does not retry the job https://review.opendev.org/749758 | 13:42 |
| TheJulia | Good morning | 13:42 |
| vmud213 | iurygregory, dtantsur or anyone if u have some time can you take a look into https://review.opendev.org/#/c/745501/ | 13:42 |
| patchbot | patch 745501 - ironic - Adds ilo-uefi-https boot interface to ilo5 - 11 patch sets | 13:42 |
| iurygregory | good morning TheJulia =) | 13:42 |
| iurygregory | hey vmud213 I will add to my list | 13:43 |
| vmud213 | Thanks | 13:43 |
| TheJulia | vmud213: You have changeset feedback :) | 13:54 |
| vmud213 | Thanks TheJulia for the review. Will look into it. | 13:55 |
| dtantsur | morning TheJulia | 13:59 |
| openstackgerrit | Dmitry Tantsur proposed openstack/bifrost master: Re-download checkums on upgrades and updates https://review.opendev.org/750665 | 13:59 |
| dtantsur | the new upgrade job finding actual upgrade issues ^^ | 13:59 |
| TheJulia | fun | 14:00 |
| openstackgerrit | Dmitry Tantsur proposed openstack/bifrost master: [PoC] A naive attempt at creating an upgrade job https://review.opendev.org/750376 | 14:00 |
| openstackgerrit | Julia Kreger proposed openstack/metalsmith master: Rework jobs so failed devstack does not retry the job https://review.opendev.org/749758 | 14:05 |
| *** ricolin_ has joined #openstack-ironic | 14:11 | |
| *** ricolin_ has quit IRC | 14:13 | |
| *** mkrai_ has quit IRC | 14:15 | |
| iurygregory | "In ONE HOUR the schedule for the Virtual #OpenInfraSummit will be launched!" does that means we will know if our talks were accepted? O.o | 14:38 |
| TheJulia | yes | 14:39 |
| iurygregory | =O | 14:39 |
| *** outbrito has joined #openstack-ironic | 14:40 | |
| iurygregory | now I'm nervous | 14:40 |
| *** thiagop has quit IRC | 14:43 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/bifrost master: [WIP] test-bifrost: use existing playbooks https://review.opendev.org/750656 | 14:45 |
| *** k_mouza has quit IRC | 14:48 | |
| *** k_mouza has joined #openstack-ironic | 14:48 | |
| Qianbiao | Morning, TheJulia | 14:49 |
| *** mkrai has joined #openstack-ironic | 14:49 | |
| Qianbiao | Hello, when ipxe, it complains "no_kernel... No such file or directory". logs: http://paste.openstack.org/show/797651/ | 14:50 |
| Qianbiao | when should i start for this issue. | 14:50 |
| *** outbrito is now known as thiagop | 14:51 | |
| Qianbiao | node json: http://paste.openstack.org/show/797652/ | 14:53 |
| jamesdenton | Good morning TheJulia - couple of comments here if you have time: https://review.opendev.org/#/c/557051/ | 14:54 |
| patchbot | patch 557051 - networking-generic-switch - Support multiple links in link_local_information d... - 13 patch sets | 14:54 |
| *** k_mouza has quit IRC | 14:58 | |
| JayF | looking at that story I was pinged about yesterday | 14:59 |
| iurygregory | OMG | 15:01 |
| openstackgerrit | Radosław Piliszek proposed openstack/bifrost master: [WIP] [CI] Check with Kolla Ansible https://review.opendev.org/750637 | 15:02 |
| *** Abdallahyas has quit IRC | 15:03 | |
| TheJulia | jamesdenton: I can look a little later | 15:03 |
| TheJulia | iurygregory: yes? | 15:03 |
| TheJulia | iurygregory: is it beer time? | 15:03 |
| jamesdenton | :thumbsup" | 15:03 |
| iurygregory | YES! | 15:03 |
| TheJulia | congrats | 15:03 |
| iurygregory | and I'm crying LMAO | 15:03 |
| *** k_mouza has joined #openstack-ironic | 15:04 | |
| JayF | dtantsur: TheJulia: https://storyboard.openstack.org/#!/story/2008113 commented on this story as reqested. Sorry I was OOO yesterday with a migraine | 15:06 |
| iurygregory | congrats TheJulia arne_wiebalck janders \o/ | 15:06 |
| TheJulia | yay | 15:07 |
| dtantsur | \o/ | 15:08 |
