opendevreview | Steve Baker proposed openstack/ironic-python-agent stable/wallaby: Fix UEFI record regex https://review.opendev.org/c/openstack/ironic-python-agent/+/817014 | 01:40 |
---|---|---|
opendevreview | Steve Baker proposed openstack/ironic-python-agent bugfix/8.1: Fix UEFI record regex https://review.opendev.org/c/openstack/ironic-python-agent/+/817470 | 01:46 |
opendevreview | Steve Baker proposed openstack/ironic-python-agent bugfix/8.1: Fix UEFI record regex https://review.opendev.org/c/openstack/ironic-python-agent/+/817470 | 01:53 |
opendevreview | Steve Baker proposed openstack/ironic-python-agent stable/victoria: Fix UEFI record regex https://review.opendev.org/c/openstack/ironic-python-agent/+/817018 | 01:59 |
opendevreview | Steve Baker proposed openstack/ironic-python-agent stable/ussuri: Fix UEFI record regex https://review.opendev.org/c/openstack/ironic-python-agent/+/817021 | 02:03 |
opendevreview | Steve Baker proposed openstack/ironic-python-agent stable/train: Fix UEFI record regex https://review.opendev.org/c/openstack/ironic-python-agent/+/817026 | 02:07 |
opendevreview | Merged openstack/ironic master: [iRMC] Convert the type of irmc_port to int https://review.opendev.org/c/openstack/ironic/+/817154 | 04:26 |
opendevreview | Merged openstack/ironic-python-agent stable/victoria: Output verbose info from efibootmgr https://review.opendev.org/c/openstack/ironic-python-agent/+/817016 | 06:05 |
opendevreview | Merged openstack/ironic-python-agent stable/victoria: Delete EFI boot entry duplicate labels first https://review.opendev.org/c/openstack/ironic-python-agent/+/817017 | 06:08 |
arne_wiebalck | Good morning, Ironic! | 07:49 |
muellerbe | good morning arne_wiebalck | 07:49 |
arne_wiebalck | hey muellerbe o/ | 07:53 |
iurygregory | good morning arne_wiebalck and Ironic o/ | 07:55 |
arne_wiebalck | hey iurygregory o/ | 07:55 |
rpittau | good morning ironic! o/ | 07:57 |
iurygregory | morning rpittau o/ | 08:00 |
rpittau | hey iurygregory :) | 08:01 |
arne_wiebalck | hey rpittau o/ | 08:01 |
arne_wiebalck | iurygregory: dtantsur: from mraineri 's answer here https://redfishforum.com/thread/565/weak-etags, I would say sushy should not remove the qualifier (see also dtantsur 's intuition :) | 08:04 |
arne_wiebalck | iurygregory: dtantsur: that leaves still two options: a) remove the stripping (may break timeu 's hardware) or b) use both tags in the response's header ... thoughts? | 08:06 |
iurygregory | arne_wiebalck, a can cause problems for Lenovo HW, b I'm ok but it seems weird sending both etags..., maybe option c? c - since ironic has logic to identify vendor we can probably add an option to insert_media in sushy https://github.com/openstack/sushy/blob/master/sushy/resources/manager/virtual_media.py#L110-L111 that would tell if we need to strip the W/ or not .... | 08:17 |
iurygregory | by default the variable wouldn't remove anything from the ETag, and in ironic we can add a logic for lenovo as a vendor to strip when using vmedia | 08:19 |
iurygregory | ofc I'm not against option b =) (but I'm not sure how BMC's will handle in case we are sending two values... | 08:20 |
arne_wiebalck | iurygregory: sending two values is standard compliant | 08:28 |
iurygregory | oh \o/ | 08:28 |
iurygregory | ok cool :D | 08:28 |
arne_wiebalck | iurygregory: and it I tested this with my hardware | 08:28 |
iurygregory | <insert happy dance gif> | 08:29 |
