rpittau | good morning ironic! o/ | 07:18 |
---|---|---|
rpittau | happy friday! | 07:18 |
masghar | Good morning and happy friday! | 09:21 |
JayF | o/ | 10:17 |
JayF | rpittau: do you perhaps know the answer to the question I had from yesterday? What the gap is between in-tree and ironic-inspector (service) inspection? | 10:43 |
iurygregory | good morning ironic o/ | 10:47 |
JayF | o/ | 10:50 |
JayF | TheJulia: I don't know how long it's been, but I know some folks who used the IPMI send raw stuff, unsure if it was via the step or the api though | 10:58 |
opendevreview | cid proposed openstack/ironic master: Provision ARM (aarch64) fake-bare-metal-vms https://review.opendev.org/c/openstack/ironic/+/915441 | 11:15 |
masghar | :JayF The main gap is inspection rules, as far as I know | 11:33 |
masghar | JayF: | 11:33 |
masghar | How does the name tagging work in IRC =D | 11:33 |
JayF | if you just say the name, regardless of prefix/suffix, it will ping most clients | 11:35 |
JayF | masghar: I'm slightly embarrased to admit I don't know exactly what those are for :/ but I'll note it | 11:35 |
masghar | I see! Theyre not used a lot I think, but they are used in production so are important enough to migrate AFAIK | 11:37 |
JayF | I don't have context for "used in production", I assume you mean a downstream RH product? | 11:38 |
masghar | They're used at CERN, I heard | 11:38 |
JayF | well I had better be sure to mention it then :D | 11:40 |
JayF | thanks! | 11:40 |
masghar | No problem =D | 11:40 |
sylvr | Hello! since yesterday I found out that my idrac aren't compatible with redfish or idrac-wsman and can't be upgraded, so I configured kayobe so that ironic/bifrost use ipmi driver but my nodes seem not to be discovered/enrolled | 12:36 |
opendevreview | cid proposed openstack/ironic master: Provision ARM (aarch64) fake-bare-metal-vms https://review.opendev.org/c/openstack/ironic/+/915441 | 13:09 |
JayF | sylvr: I wonder if you need to explicitly enable IPMI | 13:15 |
JayF | Have we ever considered making a plugin-mechanism, ala hardware managers (but less dynamic) for OOB drivers? | 13:20 |
JayF | I have a use case for a custom verify step, but also I suspect this could be generally useful too | 13:20 |
TheJulia | I'm not sure I understand what your proposing | 13:29 |
TheJulia | it migh tbe the only slightly consumed cup of coffee | 13:29 |
JayF | I can pip install downstream-ironic-steps beside ironic | 13:29 |
JayF | and have a stevedore extension-like mechanism that will pull them into indicated drivers | 13:30 |
JayF | basically the equivalent of subclassing a driver and adding methods, except in a less fork-y way | 13:30 |
TheJulia | no, we haven't thought of doing that | 13:30 |
TheJulia | at least, that way | 13:30 |
JayF | If you had the requirement to run a custom verify step on a node at enrollment team, *and* had a strong anti-forking-openstack culture, you can't solve that problem today. | 13:31 |
JayF | (but it's dirt simple in IPA) | 13:32 |
JayF | I've even seen, in the wild, IPA hardware managers that would run commands to change BMC settings in-band because it wasn't something we shipped | 13:32 |
JayF | I am just trying to make sure: 1) it's not a unique use case and 2) the shape of it I'm thinking of seems sane | 13:34 |
JayF | if I went down this path I'd spec it and implement it | 13:34 |
sylvr | JayF: I did that too, maybe it helped a bit... gonna continue down that path | 13:40 |
JayF | yeah, I'd suggest if you have access -- just get a working ipmitool command first, it'll be faster cycles | 13:41 |
JayF | then you can configure ironic based on that | 13:41 |
JayF | (and if you have that working command and still can't figure out the config, we'll be able to give you pretty targeted help) | 13:41 |
TheJulia | JayF: so changing bmc settings up front is not rally a unique need, but the what needs to he changed might be | 13:44 |
JayF | exactly | 13:44 |
JayF | consider a case where I have a device, like a DPU, which needs a mode change before Ironic can manage via it | 13:44 |