| Qianbiao | TheJulia, help me with the issue upon ^^^ when u got a minute. ramdisk deploy interface could not work :( | 15:08 |
| TheJulia | JayF: yeah, The telecoms basically made an argument that they really only needed disks wiped for ceph nodes since they are not in multitenant operations which kind of made sense which is how the additional flexibility existed | 15:08 |
| Qianbiao | This may cause the issue: under ~/data/ironic/httpboot/49272d09-5acc-4390-840a-6c7a8285dbf7, there are only a config file, no ramdisk file. | 15:08 |
| dtantsur | JayF: many thanks, responded | 15:09 |
| JayF | TheJulia: I'm sure given details I could find a reason they should care to have them wiped in all cases, at least the metadata, but given those conversations happen downstream, and like I said in the RFE the cat's already out of the bag, it's just not worth fighting it anymore | 15:09 |
| JayF | I could craft a disk image that would almost certainly break the next deploy if not at least metadata-cleaned beforehand. | 15:10 |
| JayF | I guess I find it difficult to nail down the use case for "we don't even wanna run `erase_devices_metadata` between deploys" | 15:13 |
| *** penick has joined #openstack-ironic | 15:13 | |
| JayF | unless you're actually attempting to preserve data between deployments, which is not a feature we support | 15:13 |
| TheJulia | JayF: Actually I remember they were argued upstream | 15:13 |
| JayF | eh, the point is more, I wasn't here, the decision got made, cat's outta the abg | 15:14 |
| JayF | there's no point in holding a position on semantics when in practice it's already been invalidated | 15:14 |
| JayF | Yeah and literally the metal3 thread about this is using circumvention of cleaning to get "keep data between installs" as a feature | 15:17 |
| JayF | that's the ugliest possible way to provide that feature | 15:17 |
| JayF | I object more to that existing for that use case than I object to it existing generally :( | 15:17 |
| JayF | If that's a desired feature, isn't that what preserve-ephemeral+rebuilding is for? | 15:18 |
| JayF | (if it still exists?) | 15:18 |
| openstackgerrit | Dmitry Tantsur proposed openstack/bifrost master: test-bifrost: use existing playbooks https://review.opendev.org/750656 | 15:18 |
| dtantsur | JayF: ephemeral stuff only works for partition images, I think? | 15:19 |
| JayF | I'm not sure, but I'd prefer a world where we tried to make our actual API for preserving data work for their use case | 15:19 |
| dtantsur | rebuilding should have worked for them, were it not for strange interactions inside the cluster API | 15:19 |
| *** mkrai has quit IRC | 15:20 | |
| *** akahat|rover is now known as akahat | 15:20 | |
| dtantsur | I don't like it, but I like even less when people start defaulting to no-cleaning | 15:21 |
| JayF | Any way you go about it, it's circumventing a key security feature to get backdoor support for a feature that /we already support/ | 15:22 |
| openstackgerrit | Radosław Piliszek proposed openstack/bifrost master: [DNM] Dummy https://review.opendev.org/750195 | 15:22 |
| dtantsur | JayF: we don't support it; preserve ephemeral is a weird misuse of a nova feature that I doubt a lot of people are using outside of our CI | 15:22 |
| dtantsur | we definitely don't support a notion of "don't touch disk X" | 15:23 |
| *** jhesketh has joined #openstack-ironic | 15:23 | |
| JayF | I think that feature itself would be more interesting, we even have downstream use cases for it here. | 15:23 |
| dtantsur | I wish somebody could spend cycles designing it ;) | 15:23 |
| JayF | I think the proposed change is "fine" given the fact that config is already supported, even though I don't like that it's already supported either | 15:23 |
