opendevreview | Vanou Ishii proposed openstack/ironic master: [iRMC] Handle IPMI incompatibility in iRMC S6 2.x https://review.opendev.org/c/openstack/ironic/+/865075 | 00:37 |
---|---|---|
TheJulia | kubajj: I'll look in my morning | 02:56 |
TheJulia | samuelkunkel[m]: so... I have an unlucky lab server that I upgraded from 5.10 to 6.10 and it failed horribly. I actually have a ticket with our lab folks to pull its power and try resetting it because... it is that unhappy :( Which is why I asked. :( | 02:58 |
samuelkunkel[m] | Yeah, I heard stuff like that before… (never happened to me… phew). I hope the debug output helps you then to bridge the gap so far. If you need anything else - let me know. | 05:31 |
samuelkunkel[m] | (Not sure if you saw my second link to the pastebin with the conductor running on debug) | 05:33 |
arne_wiebalck | Good morning, Ironic! | 07:12 |
rpittau | good morning ironic! o/ | 08:37 |
opendevreview | Riccardo Pittau proposed openstack/ironic-inspector bugfix/10.7: CI: Various required fixes https://review.opendev.org/c/openstack/ironic-inspector/+/860171 | 09:44 |
opendevreview | Riccardo Pittau proposed openstack/ironic-inspector bugfix/10.12: Fix CI https://review.opendev.org/c/openstack/ironic-inspector/+/870581 | 09:47 |
opendevreview | Riccardo Pittau proposed openstack/ironic-inspector bugfix/10.12: Fix CI https://review.opendev.org/c/openstack/ironic-inspector/+/870581 | 09:48 |
opendevreview | Riccardo Pittau proposed openstack/sushy stable/xena: Remove setuptools workaround https://review.opendev.org/c/openstack/sushy/+/870769 | 09:58 |
kubajj | Good morning arne_wiebalck, rpittau, and Ironic! How is everyone today? | 10:35 |
arne_wiebalck | hey kubajj and rpittau o/ | 11:10 |
iurygregory | good morning Ironic o/ | 11:31 |
opendevreview | Jakub Jelinek proposed openstack/ironic master: Reorganise Inventory Storage https://review.opendev.org/c/openstack/ironic/+/870799 | 12:42 |
kubajj | dtantsur, TheJulia: that should be the functions moved to inspect_utils ^ | 12:47 |
dtantsur | nice! | 12:47 |
* dtantsur needs to finish a few things first | 12:47 | |
kubajj | dtantsur: no worries | 12:48 |
opendevreview | Riccardo Pittau proposed openstack/ironic bugfix/18.1: Use xena and focal for releasenotes job https://review.opendev.org/c/openstack/ironic/+/870802 | 13:15 |
rpittau | well, mmm, I don't care about release notes in bugfix branch to be honest | 13:15 |
opendevreview | Riccardo Pittau proposed openstack/ironic bugfix/18.1: Remove releasenotes job https://review.opendev.org/c/openstack/ironic/+/870802 | 13:16 |
opendevreview | Riccardo Pittau proposed openstack/ironic-inspector bugfix/10.7: CI: Various required fixes https://review.opendev.org/c/openstack/ironic-inspector/+/860171 | 13:17 |
opendevreview | Riccardo Pittau proposed openstack/ironic-inspector bugfix/10.12: Fix CI https://review.opendev.org/c/openstack/ironic-inspector/+/870581 | 13:22 |
waleedm | hi guys, can you please review this patch and push merging it https://review.opendev.org/c/openstack/ironic-python-agent/+/566544 | 13:30 |
dtantsur | kubajj: provided feedback on both patches | 13:34 |
kubajj | dtantsur: thanks. For the comment about accessing services, do I just remove the service_type? | 13:35 |
dtantsur | kubajj: and the mentions of auth options | 13:36 |
dtantsur | check any file in ironic.conf that is not specific to a certain service | 13:37 |
kubajj | Ok, thanks | 13:37 |
kubajj | dtantsur: about the recursive call. I wonder how the test passed. | 13:59 |
opendevreview | Jakub Jelinek proposed openstack/ironic master: Create [inventory] https://review.opendev.org/c/openstack/ironic/+/870470 | 13:59 |
TheJulia | good morning | 14:25 |
kubajj | good morning TheJulia | 14:26 |
* TheJulia would like the migraine to go away | 14:35 | |
* dtantsur banishes TheJulia's migraine using his owl powers! | 14:40 | |
TheJulia | owl powers!? | 14:41 |
TheJulia | when did that happen? | 14:41 |
dtantsur | whispers: al-way-s | 14:41 |
TheJulia | O.o | 14:41 |
* dtantsur mysteriously flies away | 14:43 | |
* TheJulia blinks | 14:48 | |
rpittau | this is finally green again! \o/ https://review.opendev.org/c/openstack/ironic-inspector/+/860171 | 14:51 |
TheJulia | \o/ | 14:56 |
TheJulia | mraineri: o/ | 15:00 |
kubajj | dtantsur: about the unit tests for additions in isolation, would this be just the swift functions or what is already tested as well? | 15:07 |