arne_wiebalck | iurygregory: it just looks like the big gun | 08:29 |
iurygregory | kinda =) | 08:29 |
arne_wiebalck | iurygregory: doing per vendor massaging of the header seems worse, though | 08:29 |
iurygregory | like in brazil we would say you are using a bazooka to kill an ant :D | 08:29 |
arne_wiebalck | iurygregory: so, maybe b) is not such a bad option after all | 08:29 |
arne_wiebalck | iurygregory: heh | 08:30 |
iurygregory | yeah, I was worried if some BMCs will complain... | 08:30 |
arne_wiebalck | iurygregory: let's see what dtantsur thinks once he comes online | 08:30 |
iurygregory | because... standards :D | 08:30 |
rpittau | "standards" | 08:30 |
arne_wiebalck | iurygregory: oh, totally possible this will create new issues | 08:30 |
iurygregory | rpittau, yeah I forgot the " " :D | 08:31 |
rpittau | :) | 08:31 |
iurygregory | arne_wiebalck, yup =) | 08:31 |
dtantsur | arne_wiebalck: well, what mraineri suggests contradicts the ETag RFC | 08:47 |
dtantsur | I'm fine with sending both, but I'm worried about the message they are sending | 08:48 |
arne_wiebalck | dtantsur: yes | 08:48 |
arne_wiebalck | dtantsur: I agree | 08:48 |
arne_wiebalck | "weak Etags on If-* should never match" | 08:48 |
* dtantsur answers there | 08:50 | |
arne_wiebalck | +1, I was about to ask you :) | 08:50 |
dtantsur | done | 08:54 |
opendevreview | Dmitry Tantsur proposed openstack/bifrost master: [WIP] iPXE fallback script https://review.opendev.org/c/openstack/bifrost/+/818000 | 08:58 |
arne_wiebalck | dtantsur: thanks | 10:01 |
arne_wiebalck | dtantsur: I suggest none the less to put up a patch for sushy to send back both etags (even if temporary): while weak etags should not be used, we seem to have two cases already where this is the case and we have two different behaviours :-S | 10:03 |
dtantsur | heh, yeah | 10:04 |
dtantsur | omg, why is bifrost job succeeding on failure? O__o | 10:04 |
arne_wiebalck | Error: SUCCESS | 10:04 |
iurygregory | WOOT?! | 10:05 |
dtantsur | I think it's a regression in https://opendev.org/openstack/bifrost/commit/b31bc66726ddcff11678a77cce8b14f629a1b19d | 10:05 |
iurygregory | I'm trying to recall the reason why TheJulia needed this | 10:07 |
dtantsur | I think it's kinda wrong, EXIT should catch it | 10:07 |
dtantsur | I'll propose a revert (which may fail because tinyIPA fails with the DNS problem) | 10:07 |
iurygregory | ouch =( | 10:07 |
opendevreview | Dmitry Tantsur proposed openstack/bifrost master: Revert "Explicitly trap on ERR" https://review.opendev.org/c/openstack/bifrost/+/817978 | 10:09 |
dtantsur | TheJulia: ^^^ | 10:09 |
dtantsur | we have tinyIPA basically broken because of eventlet: https://zuul.opendev.org/t/openstack/build/a2c04cad4f054f5a958148a91d4f03ed/log/logs/testvm1_console.log#1458 | 10:10 |
rpittau | \o/ | 10:12 |
iurygregory | ofc eventlet =( | 10:15 |
* dtantsur opens eventlet source | 10:22 | |
dtantsur | https://github.com/eventlet/eventlet/pull/737 | 10:39 |
dtantsur | now I need to do groceries shopping and then figure out how to test it | 10:40 |
dtantsur | cc janders | 10:40 |
janders | dtantsur NICE! :) | 12:25 |
janders | if anyone has time, I would really appreciate reviews on https://review.opendev.org/c/openstack/sushy/+/817137 | 12:25 |
janders | w/r/t eventlet CI issue this is exactly what I ran into the other day | 12:27 |
janders | (just didn't get far looking into eventlet code) | 12:27 |
