| *** pmannidi is now known as pmannidi|brb | 00:36 | |
| *** pmannidi|brb is now known as pmannidi | 01:02 | |
| opendevreview | Steve Baker proposed openstack/ironic-python-agent-builder stable/wallaby: [DNM] testing dib-centos9 experimental job https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/824136 | 01:37 |
|---|---|---|
| opendevreview | Steve Baker proposed openstack/ironic-python-agent-builder stable/wallaby: [DNM] testing dib-centos9 experimental job https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/824136 | 02:00 |
| opendevreview | Steve Baker proposed openstack/ironic-python-agent-builder master: Move centos9 jobs from experimental to check https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/825233 | 02:06 |
| *** pmannidi is now known as pmannidi|brb | 02:07 | |
| *** pmannidi|brb is now known as pmannidi | 02:29 | |
| opendevreview | Steve Baker proposed openstack/ironic-python-agent-builder stable/wallaby: [DNM] testing dib-centos9 experimental job https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/824136 | 03:51 |
| opendevreview | Steve Baker proposed openstack/ironic-python-agent-builder master: Move centos9 jobs from experimental to check https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/825233 | 03:52 |
| *** pmannidi is now known as pmannidi|afk | 04:45 | |
| arne_wiebalck | Good morning, Ironic! | 07:19 |
| opendevreview | Sandeep Yadav proposed openstack/ironic-python-agent-builder stable/wallaby: Preliminary work to support CentOS 9 Stream https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/823365 | 07:30 |
| opendevreview | Sandeep Yadav proposed openstack/ironic-python-agent-builder stable/wallaby: Preliminary work to support CentOS 9 Stream https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/823365 | 07:57 |
| rpittau | good morning ironic! o/ | 08:40 |
| opendevreview | chandan kumar proposed openstack/ironic-python-agent-builder stable/wallaby: Preliminary work to support CentOS 9 Stream https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/823365 | 08:52 |
| rpittau | stevebaker[m]: about https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/825233 the sysbench rpm can be found in EPEL 8 but not 9 yet, probably need to ask for a build as I don't see any request open | 09:08 |
| rpittau | ^ this is for the "extra" job | 09:09 |
| dtantsur | good morning | 09:20 |
| rpittau | good morning dtantsur :) | 09:26 |
| rpittau | dtantsur: fix for bifrost stable/xena works https://review.opendev.org/c/openstack/bifrost/+/825141 | 09:28 |
| dtantsur | yay! | 09:29 |
| rpittau | :) | 09:29 |
| rpittau | going to cherry-pick to wallaby | 09:30 |
| opendevreview | Riccardo Pittau proposed openstack/bifrost stable/wallaby: Fix incompatibility with recent rich library https://review.opendev.org/c/openstack/bifrost/+/825212 | 09:30 |
| opendevreview | Riccardo Pittau proposed openstack/bifrost master: Add CentOS Stream 9 keystone integration job https://review.opendev.org/c/openstack/bifrost/+/824193 | 09:53 |
| holtgrewe | Good morning. Is the IPA inspector supposed to be able to run with IPv6 disabled? https://pastebin.com/w6B4JtQg | 10:50 |
| holtgrewe | It fails for me with at least some logs related to IPv6. I put ipv6.disable=1 into the kernel command line and then also need to do ipa-listen-host=0.0.0.0 otherwise it uses :: which makes it fail. | 10:50 |
| dtantsur | holtgrewe: I don't think we've ever tried it, but it should work IMO. | 10:53 |
| dtantsur | "inspection_callback_url = None" are you sure you're running inspection? | 10:54 |
| holtgrewe | I deleted the node and now need to re-enroll. | 10:54 |
| dtantsur | holtgrewe: it seems like what it's doing is not inspection, maybe cleaning. | 10:56 |
| dtantsur | are 'b0:26:28:26:5a:10,b0:26:28:26:5a:11' matching what you have in Ironic for this node? | 10:56 |
| dtantsur | any errors in ironic-api? | 10:56 |
| holtgrewe | dtantsur: These are the correct MACs of the node that I'm trying to add fresh. | 10:58 |
| holtgrewe | The node is not in ironic yet. | 10:58 |
| holtgrewe | I'm trying to run IPA now without disabling IPv6. | 10:59 |
| dtantsur | holtgrewe: could you elaborate on what exactly you're trying to do and what (as far as you can figure out) process is happening now? | 10:59 |
| dtantsur | IPA alone, started on a random machine without a corresponding node, will not do anything | 11:00 |
| holtgrewe | I had problems installing an instance on a baremetal node. So I removed the baremetal node via Horizon and now I'm trying to add it back by booting it with PXE. | 11:00 |
