| EmilienM | stevebaker[m]: hey, happy new year :) | 00:26 |
|---|---|---|
| EmilienM | stevebaker[m]: i'm trying to get devstack+ironic working for some CI related to Gophercloud and since recently I'm having this error: https://paste.opendev.org/show/bMRr0OFT1VbloH9vJnRL/ I wonder if that rings a bell to you | 00:26 |
| EmilienM | iurygregory: ^ FYI | 00:26 |
| *** rcastillo|rover is now known as rcastillo | 02:15 | |
| *** pmannidi is now known as pmannidi|brb | 02:55 | |
| holtgrewe | Good morning Ironic \o | 06:48 |
| holtgrewe | dtantsur: Remember the VIF problem I had. This was caused by using "neutron" network interface instaed of "flat" ... I have not yet configured ironic-neutron interaction really. | 06:56 |
| arne_wiebalck | Good morning Ironic! | 07:30 |
| manuvakery1 | Error: Detected change in error condition on node Intel-Xeon-E-2236G-01. Error enabling the console on node 8e00140b-0116-4a6b-94c9-a2159ac0407f. Reason: Console subprocess failed to start. Timeout or error while waiting for console subprocess to start for node: 8e00140b-0116-4a6b-94c9-a2159ac0407f. Command: shellinaboxd -t -p 36080 --background=/tmp/8e00140b-0116-4a6b-94c9-a2159ac0407f.pid -s /:991:988:HOME:ipmitool | 07:45 |
| manuvakery1 | -I lanplus -H 172.16.1.100 -L ADMINISTRATOR -U ADMIN -f /tmp/8e00140b-0116-4a6b-94c9-a2159ac0407f.pw sol activate. Exit code: 0. Stdout: '' Stderr: '[server] Failed to find any available port!\n | 07:45 |
| manuvakery1 | getting above error while enabling console | 07:46 |
| manuvakery1 | when i tried to connect directly using IMPI its works well | 07:46 |
| manuvakery1 | ipmitool -I lanplus -H 172.16.1.100 -U ADMIN -P ***** sol activate | 07:46 |
| arne_wiebalck | manuvakery1: I do not know this part of the code in detail, but if it works like other IPMI interactions, Ironic may send the command and then waits for the effect (e.g. Ironic does not assume the power state was changed when the IPMI command was successful, but waits until 'power status' reports the node to be down). | 07:49 |
| arne_wiebalck | manuvakery1: My point is that the manual command may look like it works, but the actual action is not done (and Ironic is timing out while waiting for it). | 07:50 |
| arne_wiebalck | manuvakery1: So, I would suggest to have a look where Ironic is doing this command and check if there is a condition it is waiting for. | 07:51 |
| arne_wiebalck | manuvakery1: No, sorry, I think this is sth else: re-reading your error message, this seems to be an issue with shellinaboxd not being able to find a port (this is also why your manual IPMI command works, your BMC interaction is fine). | 08:11 |
| arne_wiebalck | manuvakery1: Maybe there is a shellinaboxd running already? | 08:12 |
| holtgrewe | Dear ironic, Is there any dead-tree book that you could recommend that is a bit more prose than the docs and not as comprehensive for all corner cases? | 08:19 |
| rpittau | good morning ironic! o/ | 08:27 |
| rpittau | TGIF! | 08:28 |
| manuvakery1 | arne_wiebalck: thanks for your response.. you are right there was multiple shellinaboxd was running.. once i cleaned all i am able to enable the console well | 08:30 |
| arne_wiebalck | manuvakery1: great :) | 08:37 |
| arne_wiebalck | good morning rpittau o/ | 08:37 |
| rpittau | hey arne_wiebalck :) | 08:40 |
| ajya | happy Friday ironic | 08:51 |
| arne_wiebalck | hey ajya o/ | 08:51 |
| ajya | TheJulia: thanks, I'll take a look next week probably | 08:51 |
| ajya | hi arne_wiebalck | 08:52 |
| rpittau | hey ajya :) | 08:53 |
| ajya | good morning rpittau | 08:53 |
| arne_wiebalck | manuvakery1: you have shellinabox working with your nodes? | 09:01 |
| manuvakery1 | arne_wiebalck: yes its working | 09:04 |
