| TheJulia | stevebaker[m]: no, no it doesn't | 01:55 |
|---|---|---|
| rpittau | good morning ironic! Happy Friday! o/ | 06:19 |
| rpittau | dtantsur: about showing the entire output of lsblk https://review.opendev.org/c/openstack/ironic-python-agent/+/845785, I just realized that using the -O option changes the output as we're used to, for example the partitions and the md devices are shown as children in the tree structure | 06:44 |
| rpittau | you can see an example here https://a22c29270fdd295225bf-b3e91a29e09140a8f04c9c5cb84d28f3.ssl.cf2.rackcdn.com/845785/2/check/ironic-standalone-ipa-src/ecaee22/controller/logs/ironic-bm-logs/node-1_console.log.0 | 06:44 |
| rpittau | that means that we'll have to change substantially how we parse that otherwise all the children will be excluded by the list of devices | 06:45 |
| rpittau | just wondering if that amount of info is really worth the change or if we just want to add some more columns to the output with what can be of interest for us | 06:46 |
| dtantsur | good morning | 07:04 |
| rpittau | hey dtantsur :) | 07:04 |
| dtantsur | rpittau: the logging output definitely becomes somewhat unmanageable.. | 07:06 |
| rpittau | yep, it's a lot of stuff | 07:06 |
| dtantsur | I'm surprised it does the grouping without --tree provided | 07:07 |
| dtantsur | anyway, let's leave it for another day | 07:07 |
| dtantsur | is our CI still unwell? | 07:08 |
| rpittau | just grenade AFAICS, depends on new swiftclient release | 07:08 |
| rpittau | I mean, the fix is in swiftclient but it's unreleased | 07:08 |
| dtantsur | ah, okay | 07:10 |
| opendevreview | Riccardo Pittau proposed openstack/ironic-python-agent master: Remove oslo.serialization dependency https://review.opendev.org/c/openstack/ironic-python-agent/+/846278 | 07:35 |
| opendevreview | Riccardo Pittau proposed openstack/ironic-python-agent master: Remove oslo.serialization dependency https://review.opendev.org/c/openstack/ironic-python-agent/+/846278 | 07:37 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic-python-agent master: [WIP] Collect udev properties in the ramdisk logs https://review.opendev.org/c/openstack/ironic-python-agent/+/846280 | 07:48 |
| janders | good morning Ironic o/ | 08:06 |
| janders | been banging my head against a wall for a couple days now trying to figure out why verify steps stopped working in my lab. Ironically, it's the same lab I recorded a demo covering verify-steps! :) | 08:07 |
| janders | I wonder about this log message: https://paste.openstack.org/show/bazXLIf97neRpmgTSEY7/ | 08:07 |
| janders | I've got python-dracclient and sushy-oem-idrac installed | 08:07 |
| janders | the server I am testing against is an R640 with decent FW versions (and I used this firmware for the demo!) | 08:08 |
| janders | ajya I wonder what are your thoughts on that log message ( No extensions found for "manager" under namespace "sushy.resources.manager.oems" ). Would that break iDRAC verify-steps? | 08:09 |
| rpittau | hey janders :) | 08:10 |
| ajya | good morning | 08:16 |
| ajya | janders: yes, if it can't find extension, these verify steps will fail, because they're in extension | 08:16 |
| ajya | but why they can't find - any issues locating sushy-oem-idrac package? (python-dracclient necessary for wsman only, not here) | 08:17 |
| janders | ajya hmm, maybe it's installed in the wrong place or something | 08:20 |
| janders | what is the recommended way of installing it in a bifrost environment? | 08:20 |
| ajya | question for dtantsur ? :) | 08:21 |
| dtantsur | in /opt/stack/bifrost | 08:22 |
| janders | there we go | 08:22 |
| janders | not there | 08:22 |
| janders | must have gone to stack's homedir | 08:22 |
| dtantsur | I wonder why we don't install it when idrac is enabled.. maybe simply overlooked | 08:23 |
| dtantsur | hmm, we're supposed to | 08:24 |