| holtgrewe | It picks up the IPA image and boots | 11:00 |
| holtgrewe | and ends with the logs that I pasted above. | 11:00 |
| dtantsur | holtgrewe: this mode of operation is supported, but requires additional configuration: https://docs.openstack.org/ironic-inspector/latest/user/usage.html#discovery | 11:00 |
| dtantsur | just booting IPA without configuring inspector for discovery will not achieve what you want | 11:01 |
| dtantsur | (you'll also need to tell IPA where to find inspector, currently it's not done) | 11:01 |
| holtgrewe | This worked previously for me. I'm using kayobe/kolla for deployment. https://paste.openstack.org/show/812209/ | 11:03 |
| dtantsur | holtgrewe: okay, you have the right configuration, but the PXE environment your node boots from does not have the inspector URL for callback. | 11:04 |
| dtantsur | check whatever iPXE (?) script it's loading on boot | 11:05 |
| holtgrewe | Woudl such a script be "boot.ipxe"? | 11:08 |
| holtgrewe | OK... sorry, something else urgent came up. Let me come back with this particular issue later. | 11:10 |
| dtantsur | holtgrewe: it depends, I'm not familiar with Kolla enough to say. | 11:10 |
| dtantsur | ack | 11:10 |
| dtantsur | usually with discovery another file, not boot.ipxe, is used | 11:10 |
| holtgrewe | OK, that already helps. | 11:10 |
| iurygregory | good morning Ironic o/ | 11:30 |
| dtantsur | morning iurygregory | 11:34 |
| dtantsur | arne_wiebalck: hey, has there been any resolution to the software RAID issue? | 11:35 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic master: [WIP] Explicit parameter to distinguish partition/whole-disk images https://review.opendev.org/c/openstack/ironic/+/825305 | 12:45 |
| arne_wiebalck | dtantsur: we have confirmed that more strict error handling breaks deployments (for images which do have a recent grub2-install) | 12:51 |
| arne_wiebalck | dtantsur: today, I have started to look into whether we can make it work with efibootmgr | 12:52 |
| arne_wiebalck | dtantsur: first attempt failed | 12:52 |
| arne_wiebalck | dtantsur: I keep trying and let you know | 12:52 |
| arne_wiebalck | dtantsur: first attempt == branch into uefi/efibootmgr for sw RAID as well, no more deploy error, but the node does not boot :) | 12:53 |
| iurygregory | should we try to fix https://review.opendev.org/c/openstack/sushy/+/795557 and merge in ussuri? | 13:08 |
| iurygregory | I've checked all projects and sushy is the only one with outstanding patch that we would need to merge | 13:10 |
| dtantsur | arne_wiebalck: let's maybe merge a simple hack first? I'm worried about a feature broken in master? | 13:17 |
| dtantsur | i.e. a partial revert, only with software RAID? | 13:17 |
| arne_wiebalck | dtantsur: sounds good to me | 13:18 |
| dtantsur | iurygregory: I'm undecided. It may not be worth the time. | 13:18 |
| arne_wiebalck | dtantsur: shall I give it a go? | 13:18 |
| dtantsur | arne_wiebalck: if you don't mind | 13:18 |
| iurygregory | dtantsur, yup I have the same feeling =) | 13:19 |
| arne_wiebalck | dtantsur: I can propose sth, no prob | 13:19 |
| opendevreview | Arne Wiebalck proposed openstack/ironic master: Software RAID: Continue on bootloader installation errors https://review.opendev.org/c/openstack/ironic/+/825311 | 13:43 |
| arne_wiebalck | dtantsur: ^^, let me know if this looks ok, then I go ahead and pull that into our qa env to test it on real hardware | 13:44 |
| dtantsur | arne_wiebalck: yep, but let's (once tested) add a unit test and maybe refer to the partially reverted commit in the commit message | 13:45 |
| arne_wiebalck | dtantsur: ok | 13:45 |
| opendevreview | Arne Wiebalck proposed openstack/ironic master: [WIP] Software RAID: Continue on bootloader installation errors https://review.opendev.org/c/openstack/ironic/+/825311 | 13:46 |
| arne_wiebalck | marked as WIP for now | 13:46 |
| opendevreview | Rafael Castillo proposed openstack/ironic-python-agent-builder stable/wallaby: Adds RDO deps repository for centos 9 stream https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/825312 | 13:48 |
| opendevreview | Rafael Castillo proposed openstack/ironic-python-agent-builder stable/wallaby: [DNM] testing dib-centos9 experimental job https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/824136 | 13:54 |
| opendevreview | chandan kumar proposed openstack/ironic-python-agent-builder master: Add pkg-map for sysbench https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/825313 | 14:00 |
