| rpioso | :-) | 00:12 |
|---|---|---|
| *** pmannidi|Gone is now known as pmannidi | 00:56 | |
| arne_wiebalck | Good morning, Ironic! | 06:31 |
| rpittau | good morning ironic! o/ | 06:44 |
| dtantsur | morning folks | 07:32 |
| janders | good morning arne_wiebalck rpittau dtantsur and Ironic o/ | 07:45 |
| rpittau | hey janders dtantsur :) | 07:45 |
| dtantsur | the stable CI is not feeling well, is it | 07:46 |
| rpittau | I saw failures on various master CI jobs as well, standalone, tempest | 08:19 |
| rpittau | not sure also if anyone had a chance to look at grenade in inspector | 08:19 |
| dtantsur | definitely not me | 08:34 |
| opendevreview | Riccardo Pittau proposed openstack/ironic-inspector master: [WIP] Fix grenade job https://review.opendev.org/c/openstack/ironic-inspector/+/840588 | 09:02 |
| opendevreview | Riccardo Pittau proposed openstack/ironic-inspector master: [WIP] Fix grenade job https://review.opendev.org/c/openstack/ironic-inspector/+/840588 | 09:03 |
| rpittau | looking for a quick easy review -> https://review.opendev.org/c/openstack/sushy-tools/+/838674 | 09:27 |
| rpittau | iurygregory: can you please remind me what the plan is for the bugfix releases? Can we start already with IPA or we want to wait for the multipath handling patches to merge? | 09:34 |
| opendevreview | Merged openstack/sushy-tools master: Use python Zed tests https://review.opendev.org/c/openstack/sushy-tools/+/838674 | 11:02 |
| arne_wiebalck | hey janders rpittau dtantsur o/ | 11:13 |
| opendevreview | Jacob Anders proposed openstack/sushy stable/xena: Attempt using Redfish settings resource to get URI https://review.opendev.org/c/openstack/sushy/+/839843 | 11:30 |
| opendevreview | Jacob Anders proposed openstack/sushy stable/xena: Follow up to Attempt using Redfish settings resource to get URI https://review.opendev.org/c/openstack/sushy/+/839844 | 11:30 |
| janders | iurygregory thank you for your review and pointers in https://review.opendev.org/c/openstack/sushy/+/839843, once CI is done ^ should be ready for review (plus https://review.opendev.org/c/openstack/sushy/+/839845 which is already past CI) | 11:33 |
| opendevreview | Jacob Anders proposed openstack/sushy stable/wallaby: Use Etag in precondition header when setting boot options https://review.opendev.org/c/openstack/sushy/+/840652 | 11:52 |
| opendevreview | Jacob Anders proposed openstack/sushy stable/wallaby: Attempt using Redfish settings resource to get URI https://review.opendev.org/c/openstack/sushy/+/840653 | 11:53 |
| opendevreview | Jacob Anders proposed openstack/sushy stable/wallaby: Follow up to Attempt using Redfish settings resource to get URI https://review.opendev.org/c/openstack/sushy/+/840654 | 11:55 |
| opendevreview | Jacob Anders proposed openstack/sushy stable/wallaby: Re-trying InsertMedia call with TransferProtocolType if required https://review.opendev.org/c/openstack/sushy/+/840655 | 12:00 |
| janders | ^ that should be the lot for fixing Nokia bugs in Wallaby | 12:01 |
| rpittau | I fixed grenade and broke something else in inspector :D | 12:27 |
| rpittau | another quick and painless review if anyone has a minute https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/840508 | 12:40 |
| rpittau | and I just found a mistake.... | 12:40 |
| rpittau | lol | 12:40 |
| opendevreview | Riccardo Pittau proposed openstack/ironic-tempest-plugin master: Add yoga tests, stop testing victoria https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/840508 | 12:42 |
| rpittau | now should be painless and quick | 12:42 |
| iurygregory | good morning Ironic | 12:51 |
| iurygregory | rpittau, hey! | 12:51 |
| rpittau | hey iurygregory :) | 12:51 |
| iurygregory | I was planning to do today the releases | 12:51 |
| iurygregory | we can ofc wait if the patch is in a good shape (otherwise we can just backport) | 12:51 |
