| opendevreview | Verification of a change to openstack/ironic master failed: Use patched dnsmasq from PPA https://review.opendev.org/c/openstack/ironic/+/933104 | 00:43 |
|---|---|---|
| cardoe | I tried a couple more times. | 01:22 |
| cardoe | https://review.opendev.org/q/label:Workflow=1%20and%20status:open%20and%20((project:%5Eopenstack/ironi.*)%20or%20(project:%5Eopenstack/sush.*)%20or%20(project:openstack/virtualpdu)%20or%20(project:openstack/virtualbmc)%20or%20(project:openstack/metalsmith)%20or%20(project:openstack/molteniron)%20or%20(project:openstack/networking-baremetal)%20or%20(project:openstack/networking-generic-switch)%20or%20(project:%5Eopenstack/py | 01:25 |
| cardoe | thon-ironi.*)%20or%20(project:openstack/tenks))%20and%20is:mergeable | 01:25 |
| opendevreview | Ghanshyam proposed openstack/ironic master: Remove default override for config options policy_file https://review.opendev.org/c/openstack/ironic/+/934569 | 05:41 |
| opendevreview | Ghanshyam proposed openstack/ironic-inspector master: Remove default override for oslo.policy config options https://review.opendev.org/c/openstack/ironic-inspector/+/934586 | 05:47 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic master: Inspection: account for disable_power_off https://review.opendev.org/c/openstack/ironic/+/934233 | 10:02 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic master: [WIP] Agent deploy: account for disable_power_off https://review.opendev.org/c/openstack/ironic/+/934637 | 11:06 |
| dtantsur | JayF: this ^^ should be the meat of the disable_power_off feature (module unit tests and real testing). I'd appreciate it if you could take a quick look from the hardening perspective. | 11:06 |
| dtantsur | https://opendev.org/openstack/tempest/commit/e60b31706240d989ba6d0244acbbb97154c47a13 broke us the hard way. is it the problem you're looking at cardoe? | 13:25 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic-tempest-plugin master: Use the correct exception when checking for HTTP 406 https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/934650 | 13:30 |
| dtantsur | This ^^^ should be minimum fix. But I think we need to replace cases of UnexpectedResponseCode with the base class + checking the status code. | 13:30 |
| JayF | Oh. That's a good find. And sort of my fault by proxy because I leaned on Adam to add real support to tempest instead of doing what apparently we were already doing | 14:03 |
| JayF | cid: you around to put a review on that tempest plug-in fix? 934650 -- if you are comfortable giving it a +2 I'll land it to unblock the gate | 14:06 |
| cid | Sure. Looking... | 14:06 |
| dtantsur | hold on, I think I missed one place | 14:07 |
| JayF | Get er done 💪💪💪 | 14:08 |
| JayF | By the way, today is a US holiday and after I chair the meeting like I promised Ricardo I would. I'm going to be done for the day | 14:08 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic-tempest-plugin master: Use the correct exception when checking for HTTP 406 https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/934650 | 14:08 |
| dtantsur | JayF, cid ^^ | 14:08 |
| dtantsur | I have an early stop today and will probably miss the meeting entirely | 14:10 |
| JayF | Well I might have a really easy job because we may not have quorum 😂 | 14:10 |
| JayF | I know that CID had some questions about context for inspector rules. So maybe not today, but if you have some time early this week I think like 15 minutes sync would be very very useful in getting that feature finished | 14:11 |
| JayF | I've tried to answer his questions but the reality is is. I'm not that up on inspection + I don't want to mislead | 14:11 |
| dtantsur | My calendar is open in the afternoon tomorrow. | 14:12 |
| dtantsur | (my afternoon, your morning) | 14:12 |
| cid | ++++++, also, dtantsur, we have updated the patch, worth taking a look too. | 14:12 |
