| rpittau | good morning ironic! o/ | 07:58 |
|---|---|---|
| dtantsur | The person behind the iBMC driver is, as we expected, just a freelancer and does not know the current situation. But they promised me to forward my question. | 08:47 |
| dtantsur | Good morning, happy Monday! | 08:47 |
| rpittau | Mondays will be Mondays :/ | 08:57 |
| *** Continuity_ is now known as Continuity | 08:58 | |
| qianbiao | Hi, Ironic. | 09:46 |
| qianbiao | Am I on right server of openstack-ironic? | 09:46 |
| opendevreview | Dmitry Tantsur proposed openstack/bifrost master: [PoC] Life without Inspector https://review.opendev.org/c/openstack/bifrost/+/887934 | 09:51 |
| dtantsur | qianbiao: you are indeed, hi again :) | 09:54 |
| JayF | Welcome :) | 09:54 |
| * JayF also says hello to the europe shift from London, UK | 09:54 | |
| dtantsur | JayF: so weird to see you so early, good morning :) | 09:54 |
| qianbiao | hi dtantsur | 09:56 |
| qianbiao | Hi JayF, i still remember you. :) | 09:57 |
| qianbiao | dtantsur i have forward your mail to iBMC team, if i got any further info, will let you know. | 09:57 |
| dtantsur | qianbiao: thanks much! | 09:58 |
| JayF | :D | 09:59 |
| qianbiao | feel guilty to bring inconvenience to ironic community. Due to the aanctions against Huawei, x86 business of Huawei has mostly been sold. | 10:00 |
| qianbiao | but there are still some agent they are using the iBMC driver as I know. | 10:01 |
| qianbiao | just they are running on pretty old version of openstack. | 10:01 |
| JayF | It's not your fault, don't feel bad. We're all working under a constantly changing set of rules :) | 10:03 |
| dtantsur | True that. | 10:04 |
| qianbiao | thanks | 10:06 |
| rpittau | hey JayF welcome to the old world :) | 11:09 |
| opendevreview | Dmitry Tantsur proposed openstack/bifrost master: [PoC] Life without Inspector https://review.opendev.org/c/openstack/bifrost/+/887934 | 11:14 |
| dtantsur | rpittau, JayF, should we merge https://review.opendev.org/c/openstack/ironic/+/887971/ or do we have other plans to fix the jobs? | 11:18 |
| rpittau | dtantsur: unfortauntely if that was proposed means that we don't have clues at the moment on how to fix that on jammy | 11:18 |
| rpittau | so going to approve that | 11:18 |
| dtantsur | +1 | 11:25 |
| frickler | are those devstack based jobs? devstack is going to drop focal support any moment | 11:37 |
| dtantsur | frickler: yep, they are... | 11:39 |
| dtantsur | so we may need to pin devstack as well, oh my | 11:39 |
| opendevreview | Merged openstack/bifrost master: Fix key-order[task] linter warnings https://review.opendev.org/c/openstack/bifrost/+/874854 | 11:43 |
| rpittau | yeah, we were really late with jammy :/ | 11:44 |
| opendevreview | Pierre Riteau proposed openstack/bifrost stable/zed: Skip unnecessary SDK get_machine calls https://review.opendev.org/c/openstack/bifrost/+/883116 | 11:52 |
| JayF | ironic-standalone job is devstack, not bifrost based? | 12:12 |
| JayF | I don't think I realized that | 12:12 |
| opendevreview | Merged openstack/ironic master: Move standalone jobs to focal https://review.opendev.org/c/openstack/ironic/+/887971 | 12:31 |
| minWi | hi folks, I'm trying to wrap my head around the sushy-emulator and the ability to map virtual media via libvirt... I'm trying to map an image file (either raw or qcow2) as a usb stick... but it always gets mapped as floppy... I've tried creating a bigger image file (like 10 MB image file, >1,44MB so maybe it is recognised as usb...) and also modified the SUSHY_EMULATOR_VMEDIA_DEVICES variable to remove floppy fr | 12:58 |