| JayF | But it worries me that we're punting on the real user need | 15:24 |
| JayF | dtantsur: just because a situation is caused by lack of resources doesn't mean we shouldn't acknowledge that it's hacky or non-ideal | 15:24 |
| JayF | dtantsur: making explicit decisions about stuff like that is better than having it happen implicitly | 15:24 |
| dtantsur | right now I'm trying to prevent yet another case of "just disable cleaning" as a solution for weird problems | 15:25 |
| JayF | My point is this is just another permutation of "just disable cleaning" | 15:25 |
| JayF | arguably worse because it's easier to forget it being disabled on a per-node basis | 15:25 |
| arne_wiebalck | iurygregory: thanks, to you as well :) | 15:26 |
| JayF | but if folks are already turning it off, then turning it back on per-node (didn't know this existed until today), then this is clearly better | 15:26 |
| * arne_wiebalck goes on a long weekend, cu all on Monday o/ | 15:26 | |
| openstackgerrit | Radosław Piliszek proposed openstack/bifrost master: [WIP] [CI] Check with Kolla Ansible https://review.opendev.org/750637 | 15:26 |
| openstackgerrit | Radosław Piliszek proposed openstack/bifrost master: [DNM] Dummy https://review.opendev.org/750195 | 15:26 |
| JayF | I just don't think how people use that now, or how people will use this is particularly good :/ | 15:26 |
| TheJulia | I suspect the disjointed nature is use nature disconnect. Cloud versus fleet deployment for specific purpose and then that base node goes *poof* or gets powered off | 15:27 |
| openstackgerrit | Dmitry Tantsur proposed openstack/ironic master: Accept and use a TLS certificate from the agent https://review.opendev.org/747136 | 15:27 |
| dtantsur | TheJulia: with a unit test ^^^ | 15:27 |
| TheJulia | (not that that is always a thing) :( | 15:28 |
| TheJulia | dtantsur: thanks! | 15:28 |
| dtantsur | after this conversation I am, however, leaning towards disabling this feature by default (via the new policy) | 15:28 |
| JayF | my point is simply this: | 15:29 |
| JayF | a node going into AVAILABLE with data on the disk still | 15:29 |
| JayF | is *not a use case* that we should support | 15:29 |
| JayF | it doens't matter if the client intends to deploy that node immediately afterwards | 15:29 |
| *** Lucas_Gray has quit IRC | 15:29 | |
| JayF | that's not a great state to leave a node in for a stable state | 15:29 |
| JayF | I think there are potentially better workarounds for that than disabling cleaning, too | 15:30 |
| JayF | ("better" meaning "less gross" not "good") | 15:30 |
| *** penick has quit IRC | 15:30 | |
| *** Lucas_Gray has joined #openstack-ironic | 15:30 | |
| TheJulia | sigh, reworking metalsmith CI is quickly becoming a nightmare | 15:30 |
| *** zzzeek has quit IRC | 15:31 | |
| TheJulia | yeah, I concur, the downside that I've also seen is people start ripping nodes out of ironic and then trying to wedge them back in :\ | 15:31 |
| TheJulia | because they want to evade the cleaning | 15:31 |
| TheJulia | :( | 15:31 |
| JayF | that would be the less-gross workaround I refer to | 15:31 |
| JayF | because at least then, the user is aware they're doing something outside of the normal ironic workflow | 15:31 |
| *** zzzeek has joined #openstack-ironic | 15:32 | |
| dtantsur | at least deploy_interface=ramdisk may be a good reason to avoid cleaning for a specific node | 15:34 |
| dtantsur | (putting aside weird users with weird requirements) | 15:34 |
| JayF | as a note, it'll be a "free" benefit of anaconda/kickstart deploy driver | 15:35 |
| JayF | that you'll be able to rebuild with preserve | 15:35 |