| dtantsur | The taking a look will need to wait until tomorrow (ping me in the morning) | 14:12 |
| cid | Will do. | 14:13 |
| cid | My morning that is. | 14:13 |
| dtantsur | Sleeps inside tenacity are still impossible to stub out, aren't they? >_< | 14:22 |
| JayF | I'm not sure. Anything is possible with enough effort though... Not everything is worth it 😂 | 14:23 |
| dtantsur | True. The choice is between not testing a part of my code and creating unit tests that take a couple of seconds each. | 14:24 |
| dtantsur | okay, I've found a not-too-ugly way | 14:29 |
| opendevreview | Will Szumski proposed openstack/bifrost stable/2024.1: Fix leading and trailing whitespace in inspection_callback_url https://review.opendev.org/c/openstack/bifrost/+/934660 | 14:42 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic master: [WIP] Agent deploy: account for disable_power_off https://review.opendev.org/c/openstack/ironic/+/934637 | 14:56 |
| JayF | #startmeeting ironic | 15:00 |
| opendevmeet | Meeting started Mon Nov 11 15:00:17 2024 UTC and is due to finish in 60 minutes. The chair is JayF. 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 |
| dtantsur | o/ | 15:00 |
| kubajj | o/ | 15:00 |
| * dtantsur is here only for the beginning | 15:00 | |
| JayF | Welcome to the ironic meeting. I'm going to wait a little bit to see if we have quorum given it is a holiday in the US and a large part of Europe. | 15:00 |
| cid | o/ | 15:02 |
| JayF | If it's just the four of us, is there value in actually having a meeting? I don't think there's anything pressing. | 15:02 |
| dtantsur | I can only report the CI failure, but you've seen it | 15:03 |
| iurygregory | o/ | 15:03 |
| JayF | #info Ironic gate is currently broken until fix 934-650 is landed | 15:03 |
| JayF | #undo | 15:04 |
| opendevmeet | Removing item from minutes: #info Ironic gate is currently broken until fix 934-650 is landed | 15:04 |
| JayF | #Info ironic gate is currently broken until fix 934650 landed. | 15:04 |
| JayF | If Iury is here too, I'll run thru the agenda | 15:06 |
| JayF | #topic Announcements/Remidners | 15:06 |
| JayF | Standing reminder: review things #ironic-week-prio -> https://tinyurl.com/ironic-weekly-prio-dash | 15:06 |
| JayF | Please take note of the release schedule https://releases.openstack.org/epoxy/schedule.html | 15:07 |
| JayF | The PTG was October 21-25, notes are at https://etherpad.opendev.org/p/ironic-ptg-october-2024 | 15:07 |
| JayF | There are no discussion topics on the agenda. | 15:07 |
| JayF | #topic Bug Deputy updates | 15:07 |
| JayF | cid: I think that was you? | 15:07 |
| cid | Yep. | 15:07 |
| cid | Nothing much to report. I have reflected bug deputy updates on the meeting page; 10 new bugs, 2 RFEs. | 15:07 |
| cid | https://wiki.openstack.org/wiki/Meetings/Ironic | 15:07 |
| JayF | #info https://bugs.launchpad.net/ironic/+bug/2086672 - deprecate and retire ironic-lib | 15:08 |
| JayF | that's the first RFE, I think we approved it at the ptg | 15:08 |
| JayF | so if you wanna tag it rfe-approved I think that's a valid action | 15:08 |
| JayF | #info https://bugs.launchpad.net/networking-generic-switch/+bug/1653968 - support Vlan aware instances | 15:08 |
| JayF | I think 1653968 is maybe simple enough that bug is OK as is? | 15:10 |
| JayF | Giving a couple of minutes for others to weigh in; I suspect it's OK but will lay it over to next meeting if there aren't other opininos | 15:10 |
| dtantsur | I don't think anyone have cycles/energy/spoons to read a spec on it, realistically | 15:11 |
| dtantsur | So maybe we just trust the experts | 15:11 |
| * dtantsur needs to drop now o/ | 15:11 | |
| JayF | I would have the energy to read it | 15:12 |
| JayF | understand it; whole different story | 15:12 |
| JayF | lol | 15:12 |
| JayF | Going to move on. | 15:12 |