opendevreview | Riccardo Pittau proposed openstack/ironic-inspector bugfix/10.7: CI: Various required fixes https://review.opendev.org/c/openstack/ironic-inspector/+/860171 | 15:07 |
dtantsur | kubajj: ideally, there should be tests for these new functions in isolation, while test_inspector should mock them or otherwise rely on these functions just working. | 15:08 |
kubajj | dtantsur: ok, I'll try to come up with something | 15:08 |
kubajj | also, I think the recursive call passed because the return value from it was mocked | 15:09 |
opendevreview | Riccardo Pittau proposed openstack/ironic-inspector bugfix/10.12: Fix CI https://review.opendev.org/c/openstack/ironic-inspector/+/870581 | 15:09 |
opendevreview | Riccardo Pittau proposed openstack/ironic bugfix/18.1: Pin tox to version lower than 4 https://review.opendev.org/c/openstack/ironic/+/870802 | 15:11 |
kubajj | dtantsur: is it possible to obtain inventory from redfish now? | 15:17 |
opendevreview | Jakub Jelinek proposed openstack/ironic master: WIP: Reorganise Inventory Storage https://review.opendev.org/c/openstack/ironic/+/870799 | 15:18 |
rpittau | JayF, TheJulia, dtantsur, https://review.opendev.org/c/openstack/ironic/+/868521 has already 2 + 2, can we give the +W ? | 15:19 |
dtantsur | kubajj: we have enough code in place to be able to collect most of it. but there is no trivial get_inventory_from_redfish function to get you there, more like different pieces to combine. | 15:19 |
TheJulia | rpittau: done | 15:42 |
rloo | if any cores around, zuul is happy with this (for now) xena backport, so would be good to get merged (cuz the wallaby backport merged): https://review.opendev.org/c/openstack/ironic/+/868025 | 16:03 |
rpittau | thanks TheJulia :) | 16:09 |
opendevreview | Riccardo Pittau proposed openstack/ironic-inspector bugfix/10.12: Fix CI https://review.opendev.org/c/openstack/ironic-inspector/+/870581 | 16:27 |
TheJulia | samuelkunkel[m]: out of curiosity, do you get the same behavior in bios boot mode? | 16:28 |
opendevreview | Riccardo Pittau proposed openstack/ironic-inspector bugfix/10.7: CI: Various required fixes https://review.opendev.org/c/openstack/ironic-inspector/+/860171 | 16:44 |
rpittau | good night! o/ | 16:46 |
dtantsur | JayF: hey, you available right now? re summit | 16:47 |
TheJulia | samuelkunkel[m]: also, I'm drawing a blank, what version of ironic/sushy are you running? | 16:50 |
JayF | dtantsur: In a meeting until the top of the hour | 16:53 |
JayF | dtantsur: happy to meet at ~1700 UTC, but I need time to grab breakfast and will be scarfing it while we chat lol | 16:53 |
dtantsur | JayF: so 17:15? | 16:57 |
JayF | sure | 16:57 |
JayF | or even 1710 | 16:57 |
dtantsur | even better | 16:57 |
JayF | arne_wiebalck going to be there too? | 16:57 |
dtantsur | yep, we've chatted in private | 16:57 |
arne_wiebalck | totally | 16:57 |
JayF | \o/ | 16:57 |
opendevreview | Merged openstack/ironic master: Fix selinux context of published image hardlink https://review.opendev.org/c/openstack/ironic/+/868521 | 16:58 |
JayF | dtantsur: I gotta get a sales pitch for metal3 from you, too :D | 16:58 |
JayF | dtantsur: I just found a downstream use case which seems tailored for it | 16:58 |
dtantsur | always! | 16:58 |
JayF | dtantsur: we're in | 17:05 |
dtantsur | okay, emails sent, I can disconnect now o/ | 18:08 |
samuelkunkel[m] | TheJulia: I did not try bios boot mode. We only use uefi. But I can modify that quickly tomorrow. | 18:28 |
samuelkunkel[m] | TheJulia: Versions: ironic (api/conductor (do you need the exact version?) is running zed, sushy is 4.3.2 | 18:29 |
TheJulia | samuelkunkel[m]: sigh, so you have the full etag handling improvements | 18:32 |
* TheJulia wonders if we need to improve logging | 18:32 | |
samuelkunkel[m] | And its confirmed that idrac versions 5.10 / 6.00 are not affected? | 18:34 |
TheJulia | yup | 18:35 |
TheJulia | I guess the thing that I'm wondering is if your fully falling back to no etags at all | 18:35 |
TheJulia | I think you might be, but I don't see the log entry | 18:36 |
opendevreview | Jakub Jelinek proposed openstack/ironic master: Reorganise Inventory Storage https://review.opendev.org/c/openstack/ironic/+/870799 | 18:37 |
opendevreview | Jakub Jelinek proposed openstack/ironic master: Create [inventory] https://review.opendev.org/c/openstack/ironic/+/870470 | 18:42 |
opendevreview | Jakub Jelinek proposed openstack/ironic master: Reorganise Inventory Storage https://review.opendev.org/c/openstack/ironic/+/870799 | 18:42 |
