| opendevreview | Hayato Morita proposed openstack/ironic master: [WIP][Redfish] Add warning log to notify no means to pass token https://review.opendev.org/c/openstack/ironic/+/854735 | 00:45 |
|---|---|---|
| opendevreview | SONG SHUKUN proposed openstack/ironic stable/xena: Add SNMPv3 authentication functionality https://review.opendev.org/c/openstack/ironic/+/850554 | 01:40 |
| opendevreview | SONG SHUKUN proposed openstack/ironic stable/yoga: Add SNMPv3 authentication functionality https://review.opendev.org/c/openstack/ironic/+/850553 | 01:44 |
| opendevreview | OpenStack Proposal Bot proposed openstack/ironic master: Imported Translations from Zanata https://review.opendev.org/c/openstack/ironic/+/856004 | 04:05 |
| opendevreview | OpenStack Proposal Bot proposed openstack/ironic-inspector master: Imported Translations from Zanata https://review.opendev.org/c/openstack/ironic-inspector/+/856010 | 04:20 |
| opendevreview | Michal Nasiadka proposed openstack/tenks master: Add support for EL9 https://review.opendev.org/c/openstack/tenks/+/854550 | 05:45 |
| opendevreview | Michal Nasiadka proposed openstack/tenks master: Add support for EL9 https://review.opendev.org/c/openstack/tenks/+/854550 | 05:45 |
| opendevreview | Michal Nasiadka proposed openstack/tenks master: Fix ansible-lint errors https://review.opendev.org/c/openstack/tenks/+/856023 | 05:58 |
| opendevreview | Michal Nasiadka proposed openstack/tenks master: Add support for EL9 https://review.opendev.org/c/openstack/tenks/+/854550 | 05:59 |
| arne_wiebalck | Good morning, Ironic! | 06:53 |
| opendevreview | Michal Nasiadka proposed openstack/tenks master: Add support for EL9 https://review.opendev.org/c/openstack/tenks/+/854550 | 07:07 |
| opendevreview | Michal Nasiadka proposed openstack/tenks master: Fix ansible-lint errors https://review.opendev.org/c/openstack/tenks/+/856023 | 07:22 |
| opendevreview | Michal Nasiadka proposed openstack/tenks master: Fix ansible-lint errors https://review.opendev.org/c/openstack/tenks/+/856023 | 07:35 |
| opendevreview | Michal Nasiadka proposed openstack/tenks master: Add support for EL9 https://review.opendev.org/c/openstack/tenks/+/854550 | 08:02 |
| opendevreview | wangjiaqi proposed openstack/tenks master: Use TOX_CONSTRAINTS_FILE https://review.opendev.org/c/openstack/tenks/+/851821 | 08:31 |
| opendevreview | Radosław Piliszek proposed openstack/tenks master: Use TOX_CONSTRAINTS_FILE https://review.opendev.org/c/openstack/tenks/+/851821 | 08:44 |
| opendevreview | Aija Jauntēva proposed openstack/ironic master: Fix idrac-redfish RAID controller mode conversion https://review.opendev.org/c/openstack/ironic/+/855872 | 09:16 |
| ajya | iurygregory: added story ^ | 09:17 |
| ftarasenko | Gm Ironic! Team, need advice from you. My tenant creates software raid from target instance (he got standalone nvme drives and created mdraid). After he undeploys instance from baremetal server, it goes to cleaning, but RAID Member device is not cleaned because of _is_linux_raid_member function, only md device is shred instead of secure erase. I see that this is expected behaviour, but maybe it's not correct? | 09:26 |
| opendevreview | Michal Nasiadka proposed openstack/tenks master: Add support for EL9 https://review.opendev.org/c/openstack/tenks/+/854550 | 09:53 |
| opendevreview | Michal Nasiadka proposed openstack/tenks master: Add support for EL9 https://review.opendev.org/c/openstack/tenks/+/854550 | 10:00 |
| opendevreview | Michal Nasiadka proposed openstack/tenks master: Add support for EL9 https://review.opendev.org/c/openstack/tenks/+/854550 | 10:11 |
| opendevreview | Merged openstack/tenks master: Fix ansible-lint errors https://review.opendev.org/c/openstack/tenks/+/856023 | 10:17 |
| opendevreview | Merged openstack/tenks master: Use TOX_CONSTRAINTS_FILE https://review.opendev.org/c/openstack/tenks/+/851821 | 10:17 |
| iurygregory | good morning Ironic | 10:59 |