| minWi | om the mediatypes array... but still it is created as "<disk type='file' device='floppy'>". what am I doing wrong? Thanks! | 12:58 |
| JayF | I'm curious what kind of testing you're doing to need to do that with sushy-tools | 12:59 |
| opendevreview | Pierre Riteau proposed openstack/bifrost stable/yoga: Skip unnecessary SDK get_machine calls https://review.opendev.org/c/openstack/bifrost/+/883117 | 13:00 |
| opendevreview | Pierre Riteau proposed openstack/bifrost stable/xena: Skip unnecessary SDK get_machine calls https://review.opendev.org/c/openstack/bifrost/+/883118 | 13:00 |
| JayF | We discourage, in the strongest terms, use of sushy-tools, VirtualBMC, VirtualPDU in a production environment | 13:01 |
| TheJulia | good morning | 13:02 |
| JayF | o/ | 13:02 |
| minWi | yeah, this is not production environment don't worry :) basically I'm trying to deploy SUSE Linux Enterprise Micro in an automated fashion... and I'm trying to use combustion to do the initial setup... but how it works is SLE Micro expects a disk with a "combustion" label, so if the device is mapped as a floppy it doesn't work... I've tried this in a real baremetal server and it gets mapped as usb and it works.. | 13:04 |
| minWi | . but for testing purposes I prefer to use VMs for many reasons :) | 13:04 |
| JayF | aha, yeah that's good stuff, we just give the disclaimer | 13:04 |
| JayF | you'd be surprised the crazy ideas we hear :) | 13:04 |
| minWi | :) | 13:04 |
| JayF | now my next comment is, I didn't know sushy-tools did disk emulation at all, so I'm not sure how much I'll be helpful :D | 13:04 |
| TheJulia | I'd remove all logic for floppies at this point, tbh | 13:04 |
| rpittau | yes please | 13:04 |
| * TheJulia attempts to wake up | 13:07 | |
| * TheJulia caffinates | 13:08 | |
| TheJulia | https://github.com/openstack/sushy-tools/blob/master/sushy_tools/emulator/resources/systems/libvirtdriver.py#L1074 seems pretty specific, it comes down to the device it looks like | 13:21 |
| dtantsur | JayF: re standalone vs bifrost: we don't have tempest support in bifrost, nor do we have neutron there | 13:35 |
| JayF | What do folks think about a partially-baked Arista and Cisco (two drivers) support in NGS? | 13:43 |
| JayF | Caveats are primarily that I don't think they've been tested against a wide variety of switches, and I don't think they support all use cases | 13:43 |
| JayF | but obviously documenting those limitations would have to be part of getting it upstream | 13:43 |
| TheJulia | okay, I think I've reproduced the standalone job failure | 13:58 |
| TheJulia | at least, playing with the held CI node | 13:58 |
| rpittau | TheJulia: on jammy ? | 13:58 |
| TheJulia | yup | 13:59 |
| rpittau | oh yeah, of course | 13:59 |
| rpittau | cool | 13:59 |
| TheJulia | long story short, it is neutron not being ready, and ipxe changes gears, gets the same response, but not the response it wants to try the next overall capability so it just retries again until it fails | 13:59 |
| TheJulia | at least, that is my *high level* guess | 13:59 |
| TheJulia | and... dnsmasq disappeared on the job run causing it all to start off | 14:05 |
| TheJulia | standalone, of all of our jobs, has the most discrete scenarios tested | 14:11 |
| TheJulia | so all of the dhcp updates increase the odds of issues I guess | 14:12 |
| TheJulia | Jul 10 13:52:01 np0034614991 kernel: dnsmasq[402665]: segfault at c378 ip 00007f81458f047e sp 00007fff0db334a0 error 4 in libc.so.6[7f8145873000+195000] | 14:14 |
| TheJulia | Jul 10 14:14:03 np0034614991 kernel: dnsmasq[402863]: segfault at 7318 ip 00007f3940b9147e sp 00007ffc8df4f010 error 4 in libc.so.6[7f3940b14000+195000] | 14:14 |