| arne_wiebalck | manuvakery1: nice! does this require kernel boot parameters in order to work? | 09:05 |
| manuvakery1 | yes i had to update the grub as follows to make it work | 09:08 |
| manuvakery1 | #!/bin/bash | 09:08 |
| manuvakery1 | for file in /etc/default/grub /etc/default/grub.d/*; do | 09:08 |
| manuvakery1 | sed -i 's/ttyS0/ttyS1,115200n8/g' $file | 09:08 |
| manuvakery1 | done | 09:08 |
| manuvakery1 | update-grub | 09:08 |
| manuvakery1 | reboot | 09:08 |
| arne_wiebalck | manuvakery1: thanks! this is on the instance you deployed, right? | 09:09 |
| manuvakery1 | yes | 09:09 |
| arne_wiebalck | manuvakery1: yeah, this was my understanding and one of the reasons why I did not try it: we would need to control the kernel parameters of the user's image to make this work | 09:11 |
| manuvakery1 | arne_wiebalck: you are right .. even i could not find any other way to make it work | 09:12 |
| arne_wiebalck | manuvakery1: I guess you could pass the above config via kernel boot parameters, no? | 09:12 |
| manuvakery1 | also grub changes varies per distribution | 09:12 |
| manuvakery1 | arne_wiebalck: i have tried passing it via kernal boot parameters but did not succeed .. not able to recall what exactly i faced | 09:14 |
| arne_wiebalck | manuvakery1: ok, thanks! | 09:14 |
| manuvakery1 | arne_wiebalck: you meant using pxe_append_params right? | 09:16 |
| arne_wiebalck | manuvakery1: one last thing if I may | 09:16 |
| arne_wiebalck | manuvakery1: were the docs in https://docs.openstack.org/ironic/latest/admin/console.html useful/helpful to set things up? | 09:16 |
| manuvakery1 | it was helpful but required more work related using COM2 for SOL port and because COM2 translates to ttyS1 and update the grub config | 09:21 |
| dtantsur | morning ironic | 09:21 |
| dtantsur | holtgrewe: writing a book was something TheJulia and I pondered in the past, but it never happened | 09:22 |
| opendevreview | Arne Wiebalck proposed openstack/ironic master: [doc] Add note about shellinabox console enable error https://review.opendev.org/c/openstack/ironic/+/823778 | 09:38 |
| holtgrewe | dtantsur: :-D | 09:48 |
| holtgrewe | So, now I have 239 ironic nodes. Does that count as a mid-sized installation already? | 09:49 |
| holtgrewe | dtantsur: Yeah, my problem with OS is that it's hard to find good introductory books (Packt are really hit and miss) or courses. | 09:51 |
| holtgrewe | (with missing being more likely at Packt) | 09:51 |
| dtantsur | holtgrewe: I'd say 100-1000 is mid-size, yes :) | 09:52 |
| dtantsur | well.. re books: I suspect that's because software vendors prefer to publish books on their products, and upstream communities don't have technical writers | 09:52 |
| holtgrewe | dtantsur: ;-) yeah, I'm seeing these 1/100 cases failures ... but can still handhold them through | 09:53 |
| holtgrewe | dtantsur: and also, OS is pretty well documented. I've seen worse documented commercial products (granted relatively niche). The commercial grid engine was ... well supported but training and docs were sub par. | 09:59 |
| dtantsur | that's nice to hear (re well documented ofc)! | 09:59 |
| holtgrewe | It's just that it's too much. | 10:00 |
| holtgrewe | :-D | 10:00 |
| holtgrewe | And hard to know where to start. | 10:00 |
| arne_wiebalck | In that spirit, there are some open doc patches if someone wants to have a look :-D https://review.opendev.org/c/openstack/ironic/+/821685 https://review.opendev.org/c/openstack/ironic/+/821272 https://review.opendev.org/c/openstack/ironic/+/823778 | 10:01 |
| arne_wiebalck | "in that spirit" == "well document", not "too much" :-) | 10:02 |
| dtantsur | arne_wiebalck: I'm double-checking the whole statement in the 1st patch since "core deploy steps" is no longer a thing | 10:10 |