| opendevreview | chandan kumar proposed openstack/ironic-python-agent-builder stable/xena: Add pkg-map for sysbench https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/825217 | 14:02 |
| opendevreview | chandan kumar proposed openstack/ironic-python-agent-builder stable/wallaby: Add pkg-map for sysbench https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/825218 | 14:02 |
| opendevreview | chandan kumar proposed openstack/ironic-python-agent-builder stable/wallaby: [DNM] testing dib-centos9 experimental job https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/824136 | 14:07 |
| TheJulia | is chandan kumar in the channel? | 14:08 |
| TheJulia | Good morning ironic | 14:09 |
| iurygregory | good morning TheJulia =) | 14:09 |
| iurygregory | maybe you can find chandan in the tripleo channel I would say | 14:09 |
| TheJulia | well, ideally they need to merge first to master before doing a bunch of backporting. :( | 14:10 |
| TheJulia | Rafael... well... I just -2'ed his patches. | 14:11 |
| opendevreview | Rafael Castillo proposed openstack/ironic-python-agent-builder master: Adds RDO deps repository for centos 9 stream https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/825314 | 14:12 |
| iurygregory | truth, master 1st than backport =) | 14:12 |
| opendevreview | Rafael Castillo proposed openstack/ironic-python-agent-builder stable/xena: Adds RDO deps repository for centos 9 stream https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/825219 | 14:15 |
| TheJulia | so nova has a likely good idea https://github.com/openstack/nova/blob/master/tools/check-cherry-picks.sh#L40-L43 | 14:21 |
| iurygregory | interesting! | 14:22 |
| iurygregory | TheJulia, any thoughts about https://review.opendev.org/c/openstack/sushy/+/795557 ? shoudl we try to fix the conflict and merge in ussuri and push a release for sushy ? | 14:23 |
| TheJulia | when did I backport that? | 14:23 |
| TheJulia | wow | 14:24 |
| iurygregory | Jun 09, 2021 11:27 AM | 14:24 |
| iurygregory | :D | 14:24 |
| opendevreview | Julia Kreger proposed openstack/sushy stable/ussuri: Automatically retry HTTP 5xx on GET requests https://review.opendev.org/c/openstack/sushy/+/795557 | 14:24 |
| TheJulia | iurygregory: cleanly rebased on top of master branch it seems | 14:25 |
| iurygregory | amazing! | 14:25 |
| iurygregory | I will push a patch with all releases for ussuri for our projects after we have this one merged | 14:26 |
| iurygregory | later today/tomorrow morning max | 14:26 |
| opendevreview | LiZekun proposed openstack/ironic-prometheus-exporter master: Remove unicode from code https://review.opendev.org/c/openstack/ironic-prometheus-exporter/+/825322 | 14:35 |
| arne_wiebalck | dtantsur: the patch works with real hardware | 14:39 |
| dtantsur | \o/ | 14:46 |
| dtantsur | good morning TheJulia | 14:46 |
| iurygregory | arne_wiebalck,++ | 14:48 |
| opendevreview | Arne Wiebalck proposed openstack/ironic master: Software RAID: Continue on bootloader installation errors https://review.opendev.org/c/openstack/ironic/+/825311 | 15:14 |
| arne_wiebalck | dtantsur: ^^ i referred to the "reverted" commit and added a unit test | 15:15 |
| * arne_wiebalck makes a coffee as a reward | 15:15 | |
| TheJulia | coffee++++ | 15:16 |
| * TheJulia awaits water to boil for coffee | 15:27 | |
| * arne_wiebalck hears the coffee machine is working on the last drops | 15:29 | |
| *** pmannidi is now known as pmannidi|AFK | 16:14 | |
| micosta | HI all, I'm using sushy-tools and I'm hitting an internal server error when trying to eject the virtual media iso. It seems that the file located in /tmp/.../iso_name.iso was removed which means os.unlink is throwing a FileNotFound exception and gets stuck in that state. If no one has bandwidth, I'm happy to add an except for this myself but Is there any reason this exception shouldn't just be caught? My | 16:20 |
| micosta | thinking is the desired state already exists so it should just move on. | 16:20 |
| opendevreview | Riccardo Pittau proposed openstack/ironic master: Replace mkisofs with pycdlib calls https://review.opendev.org/c/openstack/ironic/+/819121 | 16:25 |
| TheJulia | micosta: That does kind of sound reasonable to do if it was otherwise already removed | 16:38 |
| TheJulia | micosta: patches are welcome! :) | 16:38 |
| iurygregory | CI is unhappy with the patch in ussuri D: | 16:39 |
| TheJulia | CI may just be dead for ussuri | 16:40 |
| iurygregory | yeah, let's see if recheck helps | 16:47 |