| JayF | whoa that's kinda crazy | 14:14 |
| JayF | upstream bug reporting time I guess | 14:15 |
| TheJulia | so boom, there is our problem | 14:15 |
| opendevreview | Jay Faulkner proposed openstack/networking-generic-switch master: WIP: Add VLAN and Port whitelist https://review.opendev.org/c/openstack/networking-generic-switch/+/888047 | 14:16 |
| dtantsur | dnsmasq segfault, omg | 14:16 |
| rpittau | so dnsmasq issue on jammy, yay | 14:17 |
| rpittau | could be also why the other job was failing | 14:18 |
| dtantsur | frickler: I wonder if devstack could keep focal support until this segfault is solved | 14:18 |
| TheJulia | so looks like there *is* a fix for netbooting related crashes in 2.87 | 14:26 |
| TheJulia | ubuntu ships 2.86 | 14:26 |
| rpittau | \o/ | 14:27 |
| TheJulia | but we dont' seem to invoke the config to cause it | 14:27 |
| rpittau | let's install it from lunar? I'm semi-kidding | 14:27 |
| TheJulia | https://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/2022q3/016562.html | 14:35 |
| TheJulia | different location, but likely same casue. It is happening when changes are getting loaded, I'm wondering if it still attempts it even if disabled | 14:35 |
| TheJulia | so ubuntu appears to be shipping only 2.86 | 14:38 |
| TheJulia | https://packages.ubuntu.com/search?keywords=dnsmasq&searchon=names&suite=kinetic§ion=all | 14:38 |
| TheJulia | oh, lunar has latest | 14:38 |
| dtantsur | it's probably possible to install an isolated package from lunar.. | 14:41 |
| TheJulia | looks like a fast path to dependency hell | 14:53 |
| TheJulia | so... build dnsmasq?! | 14:56 |
| rpittau | it has one more deps though | 14:56 |
| opendevreview | Jay Faulkner proposed openstack/networking-generic-switch master: Fix delete_network_postcommit KeyError https://review.opendev.org/c/openstack/networking-generic-switch/+/888051 | 14:57 |
| rpittau | it's probably doable directly with the same deps for 2.86 | 14:57 |
| rpittau | orrrrr | 14:57 |
| rpittau | isntall the one from focal :D | 14:57 |
| rpittau | we used to do that for another pkg in the past | 14:57 |
| TheJulia | sudo dpkg -i dnsmasq-base_2.89-1_amd64.deb libnftables1_1.0.6-2_amd64.deb libgmp10_6.2.1+dfsg1-1.1ubuntu1_amd64.deb libjansson4_2.14-2_amd64.deb libnftnl11_1.2.4-2_amd64.deb dnsmasq-utils_2.89-1_amd64.deb | 14:59 |
| TheJulia | <-- finally works | 14:59 |
| rpittau | cool! | 14:59 |
| TheJulia | JayF: you going to run the meeting since you seem semi-around? | 14:59 |
| * rpittau waits before pulling the trigger | 15:00 | |
| TheJulia | Pull it! There is always the chairs command | 15:00 |
| rpittau | right! | 15:00 |
| rpittau | #startmeeting ironic | 15:00 |
| opendevmeet | Meeting started Mon Jul 10 15:00:46 2023 UTC and is due to finish in 60 minutes. The chair is rpittau. Information about MeetBot at http://wiki.debian.org/MeetBot. | 15:00 |
| opendevmeet | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 15:00 |
| opendevmeet | The meeting name has been set to 'ironic' | 15:00 |
| TheJulia | o/ | 15:00 |
| rpittau | Welcome to our weekly meeting! | 15:00 |
| rpittau | The meeting agenda can be found here: | 15:00 |
| rpittau | #link https://wiki.openstack.org/wiki/Meetings/Ironic#Agenda_for_next_meeting | 15:00 |
| rpittau | it's been a while I haven't done this, so be patient :) | 15:01 |
| rpittau | #topic Announcements/Reminder | 15:01 |
| rpittau | great news everyone! | 15:01 |
| rpittau | the next PTG has been announced | 15:01 |
| rpittau | The next PTG will take place virtually October 23-27, 2023! https://openinfra.dev/ptg/ | 15:01 |