TheJulia | yeah, that is definitely a thing | 13:45 |
shajizad | iurygregory: Hello Iury! I was running unit tests on openstacksdk as mentioned in this weeks tasks and 2 tests failed. Is this normal or does that mean there's something wrong with my environment? | 13:51 |
shajizad | The logs are 22000 lines long so I can't really find the error | 13:51 |
iurygregory | hey shajizad o/ | 13:53 |
iurygregory | ok, let me test things here, what python version are you using? | 13:53 |
rpittau | 22000 lines sounds like a lot! | 13:54 |
rpittau | but there should be a recap at the end of the run with the failed tests | 13:54 |
shajizad | iurygregory: Python 3.12.3 | 13:54 |
shajizad | rpittau: Its mostly filled with deprecation warnings | 13:54 |
shajizad | rpittau: Yes, here is what it says: | 13:55 |
rpittau | shajizad: version 3.12 is not supported yet, you should use 3.11 | 13:55 |
iurygregory | jesus, quite a lot of deprecation warnings for sure... | 13:55 |
shajizad | ======... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/KummCDLtqjxchigCYaEVZZxO>) | 13:55 |
rpittau | yep, with py3.12 it will be a wall of deprecations :) | 13:55 |
shajizad | rpittau: Alright, I will try that. Thanks! | 13:56 |
iurygregory | yeah, with 3.12 it fails .. I thought we had support for 3.12 already... | 13:56 |
rpittau | should appear during this cycle at some point | 13:56 |
shajizad | /home/salman/Desktop/Coding/openstacksdk/.tox/py3/lib/python3.12/site-packages/keystoneauth1/_utils.py:72: DeprecationWarning: datetime.datetime.utcnow() is deprecated and scheduled for removal in a future version. Use timezone-aware objects to represent datetimes in UTC: datetime.datetime.now(datetime.UTC). | 13:56 |
shajizad | now = datetime.datetime.utcnow() | 13:56 |
shajizad | This was basically half of the log file | 13:56 |
shajizad | Deprecation warning about datetime | 13:56 |
rpittau | shajizad: yep, that's normal, you won't see that with py3.11 | 13:57 |
iurygregory | you can also try python 3.9, 3.10 =) they should also work | 14:01 |
iurygregory | in case you want to share things that are long you can use the paste service https://paste.opendev.org/ | 14:02 |
shajizad | Alright, thank you | 14:02 |
shajizad | by the way was this the right channel for this question or should I have messaged in #openstack-sdks | 14:03 |
opendevreview | Julia Kreger proposed openstack/ironic master: docs: augment security docs https://review.opendev.org/c/openstack/ironic/+/920705 | 14:04 |
opendevreview | Julia Kreger proposed openstack/ironic master: Enhance security notes in config description https://review.opendev.org/c/openstack/ironic/+/920706 | 14:04 |
iurygregory | shajizad, it would be in the #openstack-sdks indeed, but no problem asking here | 14:06 |
opendevreview | Julia Kreger proposed openstack/ironic master: Follow-up: minor wording change from 920706 https://review.opendev.org/c/openstack/ironic/+/920893 | 14:07 |
TheJulia | $ grep -ir "utcnow()" ./ironic|wc -l | 14:08 |
TheJulia | 44 | 14:08 |
TheJulia | sadness | 14:08 |
rpittau | heh deprecation change incoming I guess | 14:09 |
iurygregory | yeah | 14:09 |
iurygregory | low-hanging-fruit maybe :D | 14:09 |
TheJulia | on a plus or awful side, it is mostly db layer stuffs | 14:09 |
JayF | are you filing the bug or shall I | 14:09 |
TheJulia | I'm shifting gears back to my audit spreadsheet | 14:10 |
TheJulia | so if you wouldn't mind, that would be awesome | 14:10 |
TheJulia | if not, I can do so later today | 14:10 |
JayF | I'll do it r/n | 14:16 |
JayF | https://bugs.launchpad.net/ironic/+bug/2067740 triaged it as medium | 14:19 |
JayF | cid: you were asking me to point you at quick wins ^ that's one | 14:19 |
shajizad | All tests passed with Python 3.11.8, thanks! | 14:24 |
iurygregory | shajizad, awesome \o/ | 14:24 |
JayF | shajizad: thanks for saving us some embarassment, I'm unsure we would've noticed that in ironic too | 14:26 |
JayF | shajizad: so the good vibes go both ways \o/ | 14:26 |