| iurygregory | ftarasenko, if I recall this is the expected behaviour, if you want nvme to go trough cleaning there is an option you need to set, janders worked on it (let me see if I can find) | 11:01 |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/train: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852806 | 11:38 |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/victoria: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852804 | 11:39 |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/ussuri: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852805 | 11:40 |
| opendevreview | Nisha Agarwal proposed openstack/ironic master: Adds create_csr and add_https_certificate clean step https://review.opendev.org/c/openstack/ironic/+/804486 | 11:40 |
| dtantsur | ftarasenko, arne_wiebalck, I wonder if we need to make it straightforward to include delete_configuration as an automated clean step | 12:10 |
| dtantsur | I think somebody asked for literally the same thing last week (?) | 12:10 |
| iurygregory | ^ I remember something about this | 12:13 |
| dtantsur | ftarasenko: I think starting with Yoga (?), you can add priority to any clean step, making it automated clean step | 12:31 |
| dtantsur | ftarasenko: https://opendev.org/openstack/ironic/commit/71481ac483f0d1784cb10643da4c734abbea352f | 12:32 |
| dtantsur | maybe we should document that better | 12:32 |
| arne_wiebalck | dtantsur: we discussed making raid cleaning automatic at the time, but decided to follow what we do for h/w raid | 12:54 |
| arne_wiebalck | dtantsur: it is an automated clean step in our deployment (via our h/w manager) | 12:54 |
| arne_wiebalck | dtantsur: so, I am all for it, the only risk is that md devices *not* set up by ironic would be cleaned as well | 12:54 |
| arne_wiebalck | dtantsur: we have it here and it runs on *all* nodes (w/ and w/o s/w RAID) and so far it seemed ok | 12:55 |
| dtantsur | hmmm.. we probably need a way to remove only software RAID | 12:57 |
| opendevreview | Adam Rozman proposed openstack/ironic-python-agent master: add fallback logic for 'serial' root device hint https://review.opendev.org/c/openstack/ironic-python-agent/+/855866 | 13:04 |
| TheJulia | good morning | 13:20 |
| TheJulia | dtantsur: yeah, I think that was like 2 weeks ago, but it does seem like that has been asked a few times recently | 13:21 |
| dtantsur | morning TheJulia | 13:31 |
| dtantsur | arne_wiebalck: should we maybe at least delete software RAID if it was not created by us? or is it too hard? | 13:31 |
| TheJulia | I think that is a wonderful idea | 13:32 |
| TheJulia | tbh | 13:32 |
| kubajj | dtantsur: do you mean if it is not on the target raid config? | 13:33 |
| dtantsur | yep | 13:33 |
| kubajj | that should be quite straightforward | 13:34 |
| kubajj | we could just slightly modify the code around https://review.opendev.org/c/openstack/ironic-python-agent/+/852999/11/ironic_python_agent/hardware.py#2413 | 13:35 |
| ftarasenko | dtantsur: do I correctly understand that delete_configuration can be enabled per conductor as automated clean step? but in fact, I need create_configuration as well to recreate my own hardware raid. | 13:46 |
| dtantsur | mmm, okay, this is a bit more complicated then | 13:46 |
| dtantsur | I think we need a new clean step that will delete any software RAID, *regardless of the raid_interface used* | 13:46 |
| ftarasenko | iurygregory: tnx, but nvme clenaing is also disabled for raid member) | 13:46 |
| dtantsur | otherwise, delete_configuration will delete hardware RAID | 13:46 |
| TheJulia | Perhaps, just have the agent default to trying to clean it up in erase_devices | 13:46 |
| TheJulia | or erase_device_metadata | 13:47 |
| TheJulia | that is if... it is not defined | 13:47 |
| dtantsur | TheJulia: well... we decided that software RAID should behave like hardware RAID | 13:47 |
| dtantsur | could have been a mistake, but it's a part of the contract already | 13:47 |
| TheJulia | well.... it can't really... | 13:47 |
| dtantsur | a new opt-in clean step could be a solution | 13:47 |
| TheJulia | yeah | 13:47 |
| TheJulia | likely better, but also agent side could allow operators to get a fix if they are broken by a broken raid or something along those lines | 13:48 |
| ftarasenko | customer can create hardware raid as well) i think we might need step(s) that deletes hw/sw raid and creates right one. | 13:50 |