| dtantsur | okay, we do still have that, but it's only about deploy.deploy | 10:11 |
| dtantsur | mgoddard: hi! do you remember why we're preventing changing priorities of the "core" deploy step (aka deploy.deploy)? | 10:12 |
| arne_wiebalck | dtantsur: tbh, I did not look at the content, I just ran into the broken link by accident | 10:14 |
| arne_wiebalck | dtantsur: if the target is changed or removed, the patch can be abandoned ofc | 10:15 |
| dtantsur | arne_wiebalck: I've proposed the wording change, see inline | 10:15 |
| opendevreview | Arne Wiebalck proposed openstack/ironic master: [doc] Fix broken link in "Node Deployment" https://review.opendev.org/c/openstack/ironic/+/821685 | 10:19 |
| arne_wiebalck | dtantsur: thanks, updated | 10:19 |
| arne_wiebalck | dtantsur: and thanks for the reviews on the other patches | 10:27 |
| dtantsur | sure | 10:27 |
| opendevreview | Merged openstack/ironic master: [doc] Add parameters for burn-in log files https://review.opendev.org/c/openstack/ironic/+/821272 | 10:27 |
| opendevreview | Aija Jauntēva proposed openstack/ironic master: Add more sources to redfish firmware upgrade https://review.opendev.org/c/openstack/ironic/+/822781 | 11:33 |
| iurygregory | good morning Ironic happy friday | 11:44 |
| iurygregory | EmilienM, I will take a look o/ | 11:44 |
| opendevreview | Aija Jauntēva proposed openstack/ironic master: Add more sources to redfish firmware upgrade https://review.opendev.org/c/openstack/ironic/+/822781 | 11:44 |
| EmilienM | iurygregory: thanks. Let me know if you find anything, I'm working on the PR for gophercloud where I pinged you | 11:55 |
| iurygregory | EmilienM, sure! | 11:59 |
| opendevreview | Vanou Ishii proposed openstack/ironic master: [WIP] Deal with iRMC virtual media incompatibility https://review.opendev.org/c/openstack/ironic/+/823790 | 12:25 |
| iurygregory | EmilienM, newbie question.. I'm trying to find neutron logs from the job https://github.com/gophercloud/gophercloud/runs/4732330343 any ideas where it would be? .-. | 12:35 |
| EmilienM | iurygregory: in https://github.com/gophercloud/gophercloud/actions/runs/1664963957 you click on one of the Artifacts, e.g. functional-baremetal-master and then you have logs in there | 12:45 |
| EmilienM | iurygregory: however I thought we were collecting the journal, but I don't see the file. I'll look at it | 12:45 |
| iurygregory | EmilienM, ack, I saw that we have a commit from stevebaker[m] to improve the problem related to tap device, but it seems is not helping in the gophercloud scenario | 12:46 |
| EmilienM | iurygregory: oh I know why I don't have systemd logs, I'll resend a patchset to have it, if this helps | 12:47 |
| iurygregory | so I'm wondering if could be the problem we had in neutron related to the firewall (just thinking out loud XD) | 12:48 |
| iurygregory | oh nice! | 12:48 |
| dtantsur | make sure you're not trying to use OVN | 12:48 |
| EmilienM | oh | 12:49 |
| EmilienM | I'm using devstack defaults | 12:50 |
| dtantsur | EmilienM: https://opendev.org/openstack/ironic/src/branch/master/zuul.d/ironic-jobs.yaml#L54-L56 | 12:50 |
| EmilienM | and these overrides: https://github.com/gophercloud/gophercloud/blob/0735c975876addd65d9725faefe754813e794e7f/.github/workflows/functional-baremetal.yaml#L41-L62 | 12:50 |
| dtantsur | actually, copying this definition is highly recommended | 12:50 |
| iurygregory | oh right :D | 12:50 |
| iurygregory | we don't like OVN :D | 12:50 |
| iurygregory | or better, OVN doesn't like Ironic :D | 12:50 |
| EmilienM | ok so I guess I can stop investigating and add that to my overrides | 12:51 |
| iurygregory | ++ | 12:51 |
| EmilienM | and hi dtantsur :) (and thanks) | 12:51 |