| dtantsur | janders: do you have idrac enabled on installation or manually afterwards? | 08:24 |
| janders | dtantsur yes | 08:27 |
| janders | [DEFAULT]/enabled_hardware_types = ipmi,redfish,manual-management,idrac | 08:28 |
| janders | this one right? | 08:28 |
| * dtantsur likes "yes" in response to "or" questions :D | 08:29 | |
| janders | :D | 08:30 |
| janders | I did enable it manually afterwards | 08:30 |
| janders | how to enable it at install time? looking at docs, haven't spotted it yet | 08:31 |
| dtantsur | janders: bifrost-cli has a parameter, for playbooks it's enabled_hardware_types IIRC | 08:38 |
| opendevreview | Derek Higgins proposed openstack/ironic-python-agent master: Gather details about bond interfaces if present https://review.opendev.org/c/openstack/ironic-python-agent/+/846177 | 08:44 |
| opendevreview | Merged openstack/ironic master: Docs: replace nova cli calls with openstack https://review.opendev.org/c/openstack/ironic/+/846072 | 08:48 |
| janders | On another topic: can I please ask for reviews on https://review.opendev.org/c/openstack/devstack/+/845675 (IIUC it is needed to fix Grenade). Thank you! | 08:52 |
| rpittau | janders: not anymore, the real fix has landed in swiftclient | 08:52 |
| janders | (it's a Yoga backport of https://review.opendev.org/c/openstack/devstack/+/844539) | 08:52 |
| rpittau | we're waiting for a release | 08:52 |
| janders | rpittau thank you! | 08:53 |
| rpittau | janders: if you look my last comment there you'll see the link to the fix | 08:53 |
| janders | rpittau thank you! | 08:54 |
| janders | dtantsur ajya I re-ran bifrost-cli install with idrac parameter, problem solved | 08:55 |
| janders | it's verifying R640s now no problem | 08:55 |
| ajya | \o/ | 08:55 |
| janders | thank you! | 08:55 |
| janders | now I will add an older-gen box, have it fail and start looking at code improvements | 08:56 |
| janders | after reading bifrost-cli code I finally realised I was using help wrong | 08:59 |
| janders | I thought there isn't much there | 08:59 |
| janders | WRONG | 08:59 |
| janders | gotta specify the verb | 08:59 |
| janders | THEN it starts showing you the good stuff | 08:59 |
| janders | 🤦♂️ | 08:59 |
| dtantsur | :) | 09:00 |
| janders | I would have deployed it with idrac enabled but I couldn't figure out how | 09:00 |
| janders | so I figured I will just enable it after the fact | 09:00 |
| janders | but then I got bitten by packages being in the wrong place and nothing worked... | 09:00 |
| janders | ajya (and CC dtantsur) I reproduced iDRAC8 verification issue ( https://bugzilla.redhat.com/show_bug.cgi?id=2096944 ) on bifrost and thinking how to go about improving the code to resolve this. There are two distinct failure modes: 1) the server could do LC queue clear / iDRAC reset but has too old firmware and 2) it can't do LC queue clear / iDRAC | 09:39 |
| janders | reset cause it's one of the older generations. We could check for presence of URLs needed to perform these tasks and skip if they are absent for both cases, that would be my default thinking. | 09:39 |
| janders | Or - we could try be smarter and fail out with an error advising to upgrade firmware if the generation is right but the firmware is old - and skip on older generations. | 09:39 |
| janders | it's 8pm on a Friday here (and I've been shuffling boxes around the garage in prep for IKEA truck unloading) so not at my sharpest - but wanted to make some progress on this after being bogged down in lab issues for two days - so suggestions welcome! :) | 09:40 |
| janders | WDYT? | 09:41 |
| ajya | janders: yes, could check presence of URLs and if missing then iDRAC version. Yesterday I mentioned that maybe do something on sushy-oem-idrac, but maybe it's not necessary | 09:44 |
| janders | ajya ACK. I will now see what happens if I request iDRAC reset on the older node | 09:45 |