| rpittau | I vote for doing the releases and then backport :) | 12:52 |
| iurygregory | ok o/ | 12:53 |
| iurygregory | pushing patches in a few, I just got back from the doctor | 12:54 |
| TheJulia | brraaaaaiinnnssss | 13:23 |
| rpittau | TheJulia: I can offer an alternative https://www.cuisineaz.com/recettes/bloody-brain-le-cocktail-cerveau-99381.aspx | 13:29 |
| TheJulia | Neat, although really not a fan of grenadine | 13:32 |
| opendevreview | Riccardo Pittau proposed openstack/ironic-inspector master: Fix CI https://review.opendev.org/c/openstack/ironic-inspector/+/840588 | 13:33 |
| TheJulia | That kind of reminds me, a little bit of... Tuaca Alexander (think, a Brandy Alexander, but made with Tuaca) | 13:33 |
| rpittau | oh gosh, that sounds dangerously delicious | 13:34 |
| TheJulia | it is | 13:34 |
| TheJulia | so, multipath! | 13:34 |
| rpittau | yep | 13:35 |
| TheJulia | where are we at, where do we need to go, is there anything I can do to help? or should I just let you guys handle it? | 13:35 |
| * TheJulia wonders if she can silence matrix on ashinclouds[m]'s phone :) | 13:35 | |
| * TheJulia is talking about herself too | 13:35 | |
| rpittau | we need to adjust https://review.opendev.org/c/openstack/ironic-python-agent/+/837039 mainly making it more generic (plus fix all the tests) | 13:36 |
| rpittau | dtantsur already adjusted his patches and backported them | 13:37 |
| TheJulia | okay, and combined things apparently work | 13:37 |
| rpittau | correct | 13:37 |
| rpittau | we need to fix _load_multipath_modules to add a non-redhat way to deal with multipath config | 13:38 |
| TheJulia | lets see if a default is dropped in | 13:38 |
| TheJulia | because I didn't have to do anything besides run commands | 13:38 |
| TheJulia | Interesting! I have no config files | 13:39 |
| rpittau | lol | 13:40 |
| TheJulia | soo.... maybe my original approach was right for debian | 13:40 |
| rpittau | multipath.conf is definitely needed by mpathconf | 13:40 |
| * TheJulia will need to dig through debian's docs | 13:40 | |
| rpittau | yeah | 13:40 |
| TheJulia | and whatever is in debian is likely identical in ubuntu | 13:41 |
| rpittau | I really hope so :) | 13:41 |
| TheJulia | Although, kind of hearing of the downstream trials and tribulations is why I had always hoped we would never have to do it | 13:45 |
| TheJulia | trials and tribbleations | 13:45 |
| opendevreview | Riccardo Pittau proposed openstack/ironic-inspector master: Fix CI https://review.opendev.org/c/openstack/ironic-inspector/+/840588 | 13:45 |
| TheJulia | https://external-content.duckduckgo.com/iu/?u=https%3A%2F%2Ffsmedia.imgix.net%2F85%2F8f%2F28%2Fdf%2F4293%2F4740%2Fb003%2F58a281bb6c77%2Fkirk-has-had-it-up-to-here-with-these-damn-tribbles.png%3Fdpr%3D2%26auto%3Dformat%252Ccompress%26w%3D650&f=1&nofb=1 | 13:46 |
| * TheJulia assumes the cats would be in heaven with a pile of tribbles | 13:46 | |
| rpittau | :D | 13:47 |
| dtantsur | TheJulia: hi! do you remember why we didn't do https://review.opendev.org/c/openstack/ironic-python-agent/+/840499 from the start? | 13:47 |
| TheJulia | we needed to also be able to match against partuuid | 13:49 |
| TheJulia | I think | 13:49 |
| TheJulia | findfs might also be doing something funky | 13:50 |
| TheJulia | eh, it supports it | 13:51 |
| dtantsur | yup | 13:52 |
| TheJulia | dtantsur: also possible it wasn't around in the util-linux version in tinycore from ages ago | 13:52 |
| dtantsur | possibly, yeah | 13:52 |
| TheJulia | the man page is from 2014, also noting a rewrite from the original for util-linux | 13:53 |
| TheJulia | so there could have been some ships passing in the night sort of stuff there | 13:54 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic-python-agent stable/yoga: Do not try to guess EFI partition path by its number https://review.opendev.org/c/openstack/ironic-python-agent/+/840670 | 13:57 |