| dtantsur | EmilienM: also IRONIC_VM_SPECS_RAM=384 is too optimistic for year 2022 :( | 12:51 |
| dtantsur | I'd start with 1G and work you way down if possible | 12:51 |
| iurygregory | 1G for tinyipa? we are already on this? O.o | 12:52 |
| iurygregory | we don't test with DIB in gophercloud I think | 12:52 |
| EmilienM | iurygregory: I hope not | 12:52 |
| EmilienM | but i'll update to 1024 in case | 12:52 |
| dtantsur | I think we *can* do 512M or even less, but I'd start with a safer value | 12:52 |
| dtantsur | 1G is not for DIB, DIB requires at least 2.5G | 12:52 |
| iurygregory | oh right IRONIC_VM_SPECS_RAM: 1024 | 12:53 |
| iurygregory | for tinyipa jobs :D | 12:53 |
| dtantsur | yeah | 12:53 |
| dtantsur | it's a bit more than needed, but it makes the jobs more reliable | 12:53 |
| arne_wiebalck | iurygregory: two doc reviews if you have moment: https://review.opendev.org/c/openstack/ironic/+/823778 https://review.opendev.org/c/openstack/ironic/+/821685 | 12:58 |
| iurygregory | arne_wiebalck, sure! | 13:00 |
| EmilienM | iurygregory: same error. I'm looking at logs | 13:19 |
| iurygregory | EmilienM, ack I will check in a few (finishing some reviews) | 13:20 |
| dtantsur | iurygregory: please add https://review.opendev.org/c/openstack/bifrost/+/823671 to your plans if possible | 13:20 |
| dtantsur | we've had a few people recently confused by our docs | 13:20 |
| iurygregory | dtantsur, sure =) | 13:24 |
| *** rcastillo is now known as rcastillo|rover | 13:28 | |
| opendevreview | Merged openstack/ironic bugfix/19.0: Configure CI for bugfix/19.0 https://review.opendev.org/c/openstack/ironic/+/822337 | 13:32 |
| opendevreview | Arne Wiebalck proposed openstack/ironic master: [doc] Add note about shellinabox console enable error https://review.opendev.org/c/openstack/ironic/+/823778 | 13:45 |
| EmilienM | iurygregory: I'm wondering if 20s can be too short, in some environments | 14:04 |
| EmilienM | I have collected the neutron logs: https://paste.opendev.org/show/b1yCOJ2gxBmtwBXolLP6/ | 14:06 |
| opendevreview | Merged openstack/ironic master: [doc] Fix broken link in "Node Deployment" https://review.opendev.org/c/openstack/ironic/+/821685 | 14:06 |
| EmilienM | i'm actually thinking it's because the network was not created, q-dhcp seems not deployed | 14:06 |
| iurygregory | let me check the list of services we have enable in the job in ironic | 14:07 |
| iurygregory | I think 20sec would be enough since we also retry.. | 14:07 |
| EmilienM | iurygregory: don't investigate yet, sorry for noise. Let me retry by making sure I ahve q-dhcp | 14:09 |
| iurygregory | EmilienM, ack =) and don't worry because my timezone now is in Brazil so I still have some time to look today :D | 14:10 |
| EmilienM | iurygregory: oh cool. you have sun. I have snow. | 14:10 |
| iurygregory | send snow please :D | 14:12 |
| arne_wiebalck | dtantsur: can you move this one fwd once more, please? (iurygregory spotted a missing ' which broke rendering in CI), so I had to push an update: https://review.opendev.org/c/openstack/ironic/+/823778 | 14:12 |
| arne_wiebalck | EmilienM: iurygregory: just started snowing here, too :) | 14:13 |
| iurygregory | =( | 14:13 |
| opendevreview | Merged openstack/bifrost master: Clean up the "How to" documentation https://review.opendev.org/c/openstack/bifrost/+/823671 | 14:18 |
| dtantsur | I really loved spending a week in the snow | 14:20 |
| EmilienM | dtantsur: did you have to shovel? or just enjoy the snow :D | 14:22 |
| dtantsur | EmilienM: does skiing count for shoveling? :) | 14:23 |
| dtantsur | I was in the alps for the holidays. There is little hope for snow where I normally live. | 14:24 |
| EmilienM | skiing doesn't count no, :D | 14:25 |