| JayF | Who is the next bug deputy? | 15:12 |
| JayF | I can probably take it? I haven't had a go as bug deputy in a while | 15:12 |
| JayF | #info JayF next bug deputy | 15:12 |
| JayF | Nothing else on the agenda. | 15:13 |
| JayF | #topic Open Discussion | 15:13 |
| JayF | Anything for open discussion before we close out the meeting? | 15:13 |
| cid | None from me | 15:13 |
| iurygregory | none from me | 15:14 |
| cid | Erm, maybe, when tagging an rfe-approved, do I remove the 'rfe' first | 15:14 |
| JayF | I leave both tags on, but I'm not certain that's correct tbh | 15:15 |
| JayF | real answer is it probably doesn't matter that much | 15:15 |
| kubajj | I have a question about the baremetal networking working group, is this going to be just an IRC meeting or something like zoom? | 15:15 |
| JayF | I do not know. Julia is running that right now and she's not here (US holiday). I'd suggest inquiring on the email thread on the list. | 15:15 |
| JayF | thanks everyone, I'm closing this out | 15:18 |
| JayF | o/ | 15:18 |
| JayF | #endmeeting | 15:18 |
| opendevmeet | Meeting ended Mon Nov 11 15:18:26 2024 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 15:18 |
| opendevmeet | Minutes: https://meetings.opendev.org/meetings/ironic/2024/ironic.2024-11-11-15.00.html | 15:18 |
| opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/ironic/2024/ironic.2024-11-11-15.00.txt | 15:18 |
| opendevmeet | Log: https://meetings.opendev.org/meetings/ironic/2024/ironic.2024-11-11-15.00.log.html | 15:18 |
| * JayF gone for the holiday o/ | 15:19 | |
| cid | \o | 15:20 |
| cardoe | So the vlan aware is just doing trunk ports so that the machines themselves tag their traffic? | 15:34 |
| cardoe | I’ve got someone looking at that. James was looking at it too. We need to tweak stuff on the nova side for cloud-init to work correctly. And we need to have child ports created in neutron. | 15:34 |
| cardoe | The child port pieces were very OVS specific so needed to visit that. | 15:35 |
| opendevreview | Takashi Kajinami proposed openstack/ironic master: Skip functional tests on pre-commit config update https://review.opendev.org/c/openstack/ironic/+/933194 | 15:56 |
| Pcmalih_ | Hi, while trying to boot PXE using ipmi driver. The controller sends BOOTP/DHCP Request (length 298) and baremetal machine send reply message (BOOTP/DHCP, Reply, length 308). these two message keep exchanging but nothing happens after multiple of these two messages. Image is not booting. I'm using bootloader and full disk image. | 16:31 |
| Pcmalih_ | How can I troubleshoot the PXE boot in this case? | 16:32 |
| shermanm | I'd recommend using tcpdump on the controller and double-checking that the dhcpreply is actually being received at the controller | 16:33 |
| Pcmalih_ | Controleller ethernet interface (not IPMI) is able to receive the reply. That is same interface set as ironic_dnsmasq_interface as well. | 16:35 |
| shermanm | the other sanity check I'd do is that the correct NIC on the node is making the request, e.g. the mac address matches the ironic port | 16:35 |
| Pcmalih_ | @shermanm: Using tcpdump. Just another sanity check. IMPI interface is not used to boot image? | 16:36 |
| shermanm | not if you're pxe booting it, ipmi is being used for power on / power off / set boot order commands, but the ironic agent ramdisk is pxe booting over one of the in-band network interfaces | 16:38 |
| Pcmalih_ | One more question is the NIC port MAC need to be the MAC of ethernet end connected to Compute or Controller? | 16:38 |
| Pcmalih_ | I added all the NICs MAC as port in controller. | 16:40 |
| Pcmalih_ | and creating ironic port with MAC, also linking it with nova using VIF ID. | 16:41 |
| Pcmalih_ | #shermanm | 16:41 |