| TheJulia | NobodyCam: wowzers https://review.opendev.org/c/openstack/ironic/+/839995/1/ironic/drivers/modules/snmp.py | 13:57 |
| TheJulia | pep8 job has opinions... many opinions | 13:57 |
| dtantsur | omg, that's plenty of opinions | 13:59 |
| opendevreview | Iury Gregory Melo Ferreira proposed openstack/ironic master: Ironic Release 20.2 https://review.opendev.org/c/openstack/ironic/+/840671 | 14:00 |
| dtantsur | wow, no API changes sine Xena | 14:00 |
| TheJulia | It has happened before | 14:01 |
| TheJulia | For a long time it was like "all the api changes/additions in one release" and then "all the backend/internal stuffs the next release" | 14:01 |
| dtantsur | heh fair | 14:02 |
| dtantsur | next time we should make some breaking changes :) like removing netboot | 14:02 |
| TheJulia | ++ | 14:02 |
| TheJulia | some projects are asking if they can nuke legacy policy support early | 14:03 |
| dtantsur | what is the target nuke release? | 14:03 |
| TheJulia | AA | 14:03 |
| * dtantsur thought TheJulia is screaming at first :D | 14:03 | |
| TheJulia | or now... I guess that would be 2023.0? | 14:03 |
| dtantsur | OpenStack Aaaaaaaaaaaaa!!1 | 14:04 |
| TheJulia | lol | 14:04 |
| rpittau | can we do 2023.tic and 2023.toc instead ? | 14:06 |
| TheJulia | I haven't read the latest emails | 14:07 |
| TheJulia | and I'm not updating the code for version handling | 14:07 |
| TheJulia | just nope | 14:07 |
| TheJulia | nope nope nope | 14:07 |
| TheJulia | all the nopes | 14:07 |
| dtantsur | :D | 14:07 |
| rpittau | :) | 14:07 |
| TheJulia | http://memecrunch.com/meme/BGOEI/nope/image.png | 14:08 |
| dtantsur | we used to have year-based versions long ago or am I confusing something? | 14:08 |
| TheJulia | yes | 14:08 |
| TheJulia | back in the days of 2014 | 14:08 |
| TheJulia | so... like... omfg 8 years ago | 14:08 |
| dtantsur | *nod* | 14:08 |
| TheJulia | wow, so I actually joined HPE ~..... just over 8 years ago if memory serves | 14:09 |
| rpittau | I think 8 years ago is when I started playing with Openstack, Paris Summit was that year | 14:10 |
| * dtantsur celebrated 8 years in RH in February | 14:10 | |
| dtantsur | oh yeah, Paris summit. wild parties :) | 14:10 |
| TheJulia | I didn't get travel approval until like 7 days before the summit | 14:12 |
| TheJulia | "monty needs you to be there".... | 14:12 |
| dtantsur | I had to use the foundation funding.. | 14:13 |
| dtantsur | on an unrelated note, I already regret moving so much code around in IPA :D | 14:14 |
| TheJulia | backport pain? | 14:14 |
| dtantsur | yeeeeah | 14:15 |
| dtantsur | nothing applies cleanly | 14:15 |
| TheJulia | Yeah | 14:15 |
| iurygregory | HPE Proliant Merge Failed. This change or one of its cross-repo dependencies was unable to be automatically merged with the current state of its repository. Please rebase the change and upload a new patchset. | 14:17 |
| iurygregory | funny... | 14:17 |
| iurygregory | stendulker, fyi =) (I don't remember the other person from HPE to ask about the CI) | 14:18 |
| stendulker | iurygregory: no problem, I will send it to the right group. | 14:18 |
| iurygregory | Tks! | 14:19 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic-python-agent stable/wallaby: [WIP] Do not try to guess EFI partition path by its number https://review.opendev.org/c/openstack/ironic-python-agent/+/840170 | 14:23 |
| * dtantsur wishes git could detect relocated code | 14:26 | |
| TheJulia | that would be epic if it oculd | 14:28 |
| dtantsur | the code I'm trying to backport changed twice. it was significantly updated in xena, then moved around in yoga. | 14:28 |
| TheJulia | yeouch | 14:35 |
| TheJulia | we hit some of that with bootloader stuffs going back to train once... I think we backported basically every change as a result for our own long term sanity | 14:35 |