| EmilienM | I'm trying to run devstack/ironic on stable/train w/ python3 and it's using py2 for installing virtualbmc: | 14:27 |
| EmilienM | https://pipelines.actions.githubusercontent.com/VpcJmuZKqOQwnBakVAncoCADDbxjEFKEE2Y2lfgYJECEuVLego/_apis/pipelines/1/runs/842/signedlogcontent/3?urlExpires=2022-01-07T14%3A27%3A13.4250133Z&urlSigningMethod=HMACV1&urlSignature=xxW%2FLGU6BnwIRHjloq7xwndmyzNG68t7RPCsBiqYHc8%3D | 14:27 |
| EmilienM | have you seen that error before? | 14:27 |
| dtantsur | Train was long ago... | 14:27 |
| EmilienM | "ImportError: No module named _internal.cli.main" | 14:27 |
| dtantsur | yeah, it rings some bells, but nothing more than bells | 14:28 |
| dtantsur | your link gives "Uri expired" btw | 14:28 |
| EmilienM | oh oops | 14:29 |
| dtantsur | we did have some jobs with py3 on train | 14:31 |
| EmilienM | yeah https://opendev.org/openstack/ironic/src/branch/stable/train/zuul.d/ironic-jobs.yaml#L208-L215 | 14:31 |
| EmilienM | and it's weird: | 14:31 |
| EmilienM | /opt/stack/ironic/devstack/lib/ironic:install_virtualbmc:800 : USE_PYTHON3=False | 14:31 |
| EmilienM | USE_PYTHON3=True | 14:31 |
| EmilienM | from the logs | 14:31 |
| EmilienM | it's like the ironic plugin doesn't read it right | 14:32 |
| dtantsur | EmilienM: https://opendev.org/openstack/ironic/src/branch/stable/train/devstack/lib/ironic#L797 | 14:32 |
| dtantsur | :( | 14:32 |
| arne_wiebalck | dtantsur: ty | 14:32 |
| EmilienM | dtantsur: oops :/ ok now I understand | 14:32 |
| dtantsur | EmilienM: I wonder if you can work around it by adding pyghmi and virtualbmc to required-projects. potentially capping at some tag. | 14:32 |
| * TheJulia attempts to wake up | 14:33 | |
| EmilienM | https://paste.opendev.org/show/811966/ my error | 14:33 |
| dtantsur | fun | 14:33 |
| dtantsur | I think it's a half-installed pip | 14:33 |
| dtantsur | I see these things nearly every Fedora upgrade | 14:34 |
| dtantsur | e.g. in 35 they updated to Python 3.10, so my packages installed into ~/.local/lib/python3.9 stopped importing | 14:34 |
| dtantsur | while ~/.local/bin/<foo> still existed | 14:34 |
| EmilienM | I'll try use_library_from_git | 14:34 |
| dtantsur | not sure how that would work with 2.7 though | 14:34 |
| dtantsur | good morning TheJulia | 14:35 |
| EmilienM | TheJulia: o/ | 14:36 |
| dtantsur | EmilienM: btw, you're running devstack on github actions? does it work well? | 14:37 |
| dtantsur | for rust-openstack I'm using openlab zuul, but it hasn't been exceptionally stable recently | 14:37 |
| EmilienM | dtantsur: i'm moving out from openlab zuul | 14:38 |
| EmilienM | it has been terrible for us | 14:38 |
| EmilienM | and i created a devstack-action for github | 14:38 |
| EmilienM | https://github.com/EmilienM/devstack-action | 14:38 |
| dtantsur | fun | 14:38 |
| EmilienM | check this out https://github.com/gophercloud/gophercloud/pull/2307 | 14:38 |
| dtantsur | I'd probably migrate under the SDK project on opendev.org though | 14:38 |
| EmilienM | in 19m, I can deploy devstack and run acceptance test | 14:39 |
| dtantsur | wow | 14:39 |
| dtantsur | I was afraid it would exceed some quotas, e.g. on maximum run time | 14:39 |
| EmilienM | dtantsur: yeah we had a thread about it and people weren't happy with the idea, most of folks in gophercloud aren't opendev contribs and aren't happy with gerrit | 14:39 |
| dtantsur | yeah, I've been following it | 14:39 |
| dtantsur | well, I'm pretty much the only contributor to rust-openstack, sooo :D | 14:39 |
| EmilienM | https://github.com/gophercloud/gophercloud/issues/2257 | 14:39 |
| opendevreview | Merged openstack/ironic master: [doc] Add note about shellinabox console enable error https://review.opendev.org/c/openstack/ironic/+/823778 | 14:42 |