| ajya | maybe it's already enough to handle that error that sushy-oem-idrac raises in Ironic | 09:45 |
| janders | (not sure yet) | 09:45 |
| ajya | need to identify which iDRAC version started to support these in redfish, don't know off hand | 09:45 |
| opendevreview | Merged openstack/ironic-prometheus-exporter master: CI: Removing ironic job queue https://review.opendev.org/c/openstack/ironic-prometheus-exporter/+/842603 | 09:55 |
| TheJulia | good morning | 11:01 |
| janders | good morning TheJulia | 11:04 |
| dtantsur | morning TheJulia | 11:05 |
| janders | ajya https://paste.openstack.org/show/bvADdkMyURAKulpoMqbb/ iDRAC reset also fails out on the older iDRACs as we suspected | 11:13 |
| janders | now that I know what to look for I can start working on the fix | 11:13 |
| TheJulia | janders: wrt your email, it would be helpful if you'd open an issue in the osp jira project against the openstack-ironic component | 11:14 |
| janders | TheJulia ACK, on it | 11:17 |
| TheJulia | janders: thanks | 11:17 |
| TheJulia | fwiw, I'm going to request clarification/changes I think | 11:18 |
| janders | TheJulia 100% | 11:18 |
| TheJulia | but visibility in jira will help our team in planning and work tracking | 11:18 |
| ajya | janders: ack | 11:20 |
| janders | TheJulia do I need to ask FJ to open a BZ or is it OK just with the change up for review? | 11:20 |
| janders | TheJulia created issue, sent in an email | 11:25 |
| iurygregory | good morning Ironic | 11:25 |
| TheJulia | janders: much appreciated | 11:26 |
| TheJulia | janders: if they want it backported at all, a BZ will be needed | 11:27 |
| TheJulia | I'm shocked people are still using snmp for this sort of stuff :( | 11:29 |
| TheJulia | janders: it also needs a release note, and ideally doc changes | 11:35 |
| janders | TheJulia thank you. Would you be happy to put a comment in the change covering this? I will get them to address it. | 11:38 |
| TheJulia | janders: just did | 11:41 |
| janders | TheJulia TY! | 11:41 |
| TheJulia | it is actually a bunch of small things, and questioning why we're adding two new parameters with no other options. | 11:41 |
| TheJulia | I'm worried that this patch breaks upon upgrades as well | 11:42 |
| TheJulia | so requested unit testing appropriate as such | 11:42 |
| janders | All valid points. Thank you TheJulia, I appreciate your help! | 12:00 |
| opendevreview | Riccardo Pittau proposed openstack/virtualbmc master: Remove support for Python 3.6 and 3.7 https://review.opendev.org/c/openstack/virtualbmc/+/845915 | 12:41 |
| opendevreview | Riccardo Pittau proposed openstack/sushy-tools master: Remove support for Python 3.6 and 3.7 https://review.opendev.org/c/openstack/sushy-tools/+/845914 | 12:42 |
| janders | see you next week Ironic, have a great weekend everyone! o/ | 12:47 |
| rpittau | you too janders :) | 12:47 |
| sam_z | heyo, out of curiosity, are there any existing plans to add nova + cinder integration into bifrost? | 14:08 |
| sam_z | or would that be beyond the scope of what it's intended to do | 14:08 |
| dtantsur | sam_z: nova is definitely out of scope for bifrost | 14:11 |
| sam_z | makes sense. | 14:12 |
| sam_z | i suppose my use case would be more of a devstack thing | 14:12 |
| sam_z | i just uh. don't like devstack haha | 14:12 |
| dtantsur | heh, understandable :) | 14:12 |
| dtantsur | but yeah, bifrost is about standalone ironic | 14:13 |
| EmilienM | hellow - I'm wondering if there is some script i can re-use, which reproduces what devstack does to create fake bm in libvirt and import nodes into ironic. my use case: i have one beefy machine and I want to play with ironic without real hardware; i have openstack installed (all-in-one, with tripleo) | 14:18 |
| dtantsur | EmilienM: do you *need* to have tripleo? if not, you can try bifrost. | 14:19 |
| EmilienM | yeah I know I can use bifrost | 14:20 |