| rpittau | let's add a note about that | 15:02 |
| rpittau | #note The next PTG will take place virtually October 23-27, 2023! https://openinfra.dev/ptg/ | 15:02 |
| dtantsur | o/ | 15:03 |
| rpittau | #note usual friendly reminder to review patches tagged #ironic-week-prio, and tag your patches for priority review | 15:03 |
| rpittau | one more reminder, I've added the bobcat timeline to the agenda | 15:04 |
| rpittau | #note we're just 6 weeks away from the Bobcat release | 15:04 |
| rpittau | nope, sorry, I can't count | 15:05 |
| rpittau | the next important date is August 24, which is 6 weeks away from the Bobcat relase | 15:05 |
| rpittau | anyone else has anything to announce / remind ? | 15:06 |
| rpittau | alright, moving on | 15:06 |
| rpittau | not huge attendance today :) | 15:06 |
| TheJulia | umm | 15:06 |
| TheJulia | if we're so far from release, why did we cut the client library release last week? | 15:07 |
| JayF | We just passed B-2, which is libraries released and mostly freezed | 15:07 |
| rpittau | yeah | 15:07 |
| JayF | B-2 is the TC+releases team documented cutoff for library releases | 15:07 |
| rpittau | next step is non-client libs freeze | 15:07 |
| JayF | I thought B-2 was all libs freeze, it was just clients? | 15:07 |
| * JayF is only like half here, is working downstream | 15:08 | |
| rpittau | client libs freese is august 31 | 15:08 |
| TheJulia | should that t-5 weeks per the schedule? | 15:08 |
| rpittau | the latest client release was not a cut | 15:08 |
| rpittau | it was a milestone | 15:08 |
| TheJulia | oh, why are we doing milestone cuts then? | 15:08 |
| rpittau | I think we're doing them since at least 2023.1 | 15:09 |
| TheJulia | okay, I thought we cut the stable branch on python-ironicclient last week, but it seems we did not | 15:09 |
| rpittau | for the cycle-with-intermediary clients and libs | 15:09 |
| TheJulia | we just did a regular releae | 15:10 |
| TheJulia | so there is still time \o/ | 15:10 |
| rpittau | yep | 15:10 |
| rpittau | :) | 15:10 |
| rpittau | glad to bring good news :) | 15:10 |
| rpittau | oky dokey, can we move on ? | 15:10 |
| TheJulia | sure | 15:11 |
| rpittau | #topic Review previous action items | 15:11 |
| rpittau | I don't see any action items from last meeting | 15:11 |
| rpittau | so skipping it | 15:11 |
| rpittau | to the fun part | 15:11 |
| rpittau | #topic Review Ironic CI Status | 15:11 |
| rpittau | how to summarize it? | 15:12 |
| rpittau | controlled chaos? | 15:12 |
| rpittau | mild enthropy vortex? | 15:12 |
| rpittau | we have 3 jobs pinned to focal, and the support to focal in devstack is going away soon(TM) | 15:12 |
| dtantsur | :D | 15:13 |
| TheJulia | so... well | 15:13 |
| rpittau | but it looks like a potential root cause has been found by TheJulia :) | 15:13 |
| TheJulia | Latest news that is on jammy, with the lunar dnsmasq packages, we still detonate | 15:13 |
| rpittau | \o/ | 15:13 |
| rpittau | do we want to try with the focal package? | 15:14 |
| TheJulia | I think, in order to provide an accurate bug report, at a minimum we'll need to pull the source and try latest, or we could punt on focal dnsmasq packages | 15:14 |
| rpittau | focal has version 2.80 | 15:14 |
| TheJulia | yeah, the race in reload issue was introduced in 2.86 it looks like | 15:15 |
| rpittau | great! | 15:15 |
| TheJulia | but we may be hitting something else | 15:15 |
| TheJulia | anyway, it crashes when we start to reload it | 15:15 |
| TheJulia | so... *shrugs* | 15:15 |
| rpittau | considering that the devstack sword is upon us, probably using focal packages while moving to jammy nodes for the time being could be viable ? | 15:16 |