| TheJulia | or nearly every change... | 14:35 |
| * TheJulia would kindly ask current cold/sinus infection to go away | 14:40 | |
| *** diablo_rojo__ is now known as diablo_rojo | 14:43 | |
| opendevreview | Merged openstack/ironic master: Fix v6 CI job - Return it to normal non-voting status https://review.opendev.org/c/openstack/ironic/+/839086 | 14:46 |
| opendevreview | Merged openstack/networking-generic-switch master: Trivial: Remove legacy devstack CI call https://review.opendev.org/c/openstack/networking-generic-switch/+/840513 | 14:47 |
| opendevreview | Merged openstack/networking-generic-switch stable/xena: CI: Fix Multinode ssh key/hostname for unit tests https://review.opendev.org/c/openstack/networking-generic-switch/+/839920 | 14:47 |
| opendevreview | Riccardo Pittau proposed openstack/virtualbmc master: Use zed jobs https://review.opendev.org/c/openstack/virtualbmc/+/840501 | 15:08 |
| opendevreview | Merged openstack/ironic-python-agent master: Do not try to guess EFI partition path by its number https://review.opendev.org/c/openstack/ironic-python-agent/+/840453 | 15:17 |
| opendevreview | Harald Jensås proposed openstack/networking-baremetal master: Add netconf-openconfig device driver https://review.opendev.org/c/openstack/networking-baremetal/+/835324 | 15:32 |
| opendevreview | Harald Jensås proposed openstack/networking-baremetal master: Add LACP support to Netconf OpenConfig driver https://review.opendev.org/c/openstack/networking-baremetal/+/835378 | 15:32 |
| opendevreview | Harald Jensås proposed openstack/networking-baremetal master: Add support for pre-configured link aggregates https://review.opendev.org/c/openstack/networking-baremetal/+/840533 | 15:32 |
| * rpittau needs to start packing | 16:05 | |
| * rpittau bye everyone, see you on Monday! o/ | 16:05 | |
| opendevreview | Julia Kreger proposed openstack/ironic master: DNM: Multinode Grenade https://review.opendev.org/c/openstack/ironic/+/840530 | 16:19 |
| TheJulia | rpittau: so yeah, no configuration actions required. per debian docs, it only overwrites the integrated table | 17:37 |
| TheJulia | looks like ubuntu is the same, default config is embedded | 17:51 |
| rpittau | TheJulia: cool, I guess we could detect the OS or just do a try, not sure how to test that in ci though | 18:12 |
| * rpittau keeps packing | 18:12 | |
| TheJulia | I think we can do binary presence | 18:12 |
| TheJulia | and if a config file is present, I *suspect* it would be safe to skip module loading and just let everything else to rescan | 18:13 |
| rpittau | right | 18:13 |
| TheJulia | I'm going to try and pull it down and work on it, but I've got a call in 10 minutes and a call in 40 minutes... and I'm getting sick, so no idea how much progress I'll make between now and monday | 18:17 |
| sdanni | Hi! I have some questions about ironic HA settings. We've set 3 controller nodes. During baremetal node inspection, if the node gets a dhcp response from controller 0 or controller 2, the process will fail; if it gets the dhcp response from controller 1, it will download ipxe.boot file and install ramdisk/kernel. Then it failed at connection timeout when it curls the control_virtual_ip. Do you have any suggestions as | 19:39 |
| sdanni | to HA configuration? | 19:39 |
| sdanni | We have a provisioning network: 192.168.11.xx and a br-ex network: 192.168.24.xx. | 19:41 |
| sdanni | The baremetal node can only ping 192.168.11.xx and 192.168.24.9 (controller1 br-ex ip) | 19:42 |
| TheJulia | so.... that sounds really quite wrong | 19:51 |
| TheJulia | how did you setup this environment? | 19:51 |
| TheJulia | sdanni: ^ | 19:52 |
| sdanni | TheJulia: these are overcloud controller nodes | 19:53 |
| sdanni | We had to manually setup the provisioning vlan network, 192.168.11.x. And the gateway is controller1's ip:192.168.11.253 | 19:55 |