| ftarasenko | big problem for all hw/sw raid is that we have ata/nvme secure erase unavailable for them. | 13:51 |
| dtantsur | "customer can create hardware RAID" may be a tough problem to solve in the generic case | 13:51 |
| dtantsur | e.g. when Ironic does not have control over RAID (proprietary tools etc) | 13:51 |
| dtantsur | if Ironic does have control over it, delete_configuration together with the new clean step we discuss will do the job | 13:53 |
| TheJulia | it is almost like we generally need a "if is_raid_found() && is_there_target_raid_config(); then delete_all_raid_configuration(); fi | 13:53 |
| dtantsur | TheJulia: having the right driver is the problem | 13:53 |
| TheJulia | indeed | 13:53 |
| dtantsur | a topic for the PTG? ftarasenko will you take part? | 13:54 |
| TheJulia | our fearless PTL should send calendar invites out for the PTG | 13:55 |
| TheJulia | so we're all reminded! | 13:55 |
| dtantsur | I've heard we're about to have elections? | 13:55 |
| dtantsur | so we don't know the fearless PTL's name yet | 13:55 |
| TheJulia | indeed | 13:55 |
| TheJulia | Then again, we don't need to defer to the soon to be elected PTL to send out meeting invites | 13:56 |
| iurygregory | yeah, we don't need to wait for the election =) | 13:57 |
| opendevreview | Jacob Anders proposed openstack/sushy master: [WIP] Improve resiliency of eTag handling https://review.opendev.org/c/openstack/sushy/+/856123 | 13:57 |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/victoria: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852804 | 13:58 |
| iurygregory | if is a PTG topic please add to https://etherpad.opendev.org/p/ironic-antelope-ptg | 13:59 |
| iurygregory | =) | 13:59 |
| dtantsur | adding already | 13:59 |
| ftarasenko | dtantsur: I'll be happy to take part in PTG) | 13:59 |
| dtantsur | ftarasenko: I've added this topic to https://etherpad.opendev.org/p/ironic-antelope-ptg | 14:00 |
| ftarasenko | We'd also discuss this problem internally with team | 14:00 |
| iurygregory | ftarasenko, https://openinfra-ptg.eventbrite.com/ link to register for the PTG =) | 14:00 |
| ftarasenko | tnx | 14:00 |
| TheJulia | iurygregory: by chance did releasing our lesser active projects come up yesterday? | 14:01 |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/victoria: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852804 | 14:02 |
| iurygregory | TheJulia, nope, it was just a quick meeting (I think it was holiday in the US, so we didn't have a lot of ppl). I'm planing to release some of them this week | 14:03 |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/ussuri: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852805 | 14:04 |
| TheJulia | I looked at ngs, networking-baremetal, ironic-ui, and ipa-b last week and they all looked like there were no outstanding patches. | 14:05 |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/train: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852806 | 14:06 |
| janders | arne_wiebalck when you have a minute, would you be able to have a look at https://review.opendev.org/c/openstack/sushy/+/856123 (early version of eTag handling improvement patch)? This is trying to make things more resilient. We've seen some servers that don't tolerate being fed both strong and weak eTags; at times it seems better to skip eTags | 14:20 |
| janders | entirely - so I try to cater for both your use case and these use cases. | 14:20 |
| opendevreview | Merged openstack/tenks master: Add support for EL9 https://review.opendev.org/c/openstack/tenks/+/854550 | 14:28 |
| TheJulia | janders: left a comment, I have a concern about other HTTP exception errors that might be captured in that. Likely just some unit testing needed to make sure we don't inadvertently capture exceptions which should be returned to the higher level caller | 14:31 |
| dtantsur | before I dive into the oslo.db problems: is anyone already looking? | 14:35 |
| opendevreview | Julia Kreger proposed openstack/ironic master: CI: Changes to support Anaconda CI jobs https://review.opendev.org/c/openstack/ironic/+/849587 | 14:38 |