| TheJulia | possibly | 15:18 |
| TheJulia | we've sort of long seen issues | 15:18 |
| TheJulia | and this is compounded by dnsmasq attempting httpboot as well | 15:18 |
| rpittau | alright, we can probably take the discussion back again outside of the meeting | 15:19 |
| rpittau | at least we have a possible cause of the issues | 15:19 |
| TheJulia | ++ | 15:19 |
| rpittau | ok, moving on | 15:19 |
| rpittau | #topic 2023.2 Workstream | 15:19 |
| rpittau | #link https://etherpad.opendev.org/p/IronicWorkstreams2023.2 | 15:19 |
| rpittau | any update worth sharing? | 15:20 |
| rpittau | we should probably review the firmware interface patches :/ | 15:20 |
| dtantsur | nothing new on the inspector front: waiting for reviews | 15:20 |
| rpittau | that's looking good :) | 15:21 |
| TheJulia | ++ | 15:21 |
| JayF | I would really, really like to get a good review pass done this week, but if you end up with anything short a +2 ping me directly | 15:21 |
| JayF | (reminder I'm in BST this week) | 15:21 |
| TheJulia | looks like we're making good progress on dpu stuff, but overall the CI issues have been really bad for us | 15:21 |
| rpittau | sad but true :( | 15:22 |
| rpittau | thanks for the updates in the etherpad | 15:22 |
| rpittau | moving on if we don't have other updates | 15:22 |
| rpittau | #topic Open Discussion | 15:23 |
| rpittau | anyone has anything for open discussion ? | 15:24 |
| JayF | If you're an Ironic contributor and have not gotten socks, let me know. | 15:24 |
| JayF | Same for stickers. | 15:24 |
| rpittau | JayF: I miss socks and stickers :D | 15:24 |
| JayF | email me at jay at gr-oss dot io and I'll get them out when I'm back home | 15:25 |
| JayF | that goes for ~anyone who is a contributor who wants some | 15:25 |
| rpittau | awesome! thank you! | 15:25 |
| JayF | (if you aren' | 15:25 |
| JayF | (if you aren't a contributor and want one, email me too, but contributors will get the first stock) | 15:25 |
| rpittau | any other topic to discuss ? | 15:26 |
| rpittau | alright, friendly reminder that I will be your host for the next meeting too :) | 15:27 |
| rpittau | thanks everyone! | 15:27 |
| rpittau | #endmeeting | 15:27 |
| opendevmeet | Meeting ended Mon Jul 10 15:27:20 2023 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 15:27 |
| opendevmeet | Minutes: https://meetings.opendev.org/meetings/ironic/2023/ironic.2023-07-10-15.00.html | 15:27 |
| opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/ironic/2023/ironic.2023-07-10-15.00.txt | 15:27 |
| opendevmeet | Log: https://meetings.opendev.org/meetings/ironic/2023/ironic.2023-07-10-15.00.log.html | 15:27 |
| TheJulia | now back to our dnsmasq topic! | 15:27 |
| * rpittau facepalm | 15:27 | |
| TheJulia | thoughts/ideas? | 15:27 |
| TheJulia | heh | 15:27 |
| TheJulia | I think forward looking wise, our path is most likely just to drop dnsmasq | 15:28 |
| TheJulia | honestly | 15:28 |
| TheJulia | but we're not entirely there... *and* I seem to be having issues on that test job | 15:28 |
| rpittau | the idea of submitting a bug is not bad, I did it for the other pkg I don't remember the name of and never got an answer! | 15:30 |
| rpittau | But at some point the issue was fixed | 15:30 |
| rpittau | and more seriously, pinning the pkg to focal would work at least for the time being, it's the same approach we followed for the other package | 15:30 |
| rpittau | it was the ovmf package! | 15:31 |
| TheJulia | well, to what bug tracker? there is not one for dnsmasq, it is the mailing list | 15:33 |