cid | JayF: Added to my todo | 14:33 |
JayF | cid: it'd be a good idea to go through all the repos, like you did for the codespell fixes too -- I don't think there'll be a lot of them | 14:34 |
JayF | cid: and if you're not aware. TZ-aware datetimes and non-TZ-aware datetimes in python are different kind of objects, so you might have to change some surrounding code (or remove the tzdata immediately after getting the TZ-aware datetime) | 14:35 |
JayF | python datetime library WHEEEE | 14:35 |
TheJulia | As the great [spock|teal'c] once said: indeed | 14:35 |
cid | TIL | 14:36 |
cid | how do I get this indentation effect in IRC https://usercontent.irccloud-cdn.com/file/oSpsNPCO/how%20do%20I%20get%20this%20indentation%20effect%20in%20IRC | 14:44 |
clarkb | cid: you use /me something something | 14:44 |
TheJulia | yes.... | 14:44 |
* TheJulia is surprised | 14:44 | |
* TheJulia is also pleased | 14:45 | |
cid | Ermm | 14:45 |
* cid This it? | 14:45 | |
clarkb | yup | 14:45 |
* TheJulia plots taking over the word for... 0.231 seconds before resuming on spreadsheet | 14:45 | |
cid | Tada :D | 14:45 |
cid | And you can make it show someone else too | 14:46 |
TheJulia | yes, typically they are used for emotives, thoughts, ora ctions | 14:46 |
TheJulia | or actions | 14:46 |
* TheJulia cries | 14:46 | |
* TheJulia cannot type | 14:46 | |
* TheJulia resumes her spreadsheet of doom | 14:46 | |
TheJulia | It is a bit different than as if your speaking outloud, hopefully that makes sense | 14:47 |
cid | What command can I use for that? Like instead of my name showing in the indented block | 14:47 |
TheJulia | There isn't any other way. Either it is sort of like speech with the server sending "<nickname>" to the client as the start of the line, or the server sends "* nickname ..." | 14:48 |
cid | this https://usercontent.irccloud-cdn.com/file/vwPGZOTQ/this.PNG | 14:48 |
cid | ^^ I meant that effect? Or isn't the two messages from the same user? :D | 14:49 |
TheJulia | no, they are entirely separate users | 14:49 |
cid | Ooh ok. | 14:50 |
* cid Goes back to vscode | 14:50 | |
cid | :D | 14:50 |
TheJulia | For bedtime reading: https://datatracker.ietf.org/doc/html/rfc1459 | 14:51 |
TheJulia | That is the original request for comments document defining IRC | 14:51 |
TheJulia | actually, not original | 14:52 |
TheJulia | that is 1993... | 14:52 |
TheJulia | I guess it was more a documentation of what had been created | 14:52 |
TheJulia | Ahh, the abstract covers that | 14:52 |
* TheJulia is old | 14:52 | |
rpittau | mmm I still remember BBS, like using them | 14:52 |
rpittau | /me is old too | 14:52 |
dmellado | and this is how ironic became retro-computing channel xD Does any of you recall ICQ? xD | 14:59 |
TheJulia | yes, getting shut down in like 1-2 weeks? | 14:59 |
dmellado | oh, still alive? :D | 14:59 |
TheJulia | yup, but not for long | 14:59 |
dmellado | https://www.theverge.com/2024/5/25/24164579/icq-shut-down-june | 14:59 |
dmellado | just saw it, lol | 14:59 |
TheJulia | "uh oh!" | 15:00 |
dmellado | My father used internet at home for his work, and I do recall when they installed ISDN. It was like a... so you can use both internet and call, and at what speeds! xD | 15:04 |
TheJulia | The incumbant telephone company where I grew up basically didn't want to run isdn services, so we had a dedicated line for dialup for quite a long time | 15:05 |
TheJulia | I was impressed when our ISP let us know were connected like 15 minutes shy of the whole month once | 15:06 |
TheJulia | I will say, our level of input validation is actually kind of impressive | 15:08 |
dmellado | xD | 15:08 |
rpittau | I used ICQ extensively back in the days | 15:09 |
rpittau | good times | 15:09 |
dmellado | and these days my 4yo asked me about why the icon of a phone looked like that | 15:10 |
TheJulia | :) | 15:10 |
dmellado | I guess that'd apply to more icons later, like the 'save' one xD | 15:11 |
TheJulia | heh | 15:11 |