| JayF | because kickstart supports preserving a disk | 15:35 |
| dtantsur | nice! | 15:35 |
| JayF | Hmmm. zer0c00l have we thought about rebuilds with the anaconda driver? Will we support passing in a different kickstart file for a rebuild? | 15:36 |
| *** zzzeek has quit IRC | 15:37 | |
| *** zzzeek has joined #openstack-ironic | 15:38 | |
| rloo | JayF: yes -- we already support that downstream, wrt rebuild + diff kickstart. I guess it ought to be mentioned in the spec, zer0c00l | 15:39 |
| openstackgerrit | Radosław Piliszek proposed openstack/bifrost master: [DNM] check https://review.opendev.org/750689 | 15:40 |
| *** belmoreira has quit IRC | 15:40 | |
| *** zzzeek has quit IRC | 15:42 | |
| TheJulia | Qianbiao: so what is going on with the ramdisk interface? now files are not being plaed when you attempt to deploy? | 15:44 |
| *** zzzeek has joined #openstack-ironic | 15:45 | |
| Qianbiao | TheJulia the console complains "no_kernel... No such file or directory" | 15:46 |
| TheJulia | Qianbiao: what is in your instance_info field? | 15:46 |
| Qianbiao | and it seems there did not has ramdisk file under ~/data/ironic/httpboot/49272d09-5acc-4390-840a-6c7a8285dbf7 | 15:47 |
| Qianbiao | http://paste.openstack.org/show/797652/ | 15:47 |
| *** penick has joined #openstack-ironic | 15:47 | |
| openstackgerrit | Julia Kreger proposed openstack/metalsmith master: Rework jobs so failed devstack does not retry the job https://review.opendev.org/749758 | 15:49 |
| *** zzzeek has quit IRC | 15:50 | |
| TheJulia | Qianbiao: so what does glance show for that image you have in image_source ? | 15:50 |
| Qianbiao | it exists. | 15:50 |
| *** zzzeek has joined #openstack-ironic | 15:51 | |
| TheJulia | but what does it have in the records? is it a whole-disk-image? is it a partition image with a kernel/ramdisk? | 15:51 |
| TheJulia | I suspect it may be the wrong image configuration | 15:51 |
| TheJulia | so good to check that | 15:51 |
| Qianbiao | http://paste.openstack.org/show/797659/ | 15:51 |
| Qianbiao | TheJulia it has a required format? | 15:52 |
| Qianbiao | current it is qcow2 | 15:52 |
| TheJulia | so, the ramdisk interface will only use a kernel/ramdisk off an image in glance | 15:53 |
| TheJulia | that image, if used that way must be a partition image, but the image contents are ignored | 15:53 |
| TheJulia | it is documented that way, or at least that is the way I wrote the docs ages ago | 15:53 |
| TheJulia | so what you've supplied it is a whole disk image | 15:54 |
| TheJulia | there is no way to directly ramdisk boot that | 15:54 |
| TheJulia | because no kernel/ramdisk is present | 15:54 |
| Qianbiao | TheJulia thanks isee. | 15:54 |
| Qianbiao | will try using a UEC image | 15:54 |
| TheJulia | we _likely_ ought to raise an exception in the case your hitting | 15:54 |
| *** zzzeek has quit IRC | 15:56 | |
| Qianbiao | TheJulia yes, if not ramdisk could be generated, an exception should raise | 15:58 |
| TheJulia | well, it is not generated, just passed through | 15:58 |
| *** zzzeek has joined #openstack-ironic | 15:59 | |
| *** k_mouza has quit IRC | 16:00 | |
| Qianbiao | TheJulia Thanks, big help. will try new image | 16:00 |
| *** zzzeek has quit IRC | 16:04 | |
| *** zzzeek has joined #openstack-ironic | 16:05 | |
| *** lucasagomes has quit IRC | 16:05 | |
| *** k_mouza has joined #openstack-ironic | 16:05 | |
| *** Lucas_Gray has quit IRC | 16:12 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/bifrost master: test-bifrost: use existing playbooks https://review.opendev.org/750656 | 16:15 |
| vmud213 | i have a question. When we add a new configuration parameter, do we mention that any where in the documentation? | 16:18 |