TheJulia | samuelkunkel[m]: I guess, actually, what we don't have is an explicit "we're about to try removing the etag completely" log entry https://review.opendev.org/c/openstack/sushy/+/856123/15/sushy/connector.py#338 | 18:43 |
TheJulia | so there is no retry there | 18:47 |
TheJulia | because it has to be a 412 for it to try | 18:47 |
* TheJulia wonders if we need to try to fallback with a 400 as well... | 18:48 | |
samuelkunkel[m] | And in this case we get a 400 were there is no handling for? And we „would“ treat a 400 same as a 412? | 18:48 |
TheJulia | well, the conudnrum is we don't know if this is etag related or not | 18:49 |
TheJulia | it is just... basically the only difference from a data point i have | 18:49 |
JayF | can we use sushy, with a small python sample script, to give samuelkunkel[m] something to test? | 18:49 |
TheJulia | but, if it works with bios mode... that would be another datapoint and would explain exactly what is going on a bit more verbosely | 18:49 |
samuelkunkel[m] | So the „quick“ patch to recognize the 400 same as 412 I can build in the code tomorrow and try | 18:49 |
samuelkunkel[m] | I am also playing around with sushy searching for a weird behaviour in requests library | 18:50 |
JayF | e.g. if you think that fallback would make it work, if we can get him (correct pronoun?) a sushy patch and a small test script | 18:50 |
TheJulia | Yeah, that is what I'm thinking, or even just change the if to true. | 18:50 |
JayF | might give us a quick turnaround | 18:50 |
TheJulia | samuelkunkel[m]: weird requests behavior? | 18:50 |
samuelkunkel[m] | Stay tuned, will submit something tomorrow ;) | 18:50 |
TheJulia | samuelkunkel[m]: ack, we *did* see some weirdness with etags | 18:51 |
TheJulia | and we did backport a fix ages ago that sought to disable compression | 18:51 |
TheJulia | well, ages ago as in back in 2021 | 18:51 |
samuelkunkel[m] | Its not related, basically we hit https://github.com/psf/requests/issues/3829 | 18:51 |
samuelkunkel[m] | Because we provide that exact env var | 18:51 |
TheJulia | sweet | 18:51 |
samuelkunkel[m] | Patch is pretty simple | 18:51 |
samuelkunkel[m] | But did not find the time to adjust the unittests | 18:52 |
TheJulia | samuelkunkel[m]: okay, lmk tomorrow | 18:52 |
samuelkunkel[m] | So will hop into that tomorrow | 18:52 |
TheJulia | Now to see if I can go to 6.0.0.0 instead of 6.10.0.0 :( | 18:52 |
TheJulia | much appreciated! | 18:52 |
opendevreview | Merged openstack/ironic stable/xena: Prevent pxe retry when agent token exists https://review.opendev.org/c/openstack/ironic/+/868025 | 18:52 |
opendevreview | Merged openstack/sushy stable/xena: Remove setuptools workaround https://review.opendev.org/c/openstack/sushy/+/870769 | 20:05 |
stevebaker[m] | good morning | 20:06 |
opendevreview | Merged openstack/ironic bugfix/18.1: Pin tox to version lower than 4 https://review.opendev.org/c/openstack/ironic/+/870802 | 20:20 |
opendevreview | Merged openstack/sushy stable/zed: Fix volume deletion on newer iDRACs https://review.opendev.org/c/openstack/sushy/+/869311 | 20:35 |
TheJulia | stevebaker[m]: looks like growpart blew up on debian :( | 21:05 |
TheJulia | re: michael still's email to the list | 21:06 |
stevebaker[m] | whoops | 21:06 |
stevebaker[m] | TheJulia: we should discuss this in #openstack-dib | 21:09 |
TheJulia | ++ | 21:10 |
sschmitt | I'm trying to get the functionality described in this spec https://github.com/openstack/ironic-specs/blob/master/specs/approved/physical-network-awareness.rst working with networking-generic-switch. I'm adding ports to ironic nodes with different physical_network attributes. Do these physnets need to be defined in the ml2 config file like other physnets for provider networks? We are trying to use this with tenant networks | 22:27 |
TheJulia | hmmmm | 22:32 |
TheJulia | this is a very good question | 22:32 |
TheJulia | so, networking-baremetal's code suggests yes based upon https://github.com/openstack/networking-baremetal/blob/30b6008c1730f0de7bb93bc7902d17731146097f/networking_baremetal/plugins/ml2/baremetal_mech.py#L533-L535 | 22:35 |
TheJulia | Although, I don't entirely remember all of the details there, but I was thinking a lot of that was to drive/inform scheduling upfront | 22:43 |
opendevreview | Merged openstack/ironic-inspector bugfix/10.7: CI: Various required fixes https://review.opendev.org/c/openstack/ironic-inspector/+/860171 | 22:47 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!