| opendevreview | Harald Jensås proposed openstack/ironic master: Ensure 'port' is up2date after binding:host_id https://review.opendev.org/c/openstack/ironic/+/823811 | 14:48 |
| EmilienM | iurygregory: btw fixing neutron services worked fine :) the job is almost ready I think | 15:02 |
| EmilienM | iurygregory: trying to make it work on train and I think it's ready to go | 15:03 |
| EmilienM | dtantsur: the reason why gophercloud hasn't moved to opendev (yet) is because for non opendev contributors, the amount of things to figure out before sending a patch can be discouraging and our community is too small to afford any member lost. | 15:05 |
| dtantsur | yep, I can imagine | 15:05 |
| EmilienM | CLA, accounts, Gerrit, Zuul, Launchpad, etc. Not that I'm a fan of Github, but in this case "the best tool is the one you already use" would apply | 15:05 |
| EmilienM | I would be very happy if we move to opendev one day but for now this decision would have bad impact I suspect | 15:06 |
| iurygregory | EmilienM, \o/ awesome | 15:11 |
| * iurygregory is wondering why hexchat is not showing notifications when he receives a ping D: | 15:12 | |
| EmilienM | dtantsur: installing vbmc from source solved the issue btw. | 15:13 |
| EmilienM | thx | 15:14 |
| EmilienM | iurygregory: I have an acceptance issue on train: https://github.com/gophercloud/gophercloud/runs/4739989972?check_suite_focus=true | 15:14 |
| EmilienM | Failure in nodes_test.go, line 103: unexpected error "Bad request with: [PUT http://10.1.0.77/baremetal/nodes/b63fd4a0-1176-4c1a-99fe-26928b189b9a/states/raid], error message: {\"error_message\": \"{\\\"faultcode\\\":\\\"Client\\\",\\\"faultstring\\\":\\\"RAID config validation error: {'size': '> 100'} is not of type 'string'\\\",\\\"debuginfo\\\":null}\"}" | 15:14 |
| * iurygregory looks | 15:14 | |
| dtantsur | I wonder if you need to set the maximum microversion | 15:14 |
| iurygregory | dtantsur, the test should set the microversion so it should be ok... | 15:16 |
| iurygregory | https://github.com/gophercloud/gophercloud/blob/18a1ee5aa4850db7055aa5a125f606ac07ba668e/acceptance/openstack/baremetal/v1/nodes_test.go#L76 | 15:16 |
| EmilienM | that could be a bug in https://github.com/gophercloud/gophercloud/blob/master/openstack/baremetal/v1/nodes/requests.go | 15:17 |
| dtantsur | EmilienM, iurygregory, requesting a microversion does not make the feature availlable :) | 15:17 |
| dtantsur | I'm quite confident I wrote this feature after Train | 15:17 |
| dtantsur | unfortunately, this particular feature is not (?) microversioned | 15:18 |
| EmilienM | maybe I should just skip this test > train | 15:18 |
| dtantsur | yes | 15:19 |
| EmilienM | client.Microversion = "1.50" | 15:19 |
| iurygregory | dtantsur, ok now i see what you meant =) | 15:19 |
| dtantsur | EmilienM: 1.50 has nothing to do with RAID https://docs.openstack.org/ironic/latest/contributor/webapi-version-history.html#id18 | 15:19 |
| dtantsur | I wonder where it's coming from | 15:19 |
| EmilienM | Stephen Benjamin from OpenShift wrote this test (git blame) | 15:20 |
| dtantsur | I think it's enough to use 1.12 to deal with RAID | 15:20 |
| dtantsur | that stbenjam :) | 15:20 |
| dtantsur | this is arbitrary as well: https://github.com/gophercloud/gophercloud/blob/18a1ee5aa4850db7055aa5a125f606ac07ba668e/acceptance/openstack/baremetal/v1/nodes_test.go#L21 | 15:20 |
| dtantsur | EmilienM: generally, https://docs.openstack.org/ironic/latest/contributor/webapi-version-history.html is your friend when you need to figure out the version you need | 15:21 |
| EmilienM | should I skip this test on train? or bump anything? | 15:21 |