| TheJulia | vmud213: depends on the context of the change | 16:18 |
| vmud213 | ok. | 16:19 |
| TheJulia | often a release note is used and if there is something it ties in with we prefer to see the appropriate documentation also updated | 16:19 |
| TheJulia | if it doesn't get updated, then it is easy to miss | 16:19 |
| *** Lucas_Gray has joined #openstack-ironic | 16:20 | |
| vmud213 | agree. Thanks TheJulia. I missed that it would be mentioned in the release notes | 16:20 |
| *** k_mouza has quit IRC | 16:21 | |
| *** k_mouza has joined #openstack-ironic | 16:22 | |
| *** dougsz has quit IRC | 16:25 | |
| *** k_mouza has quit IRC | 16:27 | |
| *** k_mouza has joined #openstack-ironic | 16:27 | |
| openstackgerrit | OpenStack Release Bot proposed openstack/python-ironicclient stable/victoria: Update .gitreview for stable/victoria https://review.opendev.org/750719 | 16:36 |
| openstackgerrit | OpenStack Release Bot proposed openstack/python-ironicclient stable/victoria: Update TOX_CONSTRAINTS_FILE for stable/victoria https://review.opendev.org/750720 | 16:36 |
| openstackgerrit | OpenStack Release Bot proposed openstack/python-ironicclient master: Update master for stable/victoria https://review.opendev.org/750721 | 16:36 |
| openstackgerrit | OpenStack Release Bot proposed openstack/python-ironicclient master: Add Python3 wallaby unit tests https://review.opendev.org/750722 | 16:36 |
| openstackgerrit | OpenStack Release Bot proposed openstack/sushy-cli stable/victoria: Update .gitreview for stable/victoria https://review.opendev.org/750730 | 16:36 |
| *** Qianbiao has quit IRC | 16:36 | |
| openstackgerrit | OpenStack Release Bot proposed openstack/sushy-cli stable/victoria: Update TOX_CONSTRAINTS_FILE for stable/victoria https://review.opendev.org/750732 | 16:36 |
| openstackgerrit | OpenStack Release Bot proposed openstack/sushy-cli master: Update master for stable/victoria https://review.opendev.org/750734 | 16:36 |
| openstackgerrit | OpenStack Release Bot proposed openstack/sushy-cli master: Add Python3 wallaby unit tests https://review.opendev.org/750735 | 16:36 |
| *** vmud213 has quit IRC | 16:43 | |
| *** Lucas_Gray has quit IRC | 16:51 | |
| *** k_mouza has quit IRC | 16:52 | |
| openstackgerrit | Radosław Piliszek proposed openstack/bifrost master: [CI] Check with Kolla Ansible https://review.opendev.org/750637 | 16:53 |
| yoctozepto | morningv2 ironic | 16:53 |
| yoctozepto | please review https://review.opendev.org/750637 | 16:54 |
| patchbot | patch 750637 - bifrost - [CI] Check with Kolla Ansible - 5 patch sets | 16:54 |
| yoctozepto | I know it is blocked by kolla changes needing to merge first but I want to know your opinion on the approach | 16:54 |
| yoctozepto | it does seem to catch the previous issue | 16:54 |
| dtantsur | looks good at first glance, will check the CI results tomorrow | 16:54 |
| yoctozepto | (see the relatated changes) | 16:54 |
| *** dtantsur is now known as dtantsur|afk | 16:55 | |
| TheJulia | looks good to me too | 16:55 |
| dtantsur|afk | o/ | 16:55 |
| yoctozepto | dtantsur|afk: ok, feel free to play with dnm patch however you like too | 16:55 |
| yoctozepto | TheJulia: :-) | 16:55 |
| TheJulia | so... | 16:56 |
| TheJulia | I feel like the job in kolla seems not to rely upon the clones zuul is making | 16:56 |
| yoctozepto | TheJulia: it does with that depends-on | 16:57 |
| yoctozepto | TheJulia: and it detects the breakage in https://review.opendev.org/#/c/750195/4 | 16:57 |
| patchbot | patch 750195 - bifrost - [DNM] Dummy - 4 patch sets | 16:57 |
| TheJulia | ahh, a second one | 16:57 |
| *** Lucas_Gray has joined #openstack-ironic | 16:58 | |