| opendevreview | Julia Kreger proposed openstack/ironic-tempest-plugin master: Tempest test for anaconda deploy https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/854031 | 14:40 |
| TheJulia | I am not at the moment, are we broken? | 14:40 |
| iurygregory | TheJulia, yeah, going to push a DNM just to be sure =) | 14:44 |
| dtantsur | TheJulia: will be with oslo.db 12.1.0, unless they revert it | 14:44 |
| TheJulia | joy | 14:44 |
| dtantsur | https://storage.bhs.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_f04/855153/1/check/cross-ironic-py38/f043efc/testr_results.html | 14:45 |
| iurygregory | dtantsur, there is a patch for it I saw yesterday | 14:45 |
| dtantsur | a possible workaround, you mean | 14:45 |
| iurygregory | I think i added to the ironic-week-prio | 14:45 |
| iurygregory | I didn't check the code tbh =) | 14:45 |
| dtantsur | yeah, I W-1 it since I'm not confident we should proceed with the direction | 14:45 |
| iurygregory | oh ack | 14:45 |
| dtantsur | TheJulia: among other things, limiting the set of fields to fetch is broken | 14:46 |
| dtantsur | aka your API performance improvement | 14:46 |
| TheJulia | oh for $*@!(#!*#@) | 14:56 |
| dtantsur | it's absolutely cryptic. it's worked around seemingly by the stephenfin's patch, but I don't have a single guess why | 14:57 |
| * TheJulia eats breakfast and works to finish current thought before pivoting | 14:57 | |
| arne_wiebalck | janders: heh, we have an eTag retry downstream patch as well :) | 15:11 |
| mallik | Hi dtantsur and JayF, I have addressed you review comments in the ilo patch https://review.opendev.org/c/openstack/ironic/+/853690. Could you please review it? | 15:11 |
| JayF | I'm literally just an hour into the first working day here in west-coast US since Friday :) | 15:12 |
| JayF | It's on my list, if I don't get to it today it'll be tomororow. | 15:12 |
| mallik | @JayF, ok thanks. | 15:12 |
| JayF | US bank holiday yesterday :) | 15:12 |
| opendevreview | Merged openstack/ironic-inspector master: Imported Translations from Zanata https://review.opendev.org/c/openstack/ironic-inspector/+/856010 | 15:13 |
| dtantsur | iurygregory: mm, this is too confusing, let's approve https://review.opendev.org/c/openstack/ironic/+/855969 for now | 15:17 |
| dtantsur | TheJulia: enabling autocommit somehow magically fixes the tests, dunno how it works.. | 15:20 |
| TheJulia | I saw the comment, that... makes... zero sense | 15:20 |
| TheJulia | peeling off the autocommit change, and running unit tests now to see what is going on | 15:21 |
| iurygregory | dtantsur, ok o/ | 15:25 |
| iurygregory | TheJulia, it's *magic* | 15:26 |
| iurygregory | TheJulia, I will just +2 the change for now | 15:26 |
| arne_wiebalck | dtantsur: is there a story for the planned inspector/ironic merger you could point me to? | 15:26 |
| dtantsur | arne_wiebalck: well.. I keep assuming it exists, but maybe only in my head :) lemme check | 15:27 |
| iurygregory | we do have info on etherpads | 15:27 |
| iurygregory | in the* | 15:27 |
| dtantsur | I'll write one if I find nothing | 15:27 |
| TheJulia | iurygregory: sounds good to me, trying to understand what is going on | 15:27 |
| arne_wiebalck | dtantsur: thanks! | 15:30 |
| TheJulia | wow, those are some weird unit test failures | 15:30 |
| arne_wiebalck | janders: I will be off for the rest of the week, but I added myself to the reviewers list and will try to check next week. | 15:33 |
| TheJulia | I *suspect* some of the tests breaking are where we've been inherently relying upon autocommit in the past | 15:35 |
| JayF | the *unit tests* have, or the unit tests expose *real code paths* that have been relying on it? | 15:36 |
| TheJulia | possibly both | 15:36 |
| TheJulia | the one in particular I'm trying to understand is test_get_node_list_Requested_fields_no_traits at the moment | 15:37 |
| TheJulia | to see if what we expect is actually occuring under the hood, or not | 15:37 |
| iurygregory | brb, lunch time | 15:37 |
| JayF | ack; I'm sorry I likely won't be able to help today. Finishing up some stuff as part of my offboarding from ft on Armada project (and I missed 3+ days last week sick) | 15:37 |
| TheJulia | eh no worries | 15:38 |