| TheJulia | ubuntu doesn't appear to have it configured for dnsmasq either | 15:33 |
| rpittau | mmm last time it was submitted to ubuntu | 15:34 |
| TheJulia | retrying on focal | 15:35 |
| TheJulia | we could file a bz in lp and tag dnsmasq/ubuntu directly | 15:37 |
| TheJulia | dunno how well that would work | 15:37 |
| rpittau | for ovmf we kept the old package for a while, it got fixed eventually.... | 15:38 |
| TheJulia | well, I've got 3 vms in clean wait | 15:40 |
| TheJulia | cleaning | 15:40 |
| TheJulia | so that is a good-ish sign | 15:40 |
| rpittau | fingers crossed | 15:42 |
| TheJulia | saving commands for later https://www.irccloud.com/pastebin/pSfKpU82/ | 15:43 |
| TheJulia | it has yet to crash in 8 minutes, with multiple updates, so far so good | 15:43 |
| TheJulia | kernel error log https://www.irccloud.com/pastebin/sTAR03K0/ | 15:45 |
| rpittau | with focal? | 15:46 |
| TheJulia | jammy host, focal packages seem to be good | 15:46 |
| rpittau | ok | 15:46 |
| TheJulia | 11 minutes and counting, multiple updates, still original process running | 15:47 |
| * TheJulia tells another node to go clean, again | 15:48 | |
| frickler | https://bugs.launchpad.net/ubuntu/+source/dnsmasq would be the ubuntu bug tracker | 15:50 |
| eandersson | You can always do something fun like this. | 15:51 |
| eandersson | https://github.com/openstack/zaqar/commit/c4502d1593ba9e417d98b018d9a74c5d3d8be920 | 15:51 |
| JayF | We need to get that bug fixed upstream of us | 15:52 |
| JayF | Man I wish I wasn't downstream this week | 15:52 |
| JayF | this is nerd sniping me hard but I don't have time for it this week :( | 15:52 |
| rpittau | eandersson: hey! long time no see! :) we actually did that for ovmf :D | 15:54 |
| eandersson | o/ | 15:54 |
| eandersson | I always feel dirty going that route.. but if it works it works I guess :D | 15:54 |
| rpittau | heh not a lot of choice here :/ | 15:55 |
| TheJulia | does it look like I'm missing anything ? https://bugs.launchpad.net/dnsmasq/+bug/2026757 | 16:09 |
| TheJulia | and tagged | 16:11 |
| TheJulia | challenge I see just adding it is if anything else gets installed, we could end up with poisoned package senario | 16:11 |
| TheJulia | I'd almost prefer to just directly download the packages and install them | 16:12 |
| rpittau | that looks good TheJulia | 16:13 |
| rpittau | TheJulia this is what we did for ovmf https://review.opendev.org/c/openstack/ironic/+/647687/5/devstack/lib/ironic | 16:14 |
| rpittau | I need to split now, see you tomorrow! o/ | 16:15 |
| TheJulia | dmesg | 16:31 |
| TheJulia | doh | 16:31 |
| opendevreview | Julia Kreger proposed openstack/ironic master: CI: Use focal dnsmasq https://review.opendev.org/c/openstack/ironic/+/888058 | 16:40 |
| TheJulia | eandersson: a little inspiration, just removing the file when done | 16:40 |
| TheJulia | thanks! | 16:40 |
| eandersson | Nice. I like that touch | 16:44 |
| eandersson | Cleaner :D | 16:45 |
| opendevreview | Verification of a change to openstack/ironic master failed: Fix the HTTP code of the BadRequest exception https://review.opendev.org/c/openstack/ironic/+/887271 | 16:55 |
| opendevreview | Julia Kreger proposed openstack/ironic master: Add wait step https://review.opendev.org/c/openstack/ironic/+/879061 | 17:18 |
| opendevreview | Julia Kreger proposed openstack/ironic master: DNM: Eventlet science https://review.opendev.org/c/openstack/ironic/+/887996 | 17:22 |
| opendevreview | Julia Kreger proposed openstack/ironic master: CI: Use focal dnsmasq https://review.opendev.org/c/openstack/ironic/+/888058 | 17:48 |