iurygregory | janders, feel free to add ironic-week-prio as a hashtag in your patch so it will be in the prio list for review =) | 12:28 |
janders | iurygregory thank you, done! | 12:30 |
janders | see you tomorrow Ironic o/ | 12:31 |
iurygregory | bye janders o/ (see you Thursday :D ) | 12:32 |
TheJulia | dtantsur: I think without catching it was failing and not being caught at all | 13:28 |
TheJulia | dtantsur: I *think* | 13:28 |
TheJulia | dtantsur: stevebaker[m] was fighting it when he spotted it, I just happened to whip up a patch first | 13:28 |
* TheJulia wonders if the McDonalds is open | 13:29 | |
iurygregory | TheJulia, fyi regarding the tinyipa problem in ussuri/train, maybe we can try using --insecure? dtantsur proposed this in the meeting yesterday =) | 13:33 |
TheJulia | iurygregory: insecure is really unacceptable imho | 13:33 |
TheJulia | because MitM attacks could occur on code in our repositories | 13:34 |
iurygregory | oh ok =) | 13:34 |
TheJulia | Also, can't patch old IPA-builder | 13:34 |
iurygregory | oh right =( we didn't have ipa-b with branches at that time... | 13:35 |
TheJulia | and infra flat out told us "absolutely not" when we asked if we could go back and branch to do a 2.2.x fix to address some of the issues we were having a while back | 13:35 |
TheJulia | yeah | 13:35 |
TheJulia | well, I think it was more the openstack release team | 13:35 |
TheJulia | than infra | 13:35 |
TheJulia | I bet infra would have just done it | 13:35 |
iurygregory | probably... | 13:35 |
TheJulia | Anyway, I need coffee and to eat | 13:35 |
TheJulia | I got 5 hours of sleep \o/ | 13:35 |
TheJulia | Amongst trucks | 13:35 |
iurygregory | coffee and food ++ | 13:35 |
iurygregory | yeah I saw the email and twitter =( | 13:36 |
iurygregory | good luck! | 13:36 |
TheJulia | iurygregory: Yeah, I guess I'm getting all my broken down in some strange city credits for life | 13:47 |
dtantsur | TheJulia: IIRC the tinyIPA build is already insecure | 13:49 |
dtantsur | good morning | 13:49 |
dtantsur | hence my suggestion | 13:49 |
TheJulia | dtantsur: are we explicitly disabling cert verification anywhere else? | 13:49 |
dtantsur | I *think* (double-checking now) TCL repos don't use TLS | 13:50 |
TheJulia | I guess the conundrum beyond that is we would have to hack 10.x and 12.x support in side by side into master branch | 13:50 |
dtantsur | TheJulia: https://opendev.org/openstack/ironic-python-agent-builder/src/branch/master/tinyipa/tc-mirror.sh#L12 | 13:50 |
TheJulia | woot | 13:51 |
dtantsur | that's one of the reasons we don't recommend tinyIPA in production | 13:51 |
TheJulia | I always viewed it as drivers, but that is definitely another solid reason | 13:52 |
Alex-dev | Hi! I'm trying to deploy OS (centos stream) via ironic (tripleo setup Victoria) and got error: | 13:59 |
Alex-dev | Failed to provision instance 2318a720-783c-41bc-8842-cc793a7cf389: Deploy step deploy.write_image failed on node 95442d75-44f8-4783-ba24-783d1d008eae. Error performing deploy_step write_image: Command execution failed: Unexpected error while running command.\nCommand: blkid /dev/sda --match-tag UUID --match-tag PARTUUID\nExit code: 2 | 13:59 |
rpittau | Alex-dev: that's a known error | 13:59 |
rpittau | this should fix the issue https://review.opendev.org/c/openstack/ironic-lib/+/817922 | 14:00 |
Alex-dev | Thanks! I will try | 14:00 |