| sdanni | We have a /var/lib/neutron/dhcp/pxelinux.cfg/default like this: | 19:56 |
| sdanni | DEFAULT ipxe | 19:56 |
| sdanni | LABEL ipxe | 19:56 |
| sdanni | KERNEL ipxe.lkrn | 19:56 |
| sdanni | APPEND dhcp && chain http://192.168.11.253:8088/boot.ipxe | 19:56 |
| TheJulia | so... that is... wrong | 19:58 |
| TheJulia | why are you doing that? | 19:58 |
| sdanni | TheJulia: because the ipxe rom in the solaflare nic is too old and we are not able to update it. | 19:59 |
| TheJulia | what happens? | 20:00 |
| * TheJulia wonders if uefi would be happier | 20:00 | |
| sdanni | TheJulia: the ipxe in the nic can't recognize commands such as "inc" "chain" | 20:01 |
| sdanni | we figured out a way to custom dnsmasq configuration and use ipxe.klrn in the conductor. And that worked with 1 controller node environment. Now we are having network issues with HA | 20:03 |
| sdanni | i haven't tried uefi boot, not sure if that would be useful | 20:04 |
| TheJulia | sdanni: so most people who run into stuff like that just remove the retry logic from the ipxe template | 20:04 |
| TheJulia | since your running in a super special modified network, your kind of off the rails w/r/t openstack networking too, at least tested networking | 20:05 |
| TheJulia | which may explain what your running into | 20:05 |
| TheJulia | and... I think my next tweet will be something like friends don't let friends use bios mode | 20:05 |
| TheJulia | UEFI doesn't use firmware in the card | 20:05 |
| sdanni | TheJulia: It's worth a try! | 20:07 |
| TheJulia | that should skip the card trying to speak ipxe | 20:08 |
| TheJulia | if it still tries it, you might have firmware loading both legacy bios and uefi firmware loading happening, just set it to uefi only in the firmware and you should be good... Since RHEL also won't support Legacy firmware loading in UEFI mode. | 20:08 |
| sdanni | TheJulia: in a general HA environment, during node inspection, does bm node only connect to one controller or many? | 20:11 |
| TheJulia | sdanni: one by default, it is blocked from seeing the others so they can't service dhcp | 20:13 |
| TheJulia | but it can be done with dhcp only configs | 20:13 |
| TheJulia | as well | 20:13 |
| TheJulia | Actually, I thought that was the default | 20:13 |
| TheJulia | hjensas might remember better, I haven't touched that area in literal ages | 20:13 |
| sdanni | gotcha! | 20:14 |
| opendevreview | Harald Jensås proposed openstack/networking-baremetal master: Add netconf-openconfig device driver https://review.opendev.org/c/openstack/networking-baremetal/+/835324 | 21:16 |
| opendevreview | Harald Jensås proposed openstack/networking-baremetal master: Add LACP support to Netconf OpenConfig driver https://review.opendev.org/c/openstack/networking-baremetal/+/835378 | 21:17 |
| opendevreview | Harald Jensås proposed openstack/networking-baremetal master: Add support for pre-configured link aggregates https://review.opendev.org/c/openstack/networking-baremetal/+/840533 | 21:17 |
| * hjensas reads scrollback | 21:31 | |
| hjensas | sdanni: I get the feeling your network setup with controller1's ip:192.168.11.253 as gateway is not working, given the node can only ping 192.168.24.9 the traffic is filtered for some reason on the controller node. | 21:32 |
| opendevreview | Steve Baker proposed openstack/sushy-tools master: novadriver, don't use instance.flavor.id https://review.opendev.org/c/openstack/sushy-tools/+/840768 | 22:04 |
| opendevreview | Steve Baker proposed openstack/sushy-tools master: novadriver, don't use instance.flavor.id https://review.opendev.org/c/openstack/sushy-tools/+/840768 | 23:10 |
| opendevreview | Steve Baker proposed openstack/sushy-tools master: Fix per-request memoization https://review.opendev.org/c/openstack/sushy-tools/+/840780 | 23:10 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