| EmilienM | yeah i'm using that cloud to run your fav k8s on top of it | 14:20 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic-python-agent master: Collect udev properties in the ramdisk logs https://review.opendev.org/c/openstack/ironic-python-agent/+/846280 | 14:20 |
| EmilienM | yeah i need ooo because we will eventually use that node for CI | 14:20 |
| EmilienM | think integration ci -- tripleo | ocp | 14:21 |
| dtantsur | EmilienM: you can look into https://opendev.org/openstack/ironic/src/branch/master/devstack/tools/ironic/scripts/create-node.sh and the accompanying python script, but they may need adjusting for non-devstack case | 14:21 |
| dtantsur | EmilienM: OR you can try reusing https://opendev.org/openstack/bifrost/src/branch/master/playbooks/roles/bifrost-create-vm-nodes but it will pull some bifrost bits with it | 14:21 |
| EmilienM | the thing with bifrost is that it installs ironic (which makes sense) but I don't need ironic, my tripleo already gives it to me | 14:22 |
| dtantsur | fair enough. maybe only the role I linked? although I'm not sure how easy it is, to use it without the rest of bifrost | 14:23 |
| EmilienM | kk, I'll look | 14:23 |
| EmilienM | I'll probably end up stealing the bits from there | 14:23 |
| EmilienM | i was hoping for an easy answer, you know, friday :) | 14:23 |
| sam_z | i was just in the opposite situation where i just finished adding nova + cinder to standalone ironic and for a bunch of stuff i did just kinda dig thru the bifrost ansible playbooks to try to piece together what to do lol | 14:24 |
| EmilienM | I'll call it hifrost | 14:26 |
| EmilienM | ;-) | 14:26 |
| rpittau | bye everyone, have a great weekend! o/ | 14:31 |
| dtantsur | hifrost lol | 14:39 |
| TheJulia | EmilienM: ... that could work ;) | 14:43 |
| dtantsur | have a nice weekend folks o/ | 15:00 |
| TheJulia | o/ | 15:00 |
| * TheJulia tries to avoid unpacking the pile of post-it notes in her bag | 15:01 | |
| NobodyCam | Good Morning Ironic folks | 17:22 |
| NobodyCam | T.G.I.F. | 17:22 |
| NobodyCam | have things changed... should A review with a +1 workflow merge? | 17:23 |
| iurygregory | hey NobodyCam o/ happy friday | 20:53 |
| iurygregory | I think you need at least one +2 and the +1 W | 20:54 |
| NobodyCam | hey hey iurygregory | 20:54 |
| NobodyCam | yea some thing strange with the review .. I issued a recheck | 20:54 |
| iurygregory | do you have the link for the review? | 20:55 |
| iurygregory | I can keep an eye on it | 20:55 |
| NobodyCam | https://review.opendev.org/c/openstack/ironic/+/840986 | 20:56 |
| NobodyCam | I just ran another recheck | 20:56 |
| iurygregory | soooo, there is one thing a bit weird about this patch | 20:57 |
| iurygregory | it's in chain with a change that is abandoned | 20:57 |
| NobodyCam | yea the abandoned first patch | 20:57 |
| NobodyCam | +++ | 20:58 |
| iurygregory | so maybe zuul will ignore the fact it has a +W, since the first change is not merged | 20:58 |
| iurygregory | I think it will need to rebase the change | 20:58 |
| NobodyCam | hummm | 20:59 |
| NobodyCam | ever used the rebase button? | 21:00 |
| NobodyCam | lol | 21:00 |
| NobodyCam | I can pull down and rebase if needed ... I'll let this last recheck complete | 21:00 |
| iurygregory | I've used before | 21:00 |
| NobodyCam | ++ | 21:00 |
| iurygregory | works most of the times =) | 21:01 |
| NobodyCam | heeheheh | 21:01 |
| iurygregory | I've subscribed to receive updates from it o/ | 21:01 |
| NobodyCam | Thank you | 21:03 |
| iurygregory | going to drop now, time to go to the gym, but latter I will check things o/ | 21:08 |
| TheJulia | hmm, iury disappeared | 21:24 |
| TheJulia | rebase button does tend to work :) | 21:24 |
| TheJulia | but will also post a bad rebase back :( | 21:25 |
| TheJulia | at least in newest gerrit | 21:25 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!