| opendevreview | Julia Kreger proposed openstack/ironic master: CI: Use focal dnsmasq https://review.opendev.org/c/openstack/ironic/+/888058 | 18:05 |
| opendevreview | Julia Kreger proposed openstack/ironic master: CI: Use focal dnsmasq https://review.opendev.org/c/openstack/ironic/+/888058 | 18:22 |
| opendevreview | Julia Kreger proposed openstack/ironic master: CI: Use focal dnsmasq https://review.opendev.org/c/openstack/ironic/+/888058 | 18:35 |
| TheJulia | so I think we're going to need to pull down the dpkgs down individually, adding it and trying to rely upon debian's package manager to do the right thing just doesn't seem to work, it can't seem to find the package in the configs | 19:25 |
| opendevreview | Julia Kreger proposed openstack/ironic master: CI: Use focal dnsmasq https://review.opendev.org/c/openstack/ironic/+/888058 | 19:57 |
| larsks | Is there a better place than here to ask questions about sushy? I have some questions about accessing actions on a redfish resource. | 19:58 |
| TheJulia | larsks: o/ best place, whats up? | 21:16 |
| larsks | TheJulia: hey. I want to access some redfish actions on our dell hardware, and I'm not seeing an api for doing that. There is support for *specific* actions (like the system_reset method on the system object), but I don't see something for arbitrary actions. | 21:17 |
| larsks | Is the best option calling sushy.Sushy._conn.post directly? | 21:17 |
| TheJulia | Eh, I'd consider calling a post directly a semi-workaround, but the reality is, afaik, we don't automatically expose internal methods. We build an appropriate wrapper as needed. System reset along the lines of reset to factory defaults? | 21:19 |
| larsks | No, I was pointing that out as an Action for which there was currently explicit support. | 21:19 |
| larsks | I am looking into using the SecureErase option on drives in a bunch of dell servers. | 21:19 |
| larsks | It looks like that means (1) get the action url from <object>.json['Actions'][<action_name>]['target'], and then (2) pass that to <Sushy>._conn(url, ...) with appropriate parameters. | 21:20 |
| larsks | Not sure if that action is actually supported on our idracs, though. Maybe a better option is just to enroll everything in ironic and then trigger a clean cycle... | 21:22 |
| TheJulia | yeah, that might be a good option :) | 21:25 |
| TheJulia | Given the whole nature of just in case things are not BMC managed/visible | 21:25 |
| TheJulia | so we did like a virtual media attachment action, and I remember correctly it took some non-zero effort to just getit to navigate/handle | 21:26 |
| TheJulia | fwiw, there is also a whole retirement logic flow for operators to help take care of wiping servers prior to them being removed | 21:28 |
| larsks | +1. I'm going to explore that option instead, I think; it looks like the SecureErase action, though advertised, isn't functional. | 21:29 |
| TheJulia | wheeeee :( | 21:30 |
| opendevreview | Julia Kreger proposed openstack/ironic master: Enable vendor interfaces to be called as steps https://review.opendev.org/c/openstack/ironic/+/879089 | 21:39 |
| TheJulia | larsks: For a minute, I was worried you were going to try to do something like the ilo "reset for handoff wipe" logic, which ... does a few truly "evil" things to the bmc. I think we actually flag the machine as in maintenance afterwards if it is invoked | 21:39 |
| TheJulia | rpittau: so, did we figure out a root cause on the snmp job failure? or was it just the dnsmasq curse | 21:44 |
| TheJulia | I've rechecked the dnsmasq fix, it appears solid and no errors in the latest job run | 21:47 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!