| dtantsur | EmilienM: just skip on Train (and consider setting microversions to realistic values) | 15:22 |
| EmilienM | 1.12 instead of 1.50 | 15:22 |
| EmilienM | gotcha | 15:22 |
| opendevreview | Harald Jensås proposed openstack/ironic master: Add dhcp options for each ip_version once https://review.opendev.org/c/openstack/ironic/+/823817 | 15:28 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic master: ImageCache: respect Cache-Control: no-store https://review.opendev.org/c/openstack/ironic/+/822329 | 15:28 |
| opendevreview | Harald Jensås proposed openstack/ironic master: Add dhcp options for each ip_version once https://review.opendev.org/c/openstack/ironic/+/823817 | 15:29 |
| opendevreview | Harald Jensås proposed openstack/ironic master: Add dhcp options for each ip_version once https://review.opendev.org/c/openstack/ironic/+/823817 | 15:33 |
| dtantsur | folks, could you please review https://review.opendev.org/c/openstack/ironic/+/821219 since it has a high potential of conflicts? | 15:43 |
| dtantsur | also the bifrost changes culminating in https://review.opendev.org/c/openstack/bifrost/+/820479/ would be cool to land | 15:44 |
| iurygregory | dtantsur, I will take a look after lunch o/ | 16:01 |
| dtantsur | thx | 16:08 |
| EmilienM | iurygregory: when time permits (even next week), please have a look at https://github.com/gophercloud/gophercloud/pull/2308 - this is going to be the CI job to test Ironic on gophercloud. Covering all releases until Train. | 16:38 |
| iurygregory | EmilienM, sure! | 16:51 |
| arne_wiebalck | bye everyone, have a good week-end o/ | 17:10 |
| jamesdenton | are the IPA images available at tarballs.opendev.org sufficient for PXE booting a VM? | 17:15 |
| iurygregory | bye arne_wiebalck o/ | 17:15 |
| iurygregory | jamesdenton, they should (dib images and even tinyipa since you mentioned VMs) | 17:18 |
| iurygregory | don't use tinyipa if is a production env | 17:18 |
| jamesdenton | good to know | 17:18 |
| jamesdenton | thanks -- the dib images are working to boot an OS, but having some issues inside connecting back to the API | 17:21 |
| jamesdenton | probably something on my side, i'll keep looking | 17:22 |
| TheJulia | can I jusst sleep all weekend?! | 17:25 |
| dtantsur | TheJulia: I think it's a sensible idea | 17:30 |
| dtantsur | I'm pondering the same, maybe with one bouldering session in the middle | 17:30 |
| dtantsur | jamesdenton: the DIB images we publish are configured to try DHCP on all interfaces, so they should work in most cases | 17:31 |
| jamesdenton | thanks dtantsur - i think this might be an issue with me attempting to perform an introspection w/o a baremetal port for the node. Seems when IPA reaches out to API:5050 with POST, inspector API kills the connection. verifying now | 17:35 |
| jamesdenton | i'm assuming because it has no way to identify the node? i dunno | 17:35 |
| dtantsur | jamesdenton: kills the connection? check for firewall, selinux and http-https mismatch | 17:38 |
| jamesdenton | by kill i mean FIN | 17:38 |
| jamesdenton | sorry | 17:38 |
| jamesdenton | just a premature FIN, though, i would think. because once IPA sends the payload then it gets RST | 17:38 |
| dtantsur | yeah, that's why I suggested firewall as the most likely source | 17:38 |
| dtantsur | inspector returns a normal HTTP 40? on failure | 17:39 |
| jamesdenton | alright, so it looks like this may be related to IPA sending to http vs https - all of the endpoints are https | 17:49 |
| jamesdenton | and i guess that's due to a bunk inspector.ipxe | 17:51 |
| jamesdenton | *facepalm* | 17:51 |
| dtantsur | time to go, have a great weekend folks o/ | 18:21 |
| jamesdenton | thanks for the help | 18:35 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!