| shermanm | I'm not totally sure what you're asking there, I've generally been using ironic with the neutron network driver, in which case the baremetal ports just correspond to NICs on each baremetal node | 16:42 |
| Pcmalih_ | Trying to tag participant, how do folks tag in chat here? | 16:42 |
| cardoe | Username | 16:43 |
| Pcmalih_ | Username shermanm I'm using flat networking, created the subnet in neutron same as of the network address where controller and compute are. | 16:44 |
| cardoe | So DHCP is a 4 packet exchange. Discover, Offer, Request, Ack. The client (PXE) is step 1 and 3. While the server is 2 and 4. Step 1 and 2 is a broadcast. | 16:44 |
| cardoe | Most often PXE doesn’t send step 3 because it decides it’s missing some info regarding next boot. Or can’t access the system. | 16:46 |
| Pcmalih_ | cardoe only BOOTP/DHCP Request and BOOTP/DHCP Reply are exchanged, no ACK message is received over controller. | 16:46 |
| shermanm | I'd recommend taking a look at this diagram for more context: https://docs.openstack.org/ironic/latest/user/architecture.html#direct-deploy-example | 16:46 |
| cardoe | So you’re looking at something weird. Cause you’re saying packets 2 and 3. | 16:48 |
| cardoe | It’s Discover, Offer, Request, Ack. | 16:49 |
| cardoe | Toss a tcpdump in there and pay attention to the next server and boot file fields. | 16:49 |
| Pcmalih_ | <cardoe> My tcpdump sniffing show only these two types of messages. | 16:49 |
| cardoe | Then you’re not listening to something. Cause you should be seeing a discover. | 16:50 |
| shermanm | or something's blocking broadcast traffic maybe? | 16:50 |
| cardoe | And if you see a request, look at your server and see why it’s rejecting it. | 16:51 |
| Pcmalih_ | Its direct connection between controller and compute. @ | 16:52 |
| Pcmalih_ | How to look at server (controller), I assume tcpdump over the NIC (controller side) is only way to trubleshoot? | 16:53 |
| cardoe | Essentially if you’re seeing a request, you should be seeing an Ack or Nack. | 16:53 |
| cardoe | You can capture on the switch. | 16:53 |
| Pcmalih_ | <cardoe> now i don't have switch but a direct ethernet connection. Still tcpdump logs are showing same messages. No ACK or NACK. | 16:55 |
| Pcmalih_ | Just request and reply. | 16:55 |
| cardoe | Listen on any interface. Could you be routing the packet out another interface. | 16:59 |
| cardoe | You should be seeing a discover at a minimum. | 16:59 |
| Pcmalih_ | tcpdump -i <interface> port 67 or port 68 or port 69 -e -n | 17:04 |
| Pcmalih_ | <cardoe>This command should all messages e.g. ACK, DISCOVER? | 17:05 |
| opendevreview | Doug Goldstein proposed openstack/ironic master: Handle Power On/Off for child node cases https://review.opendev.org/c/openstack/ironic/+/896570 | 17:09 |
| cardoe | You don't need port 69. But I would do `-i any` | 17:09 |
| cardoe | TheJulia: I touched https://review.opendev.org/c/openstack/ironic/+/896570 to adjust the locking for power_off. JayF and cid I dropped your reviews. | 17:10 |
| opendevreview | Doug Goldstein proposed openstack/sushy master: switch to pre-commit https://review.opendev.org/c/openstack/sushy/+/929704 | 17:13 |
| JayF | cardoe: I'm not here today, but -1, and I think you should restore patchset 10 | 18:47 |
| JayF | I was +2 to patchset 10 and think you misread around your concerns | 18:47 |
| cardoe | Okay. I’m on mobile at this point so not sure if I can do it from the web UI. I thought Julia said yes the locking needed to change but she didn’t have the time. | 19:04 |
| opendevreview | lkk proposed openstack/ironic-tempest-plugin master: list detailed ports by node https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/712831 | 23:17 |
| opendevreview | likui proposed openstack/ironic master: update doc and drop deprecated plugin https://review.opendev.org/c/openstack/ironic/+/898157 | 23:21 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!