rpittau | oh wait that's centos stream though, so should be compatible with the long variable names | 14:01 |
rpittau | Alex-dev can you post the full log somewhere ? | 14:01 |
rpittau | Alex-dev: just to confirm, are you using a centos stream based ipa ramdisk ? | 14:01 |
Alex-dev | I seen in ipmi console centos stream. in which log file I can find info about it? ironic, ironic-conductor, nova? | 14:03 |
rpittau | Alex-dev: the ironic-python-agent ramdisk logs, you can see how to get them here: https://docs.openstack.org/ironic/latest/admin/troubleshooting.html#retrieving-logs-from-the-deploy-ramdisk | 14:05 |
rpittau | Alex-dev: that looks like the same error from centos 7, but since you mentioned centos stream it may be something different. At least see if you can post the conductor logs somewhere, as a start | 14:09 |
Alex-dev | conductor logs: https://drive.google.com/file/d/1V1rMD7bGwkK13wira8gNTH403e5B9UZi/view?usp=sharing | 14:15 |
TheJulia | I *think* another diesel broke down in the parking lot | 14:15 |
Alex-dev | @rpittau ^ | 14:15 |
* TheJulia hears telltale signs of air break problems | 14:15 | |
TheJulia | brake | 14:16 |
rpittau | Alex-dev: I don't see any errors there, where did you get the original error from ? | 14:18 |
* TheJulia attempts to patiently wait for the mobile mechanic | 14:21 | |
Alex-dev | rpittau: in nova-conductor and nova-compute logs, I will share it now | 14:25 |
TheJulia | so it is exposing last_error | 14:27 |
TheJulia | and that gets reset if cleaning is enabled and it rolls back through successfully | 14:27 |
* TheJulia has this sneaking suspicion | 14:28 | |
Alex-dev | rpittau: https://drive.google.com/drive/folders/1pZn04cVcJskD3JpgXJ5PQnS5kMg5TUUr?usp=sharing | 14:28 |
rpittau | Alex-dev: ok, I can see the error there, it's taken from ironic conductor but I can't see the same in the ironic conductor logs, they seem cut | 14:32 |
rpittau | would be great to have the full ironic conductor logs plus understand if the deploy image is using centos 7 | 14:32 |
Alex-dev | Where I can get actual builded ipa images for update it ? this error expected only on centos 7? I want try again on latest images | 14:37 |
TheJulia | https://tarballs.opendev.org/openstack/ironic-python-agent/dib/ | 14:38 |
TheJulia | Alex-dev_: ^^ | 14:40 |
TheJulia | note the dates | 14:40 |
TheJulia | specifically we had to stop building centos7 a long time ago because we dropped python 2.x support | 14:40 |
rpittau | Alex-dev: that should happen only on centos 7, yes | 14:40 |
* TheJulia looses her mind | 14:41 | |
Alex-dev_ | ok, I will use ipa-centos8-stable-victoria.tar.gz. tryed and get back | 14:41 |
TheJulia | okay | 14:42 |
* TheJulia resumes waiting for the mechanic | 14:42 | |
* TheJulia is getting super antsy | 14:42 | |
* TheJulia wants to get home... badly | 14:42 | |
TheJulia | Anyone taking odds on if I'll get home today? | 14:44 |
rpittau | TheJulia: I truly hope you get home today :) | 14:45 |
TheJulia | I sure hope so. :( | 14:57 |
TheJulia | I can't eat any more McDonalds | 14:57 |
TheJulia | Ohhh, mechanic has arrived | 14:57 |
iurygregory | \o/ yay | 15:08 |
* TheJulia hopes by lunch time to be on the road | 15:20 | |
* TheJulia hopes | 15:20 | |
* TheJulia has been hoping a lot | 15:20 | |
TheJulia | If y'all haven't figured it out, your like my only human contact right now | 15:23 |
* dtantsur is an owl contact | 15:28 | |