* TheJulia tries to remember she last used a floppy disk | 15:12 | |
* TheJulia fails | 15:12 | |
dmellado | I do have a couple of them around at my parent's place but I do wonder if they still work... felt sad to throw away 'prince of persia' and 'the lemmings' xD | 15:13 |
* TheJulia remembers lemmings | 15:15 | |
* TheJulia wonders if pingus is still maintained | 15:15 | |
rpittau | I still have a DOS distribution in floppy disks, plus way too many games :D | 15:16 |
rpittau | I doubt they still work though, especially the 5 1/4 inch | 15:16 |
TheJulia | Magnetic media, in theory, has a lifetime of like 20+ years | 15:17 |
TheJulia | .... | 15:17 |
TheJulia | Optical.... on the other hand | 15:17 |
TheJulia | 8-10 years, I think. | 15:17 |
TheJulia | Which means my wife's giant DVD collection might start to not work so well | 15:17 |
clarkb | I have a usb floppy drive somewhere | 15:18 |
clarkb | I bought it maybe 10 years ago and even then I felt like this was silly | 15:19 |
rpittau | so I just need to find a 5 1/4 inch floppy disk reader and try to install Monkey Island again :) | 15:19 |
TheJulia | I feel like I suddenly need pingus but instead of penguins, I need little fallout guys | 15:22 |
dmellado | rpittau now I do wonder if there's any usb 5 1/4 units xD | 15:33 |
rpittau | :D | 15:37 |
JayF | dmellado: we've always been the retro computing channel ;) | 15:38 |
* JayF had a 7 digit ICQ number | 15:38 | |
TheJulia | ... | 15:41 |
TheJulia | JayF: I just had a flashback to when we last discussed ICQ number length | 15:41 |
TheJulia | for the audience, it was a *very* long time ago | 15:41 |
* JayF falls into the chasm that is his missing long term memory | 15:41 | |
JayF | although I can guess | 15:42 |
TheJulia | At that old grumpy marine's place in the woods | 15:42 |
JayF | I remember that, not the conversation though | 15:42 |
TheJulia | Speaking of which, I might be going to NC in July and visiting him | 15:42 |
JayF | nice! | 15:43 |
JayF | Can folks *please* prioritize spec review? TheJulia's networking spec has one +2 from me and zero other interactions from Ironic cores, and my runbooks spec has idled for almost two weeks with no feedback | 15:47 |
JayF | https://review.opendev.org/c/openstack/ironic-specs/+/890164 https://review.opendev.org/c/openstack/ironic-specs/+/916126 | 15:48 |
rpittau | JayF: I have both open since I came back on Monday but really couldn't find the time this week to have a proper look, I'll try next week, worst case scenario I'll review both on a train :) | 15:52 |
rpittau | need to split now, have a great weekend everyone! o/ | 15:52 |
JayF | honestly even just someone commenting "this looks generally okay" would be nice :D | 15:52 |
JayF | but yeah, train is where I'm doing some stuff too :D | 15:52 |
TheJulia | train++ | 15:53 |
JayF | "this high speed train, number 2.7, is now deprecated" | 15:55 |
* TheJulia notes the real number is not an integer | 15:55 | |
JayF | that's not a number at all, it's a string! (#VersionsAreNotNumbers #ItsAnOpenStackTrainJoke) lol | 15:59 |
JayF | oh that's unfortunate that my joking "hashtags" link in irccloud :D | 15:59 |
-opendevstatus- NOTICE: Gerrit on review.opendev.org is being upgraded to version 3.9 and will be offline. We have allocated an hour for the outage window lasting until 1700 UTC | 16:01 | |
TheJulia | .... sigh, a little more heads up might have been nice | 16:01 |
TheJulia | but different focuses | 16:01 |
clarkb | TheJulia: this was announced to service-announce some time ago | 16:02 |
TheJulia | I was more thinking, something in line with human awareness in present time | 16:02 |
clarkb | we continue to encourage people to subscribe to that list if they haven't already. it is extremely low volume | 16:02 |
TheJulia | "going down an hour" just allows better time planning when deep in the weeds | 16:03 |
* TheJulia goes back to code audit | 16:03 | |
JayF | I got a notification popup in gerrit | 16:04 |
JayF | notification #503 | 16:05 |
JayF | lol | 16:05 |
TheJulia | heh | 16:05 |