| TheJulia | yeah, I think it is still going to need to get cloned outright by zuul to do the right thing for depends-on tags to fully apply | 16:59 |
| TheJulia | but that is just a "gut feeling" at the moment | 17:00 |
| *** penick has quit IRC | 17:00 | |
| *** derekh has quit IRC | 17:00 | |
| yoctozepto | TheJulia: sure, due to zuul's magic we could actually see some different behaviour once it merges and is run from "pristine" queue | 17:01 |
| yoctozepto | TheJulia: nonetheless, I think I've worked through enough docs to get it right finally | 17:01 |
| yoctozepto | ;d | 17:01 |
| *** Lucas_Gray has quit IRC | 17:03 | |
| *** Lucas_Gray has joined #openstack-ironic | 17:07 | |
| *** k_mouza has joined #openstack-ironic | 17:08 | |
| *** k_mouza has quit IRC | 17:13 | |
| TheJulia | yeah, we did something similar in the early days of bifrost back with zuul 2.5 where we parsed through and applied changes | 17:20 |
| TheJulia | same basic risk :) | 17:20 |
| openstackgerrit | Merged openstack/python-ironicclient master: Update master for stable/victoria https://review.opendev.org/750721 | 17:30 |
| *** Lucas_Gray has quit IRC | 17:31 | |
| *** gyee has joined #openstack-ironic | 17:33 | |
| *** belmoreira has joined #openstack-ironic | 17:35 | |
| openstackgerrit | Merged openstack/sushy-cli stable/victoria: Update .gitreview for stable/victoria https://review.opendev.org/750730 | 17:37 |
| openstackgerrit | Merged openstack/sushy-cli master: Update master for stable/victoria https://review.opendev.org/750734 | 17:47 |
| openstackgerrit | Merged openstack/sushy-cli master: Add Python3 wallaby unit tests https://review.opendev.org/750735 | 17:47 |
| openstackgerrit | Merged openstack/sushy-cli stable/victoria: Update TOX_CONSTRAINTS_FILE for stable/victoria https://review.opendev.org/750732 | 17:47 |
| *** k_mouza has joined #openstack-ironic | 17:57 | |
| *** k_mouza has quit IRC | 18:02 | |
| openstackgerrit | Julia Kreger proposed openstack/metalsmith master: Rework jobs so failed devstack does not retry the job https://review.opendev.org/749758 | 18:32 |
| *** zzzeek has quit IRC | 18:33 | |
| *** zzzeek has joined #openstack-ironic | 18:34 | |
| openstackgerrit | Julia Kreger proposed openstack/metalsmith master: Change metalsmith nodeset to Focal https://review.opendev.org/750489 | 18:35 |
| openstackgerrit | Julia Kreger proposed openstack/metalsmith master: Replace the Centos7 image job with a Centos8 UEFI based job https://review.opendev.org/748315 | 18:41 |
| *** zzzeek has quit IRC | 18:41 | |
| *** zzzeek has joined #openstack-ironic | 18:42 | |
| zer0c00l | JayF Yes rebuilds should work. From ironic perspective isn't it another deploy? | 18:49 |
| zer0c00l | How does it work? | 18:49 |
| *** zzzeek has quit IRC | 18:50 | |
| *** zzzeek has joined #openstack-ironic | 18:50 | |
| TheJulia | it basically is a redeploy on the same hardware but we had a preserve_ephemeral option for way back in the day so tripleo clusters could be upgraded... as all state data should have been on the "ephemeral partition" | 18:53 |
| *** zzzeek has quit IRC | 18:55 | |
| *** zzzeek has joined #openstack-ironic | 18:56 | |
| *** priteau has quit IRC | 18:57 | |
| *** zzzeek has quit IRC | 19:03 | |
| *** zzzeek has joined #openstack-ironic | 19:05 | |
| JayF | I mean, for ks driver, someone will just have to do that in the KS file, right? | 19:09 |
| JayF | seems borderline-impossible to support the ironic opt for that, given that in ks, I think that also would mean you can't do 100% automatic partitioning | 19:10 |
| JayF | zer0c00l: I think it's worth noting that rebuilds will just work, and that if desired the ks file can be changed. Just addressing their behavior generally in the spec would probably be nice. | 19:10 |