| dtantsur | arne_wiebalck, kubajj, started a new RFE https://storyboard.openstack.org/#!/story/2010275, will populate with details | 15:38 |
| dtantsur | arne_wiebalck, kubajj, you don't need to do EVERYTHING, any step is a step forward | 15:40 |
| kubajj | dtantsur: thanks, will look into it | 15:41 |
| dtantsur | maybe we need to discuss the final plan at the PTG, wdyt iurygregory, TheJulia? | 15:42 |
| TheJulia | ++ | 15:42 |
| TheJulia | so, I *think* sqlalchemy's behavior might have changed slightly and autocommit hides it | 15:42 |
| TheJulia | err, wrong field | 15:43 |
| TheJulia | doh | 15:43 |
| TheJulia | ugh, weird | 15:46 |
| TheJulia | okay... ugh | 15:49 |
| kubajj | TheJulia: dtantsur: I assume you have better stuff to do, but I am trying to figure out what could be wrong with this change. https://review.opendev.org/c/openstack/ironic-inspector/+/853304 Ironic Inspector grenade is timing out and as I was more invested in IPA than inspector, I am not really sure how it works. I already tried to recheck a couple days ago. | 15:51 |
| dtantsur | lemme look (finished the 1st draft of the RFE meanwhile) | 15:53 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic-inspector master: Switch the grenade job away from netboot https://review.opendev.org/c/openstack/ironic-inspector/+/856156 | 15:57 |
| dtantsur | kubajj: let's see if ^^^ helps | 15:58 |
| kubajj | dtantsur: thanks | 15:58 |
| TheJulia | I +2+A'ed that change. There was another inspector thing I noticed which is *likely* the same cause | 20:51 |
| TheJulia | I've workflowed the oslo.db workaround | 21:14 |
| TheJulia | I'm a bit stumped, and very worried we're going to have to do a a heavy lift with our dbapi | 21:14 |
| TheJulia | a very heavy lift | 21:14 |
| TheJulia | but fresh eyes in the morning... hopefully | 21:14 |
| opendevreview | Jay Faulkner proposed openstack/ironic bugfix/21.0: Redfish: Consider password part of the session cache https://review.opendev.org/c/openstack/ironic/+/856191 | 21:49 |
| opendevreview | Merged openstack/ironic-inspector master: Switch the grenade job away from netboot https://review.opendev.org/c/openstack/ironic-inspector/+/856156 | 21:51 |
| iurygregory | TheJulia, ack, hopefully we can find a fix that won't cause heavy impact in dbapi | 21:53 |
| iurygregory | everyone, fyi tomorrow is holiday in Brazil (independence day) so I won't be around | 21:54 |
| TheJulia | k | 21:55 |
| JayF | https://review.opendev.org/c/openstack/ironic-python-agent/+/854805 needs some core review love, software raid output poisioning fix for stable/yoga | 21:56 |
| iurygregory | it wasn't on my radar (not in the ironic-week-prio) .-. | 22:02 |
| iurygregory | reviewing now | 22:02 |
| JayF | Am I OK to add that myself, without consensus first? | 22:03 |
| JayF | I avoid tagging stuff like that unless I mention it in the meeting, but probably just being overcautious | 22:03 |
| JayF | also many of these have been more dormant than usual b/c I've been out sick -- this is the first time I've looked at stable backports in a week+ | 22:03 |
| iurygregory | JayF, it's totally fine for a contributor to add the hashtag to a patch they pushed (don't need to wait till the weekly meeting) | 22:14 |
| JayF | ack; I'll start doing that for my stable backports | 22:15 |
| iurygregory | tks o/ | 22:15 |
| JayF | o/ | 22:15 |
| iurygregory | see you Thursday Ironic, now it's time to enjoy the pre-holiday with some nice beer :D | 22:16 |
| janders | arne_wiebalck thank you! | 22:28 |
| opendevreview | Merged openstack/ironic master: Fix compatibility with oslo.db 12.1.0 https://review.opendev.org/c/openstack/ironic/+/855969 | 22:45 |
| TheJulia | okay, so it might not be *that* invasive | 23:04 |
| TheJulia | w/r/t sqlalchemy 2 | 23:04 |
| * TheJulia is still poking and prodding it | 23:04 | |
| TheJulia | so the tl;dr is without autocommit, we break kind of hard | 23:35 |
| TheJulia | because we're heavily reliant upon autoflush in some older areas of the code | 23:36 |
| TheJulia | locking also uses a now unsupported suntax | 23:36 |
| TheJulia | so it needs to be re-written | 23:36 |
| TheJulia | at least at teh db level | 23:36 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!