dtantsur | TheJulia: fingers crossed that it's resolved this time! | 15:28 |
iurygregory | owl contact :D | 15:29 |
TheJulia | Owls are awesome | 15:31 |
dtantsur | ++ can confirm (or cat confirm?) | 15:31 |
TheJulia | Confirm... at the moment. I think | 15:32 |
TheJulia | Kittah overlord was loaded into the car and taken home today. | 15:33 |
dtantsur | good to be a kitty! | 15:35 |
*** dviroel is now known as dviroel_ | 15:37 | |
TheJulia | woot, two more bolts to come off, then four to go back on | 15:41 |
TheJulia | And then resume the belt replacement | 15:42 |
opendevreview | Arne Wiebalck proposed openstack/sushy master: Handle weak Etags https://review.opendev.org/c/openstack/sushy/+/818110 | 15:45 |
opendevreview | Arne Wiebalck proposed openstack/sushy master: Use Etag in precondition header when setting boot options https://review.opendev.org/c/openstack/sushy/+/818114 | 15:47 |
arne_wiebalck | dtantsur: iurygregory ^^ returning both etag flavors and using them for patch (boot options only) | 15:49 |
iurygregory | arne_wiebalck, ack =D | 15:49 |
opendevreview | Merged openstack/ironic-lib master: Restore blkid compatibility with Centos 7 https://review.opendev.org/c/openstack/ironic-lib/+/817922 | 15:50 |
opendevreview | Riccardo Pittau proposed openstack/ironic-lib stable/xena: Restore blkid compatibility with Centos 7 https://review.opendev.org/c/openstack/ironic-lib/+/817983 | 15:54 |
opendevreview | Riccardo Pittau proposed openstack/ironic-lib stable/wallaby: Restore blkid compatibility with Centos 7 https://review.opendev.org/c/openstack/ironic-lib/+/817984 | 15:55 |
opendevreview | Dmitry Tantsur proposed openstack/ironic-python-agent master: Move manage_uefi from the image extension to a public location https://review.opendev.org/c/openstack/ironic-python-agent/+/815651 | 16:58 |
dtantsur | arne_wiebalck: could you check that I've applied your changes correctly ^^^? | 16:58 |
arne_wiebalck | dtantsur: I will! | 17:00 |
arne_wiebalck | dtantsur: etags, so my vendor's implementation is correct when it comes to the redfish standard ... I just wonder why it took me so long to realise what to send back ;) | 17:18 |
*** sshnaidm is now known as sshnaidm|afk | 17:33 | |
dtantsur | sigh | 17:33 |
arne_wiebalck | take two standards, combine them with some implementation creativity and you end up with a nice explosive result :) | 17:42 |
arne_wiebalck | bye everyone o/ | 17:44 |
dtantsur | yeah.. | 17:45 |
dtantsur | I'm going as well, see you | 17:45 |
*** tosky is now known as Guest6054 | 18:05 | |
*** tosky_ is now known as tosky | 18:05 | |
*** sshnaidm|afk is now known as sshnaidm | 18:39 | |
rpittau | good night! o/ | 18:45 |
*** tosky_ is now known as tosky | 18:48 | |
stevebaker[m] | morning | 19:32 |
stevebaker[m] | TheJulia: I see you're still having A Time | 19:32 |
stevebaker[m] | New baremetal SIG video is live! https://www.youtube.com/watch?v=wy7T8WuYDMY | 22:11 |
stevebaker[m] | Once again I had to find a new video editor when the current one broke after a fedora upgrade | 22:11 |
stevebaker[m] | So far its been openshot -> flowblade -> kdenlive | 22:12 |
stevebaker[m] | anyway, arne_wiebalck ^^ | 22:12 |
*** tosky is now known as Guest6070 | 22:42 | |
*** tosky_ is now known as tosky | 22:42 | |
*** tosky is now known as Guest6073 | 23:07 | |
*** tosky_ is now known as tosky | 23:07 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!