| zer0c00l | JayF, TheJulia ack | 19:13 |
| TheJulia | JayF: I believe so yes, it would have to be embedded into the ks file | 19:20 |
| *** zzzeek has quit IRC | 19:21 | |
| *** zzzeek has joined #openstack-ironic | 19:23 | |
| zer0c00l | JayF: To be clear the rebuild will have to preserve_ephemeral partitions? | 19:37 |
| zer0c00l | In that case the rebuild kickstart file has to be designed in a way that it preserves whatever partition they'd like to preserve | 19:37 |
| JayF | That's usually an option that is provided. It's not required that your driver support/honor that. See above comment about it likely being difficult/impossible. | 19:38 |
| JayF | Just would maybe reference that a user can rebuild using a different kickstart file | 19:38 |
| TheJulia | You know, it might not be worth the time and braincells to worry about preserve ephemeral. We have no way of doing such with wholedisk images | 19:38 |
| JayF | and perhaps even indicate that you could preserve a disk partition in the spec (and later in the usage docs?) if you want | 19:38 |
| TheJulia | There has been some discussion of killing preserve_ephemeral too, at least that was in the past. I'm not sure if people really using it | 19:38 |
| JayF | We have strong use cases for it; even if today we implement those in a very-downstream manner. | 19:39 |
| *** hjensas_ has joined #openstack-ironic | 19:48 | |
| *** hjensas has quit IRC | 19:52 | |
| *** k_mouza has joined #openstack-ironic | 20:09 | |
| TheJulia | ok | 20:11 |
| TheJulia | ironic-cores: https://review.opendev.org/#/q/project:openstack/ironic-python-agent+status:open could use some attention review wise | 20:11 |
| *** k_mouza has quit IRC | 20:13 | |
| openstackgerrit | Julia Kreger proposed openstack/metalsmith master: Replace the Centos7 image job with a Centos8 UEFI based job https://review.opendev.org/748315 | 20:15 |
| *** openstackgerrit has quit IRC | 20:17 | |
| *** Lucas_Gray has joined #openstack-ironic | 20:21 | |
| *** k_mouza has joined #openstack-ironic | 20:27 | |
| *** k_mouza has quit IRC | 20:32 | |
| *** stevebaker has joined #openstack-ironic | 20:32 | |
| *** hjensas__ has joined #openstack-ironic | 20:36 | |
| *** hjensas_ has quit IRC | 20:39 | |
| JayF | TheJulia: not a core, but I'll do a review pass on the ones I'm familiar with. | 20:43 |
| *** stevebaker_ has joined #openstack-ironic | 20:52 | |
| *** stevebaker has quit IRC | 20:52 | |
| TheJulia | JayF: thanks | 21:01 |
| JayF | TheJulia: https://review.opendev.org/#/c/750097 you may want to remove your +2 from this, see my comment -- I'm almost certain removing StrictVersion() is a breaking change | 21:05 |
| patchbot | patch 750097 - ironic-python-agent - Refactor API version negotiation code - 1 patch set | 21:05 |
| *** adrianc has quit IRC | 21:05 | |
| *** adrianc has joined #openstack-ironic | 21:05 | |
| *** k_mouza has joined #openstack-ironic | 21:07 | |
| *** k_mouza has quit IRC | 21:12 | |
| *** thiagop has quit IRC | 21:12 | |
| *** openstackgerrit has joined #openstack-ironic | 21:13 | |
| openstackgerrit | Verification of a change to openstack/ironic-python-agent failed: Prepare to move ipa-builder jobs https://review.opendev.org/746766 | 21:13 |
| TheJulia | it definitely changes the behavior a little, breaking I'm not entirely sure about but yeah. I just wish I remembered why the strictversion route was gone down | 21:15 |
| JayF | It is absolutely broken with that change | 21:15 |
| JayF | with any API with two-digit versions | 21:15 |
| JayF | min(1.25, 1.3) = 1.3 | 21:15 |
| JayF | er, I mean, 1.25 | 21:15 |
| JayF | min(1.25, 1.3) = 1.25 | 21:15 |