JayF | unfortunate but I got all but one of my open tabs review-voted before it hit, so not too bad | 16:05 |
TheJulia | anyone got ironic running locally at the moment? | 16:13 |
cid | During the outage, can code uploads still happen? | 16:14 |
TheJulia | cid: no | 16:14 |
cid | Ah | 16:14 |
cid | TheJulia: This time or always | 16:15 |
TheJulia | When gerrit is down, there is no ability to change/update/post a change set | 16:16 |
cid | Got it | 16:16 |
* cid Wonders if that was a stupid question :D | 16:17 | |
TheJulia | not at all | 16:17 |
JayF | The only stupid questions are the ones left unasked because someone thought they were stupid ;) | 16:25 |
TheJulia | ++ | 16:26 |
JayF | as an aside: I put on my todo list to document node history because I couldn't find it | 16:26 |
* TheJulia goes and finds migrane meds before moving on to the next columns of this spreadsheet of doom | 16:27 | |
JayF | also that docs audit, I read the preliminary report today and commented on it; I'd expect the public version of it to be ready soon | 16:27 |
TheJulia | cool cool | 16:29 |
opendevreview | Merged openstack/ironic master: Enhance security notes in config description https://review.opendev.org/c/openstack/ironic/+/920706 | 16:42 |
TheJulia | looks like gerrit is back | 16:42 |
clarkb | yup we've gone through our checks at this point and things look good (zuul is doing what we expect, the web is responsive, you can push changes etc) | 16:44 |
JayF | thanks for the upgrade! | 16:48 |
JayF | gerrit go wrrr | 16:48 |
TheJulia | beep boop | 17:05 |
shajizad | Hello, I was using bitfrost on a VM and during the deployment stage it failed and gave me an error with something to do with public keys. So I just blindy did generated keys with ssh-keygen and it worked. I'm just a bit unsure about why it needed it? | 17:06 |
TheJulia | bifrost is modeled around the concept of you'll use ansible playbooks to execute additional steps on the hosts | 17:11 |
TheJulia | so, to login... the keys are needed | 17:11 |
TheJulia | or, a key, to be more precise | 17:12 |
JayF | TheJulia: I have a local ironic devstack | 17:18 |
JayF | TheJulia: you want me to put your key on it? | 17:18 |
JayF | s/local/in AWS/ | 17:18 |
JayF | I missed the message earlier or I woulda offered it earlier | 17:18 |
TheJulia | eh, I if you turn off auth and just curl the api for /v1/nodes/<valid_id_here>/xyz and see what you get, it will satisfy my curiosity | 17:18 |
JayF | turn off auth meaning -> no client | 17:19 |
JayF | or turn off auth meaning -> ironic doesn't do authentication with keystone | 17:19 |
JayF | I don't know how to do the latter | 17:19 |
TheJulia | set [DEFAULT]auth_strategy to "noauth" and restart hte api, and just use curl | 17:19 |
JayF | aight, getting in it | 17:19 |
TheJulia | much appreciated | 17:19 |
JayF | been doing a bit of downstream demoing this week | 17:21 |
JayF | node history and notifications are a hit for operability | 17:21 |
TheJulia | cool cool | 17:24 |
JayF | the return is a json doc indicating request not acceptable | 17:25 |
TheJulia | cool | 17:25 |
JayF | and I ran with -v and confirm it's a http 406 | 17:25 |
JayF | I'm about to go offline for the weekend, this devstack is essentially disposable to me, I'm happy to put your public key on it | 17:25 |
JayF | if you might need something else | 17:26 |
TheJulia | Nah, I'm good, I just looked at a point and went "i wonder what that would do!" | 17:26 |
JayF | ack | 17:27 |
JayF | I'm out o/ | 17:27 |
TheJulia | Have a great weekend | 17:27 |
JayF | Paris, here I come | 17:28 |
* JayF will undo all his weight loss in baguettes and croissants | 17:28 | |
TheJulia | om nom nom | 17:28 |
JayF | last time I went to paris I was gluten-free, so I was tortured | 17:28 |
JayF | I'm looking forward to being there without the limitation, lol | 17:28 |
cid | o// | 18:25 |
TheJulia | JayF: https://bugs.launchpad.net/ironic/+bug/1304673 updated | 22:42 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!