| JayF | whereas the strictversion understands that it's not a decimal number, it's a version string that should be parsed as two dot-separated numbers | 21:16 |
| *** k_mouza has joined #openstack-ironic | 21:16 | |
| JayF | If it's not clear, I can make a code snippet to demonstrate. I don't want that to merge with the change, I'm more certain it's broken the more I think about it. | 21:17 |
| *** k_mouza has quit IRC | 21:21 | |
| *** adrianc_ has joined #openstack-ironic | 21:22 | |
| *** adrianc has quit IRC | 21:22 | |
| *** k_mouza has joined #openstack-ironic | 21:24 | |
| *** k_mouza has quit IRC | 21:29 | |
| JayF | I'm wrong | 21:29 |
| JayF | TheJulia: https://gist.github.com/jayofdoom/cef64a13595042ca655d1e9fd88fe080 | 21:29 |
| TheJulia | RIGHT! | 21:30 |
| TheJulia | gah | 21:30 |
| TheJulia | interesting | 21:31 |
| TheJulia | It was somethign along the lines of where 1.3 was larger than 1.25, but only as a real number | 21:32 |
| JayF | I think the fact versions are all tuples completely negates that line of breakage | 21:32 |
| JayF | but IMBW | 21:32 |
| JayF | I'm never convinced something works, but I'm not convinced it's broken now so +1 :D | 21:33 |
| TheJulia | yeah, there was a case where we did have to use strictversion because we were dealing with tuples but strings. StritVersion does at least provide a little more future proofing so we don't have to revamp the code on major version changes | 21:34 |
| JayF | Hmmm. | 21:36 |
| JayF | Nah, I can't find a breakage with major version number changes, eithe | 21:37 |
| JayF | e.g. Min of 1.99 and 2.25 | 21:38 |
| JayF | (1, 99) | 21:38 |
| openstackgerrit | Jay Faulkner proposed openstack/ironic-python-agent-builder master: Followups for TLS support https://review.opendev.org/750818 | 21:55 |
| *** jtomasek has quit IRC | 21:56 | |
| JayF | ^ would appreciate a quick review/merge cycle on that, the merged DIB TLS element is broken. If the approach on tinyipa is not ideal, I can remove it into a separate change | 21:57 |
| JayF | actually, still broken, going to W-1 it | 21:58 |
| openstackgerrit | Julia Kreger proposed openstack/metalsmith master: Replace the Centos7 image job with a Centos8 UEFI based job https://review.opendev.org/748315 | 22:09 |
| *** Lucas_Gray has quit IRC | 22:13 | |
| *** k_mouza has joined #openstack-ironic | 22:18 | |
| openstackgerrit | Julia Kreger proposed openstack/ironic-python-agent master: Fix backup node lookup https://review.opendev.org/750823 | 22:19 |
| *** k_mouza has quit IRC | 22:23 | |
| *** rloo has quit IRC | 22:24 | |
| *** k_mouza has joined #openstack-ironic | 22:26 | |
| *** k_mouza has quit IRC | 22:31 | |
| janders | good morning Ironic o/ | 22:38 |
| *** k_mouza has joined #openstack-ironic | 22:43 | |
| *** k_mouza has quit IRC | 22:47 | |
| *** k_mouza has joined #openstack-ironic | 22:51 | |
| *** rcernin has joined #openstack-ironic | 22:54 | |
| openstackgerrit | Jay Faulkner proposed openstack/ironic-python-agent-builder master: Followups for TLS support https://review.opendev.org/750818 | 22:55 |
| JayF | removed the tinyipa bits, just fixing the dib element ^ | 22:55 |
| *** rcernin has quit IRC | 22:56 | |
| *** k_mouza has quit IRC | 22:56 | |
| *** rcernin has joined #openstack-ironic | 22:56 | |
| *** zzzeek has quit IRC | 23:17 | |
| *** zzzeek has joined #openstack-ironic | 23:19 | |
| *** Goneri has quit IRC | 23:21 | |
| *** zzzeek has quit IRC | 23:38 | |
| *** zzzeek has joined #openstack-ironic | 23:40 | |
| *** k_mouza has joined #openstack-ironic | 23:47 | |
| *** k_mouza has quit IRC | 23:52 | |
| *** zzzeek has quit IRC | 23:59 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!