| iurygregory | zuul here are the cookies | 16:47 |
| TheJulia | are the special zuul cookies, or just the cookies for santa? | 16:52 |
| iurygregory | I added some Brazilian flavour in the cookies to see if zuul will be happy LOL | 16:55 |
| iurygregory | :D | 16:55 |
| TheJulia | we really need to make a "book of baremetal recipes" | 16:56 |
| TheJulia | and include! pizza, zuul cookies, etc | 16:56 |
| iurygregory | ++ | 17:07 |
| dking | Is there documentation for the Conductor API? I'm having trouble finding it. | 17:18 |
| JayF | So, the Ironic Conductor is an internal service that's only addressed (over RPC) by the Ironic API. | 17:19 |
| TheJulia | What JayF said | 17:19 |
| JayF | There is no separate conductor API; there's just the Ironic API; which is documented here: https://docs.openstack.org/api-ref/baremetal/ | 17:19 |
| dking | For some reason, my conductor seems to be unresponsive intermittently, or doesn't seem to be receiving responses from the API. I was thinking that I would start by sending some requests to it to see if I could replicate with as few variables as possible. | 17:20 |
| TheJulia | hmm | 17:20 |
| TheJulia | dking: rabbit or jsonrpc? | 17:20 |
| JayF | I think I used to use `driver-list` as my hello-world for conductors | 17:20 |
| dking | That would be a good thing to check, but I don't believe that I am using them. It is a plain install of ironic inside a container. | 17:21 |
| TheJulia | so likely jsonrpc, how much work are you sending it? are there lots of ipmi nodes? | 17:22 |
| TheJulia | and by work, number of nodes I guess | 17:23 |
| dking | This one is actually a smaller setup with only about 50 nodes in it. | 17:25 |
| dking | They do get a request in to check all of them for updates fairly regularly, but I'm not having the same issues on a similar setup with 200 nodes. | 17:25 |
| dking | And of those, I'm not thinking that the conductor has much to do since they don't change state too much in this instance. | 17:26 |
| TheJulia | power sync has historically been a thing which can generate a lot of load | 17:30 |
| TheJulia | well, power state sync , and with ipmi that is a ton of overhead | 17:30 |
| dking | When the issue occurs, there's no log entries at all for conductor. My guess is that it's our internal network somehow, but I don't know as much about conductor, so I was trying to get my head around it. | 17:30 |
| dking | Oh, really? | 17:30 |
| dking | I hadn't thought about all the extra calls it would have to do for the IPMI request. | 17:31 |
| dking | Currently, while I'm trying to replicate, I send a bunch of the power requests looking for one to fail, so perhaps it's just me DoSing my own service. | 17:31 |
| TheJulia | possibly | 17:33 |
| TheJulia | we have tunables wrapped around that, fwiw | 17:33 |
| dking | Before, I had what appeared to be a related issue, but I discovered that had to do with the db connection. It's resolved now, but when testing to make sure it was, that was when I started to get these results. | 17:33 |
| TheJulia | but yeah, exec of the ipmitool command is kind of heavy for python and for the cpu in general | 17:33 |
| dking | I feel like we've sent over 20 at a time before with no problem, though, so I didn't expect to have trouble. But then, we also do expect to have some kinds of issues when we do larger groups, like 50 at a time. | 17:34 |
| TheJulia | weird, but yeah, load can be somewhat bursty depending on the settings which may be what your seeing | 17:34 |
| TheJulia | how many cpu cores are available to the container | 17:34 |
| dking | 48 | 17:35 |
| dking | Oh, wait, to that container, I'm not sure. | 17:36 |
| dking | So, it's only using a fraction of what the server has. | 17:36 |
| TheJulia | hmm | 17:38 |
| TheJulia | power state tunables would be my first thing to investigate | 17:39 |
| TheJulia | but that is just my default thought because it is a heavy load item. | 17:39 |
| dking | I might just go on and leave this alone since I might be causing the problem, or it might resolve itself before it's a real issue. Since there is a good chance this is unrelated to the actual problem I was having that was the DB connectivity, and it isn't as much of a problem, I might just revisit it later. | 17:45 |
| dking | Thank you for your help and knowledge with this. Sometimes another brain or set of eyes can help. | 17:46 |
| arne_wiebalck | dtantsur: TheJulia: I think there are (at least) two issues why the efibootmgr code path does not work out of the box with s/w RAID: | 17:51 |
| arne_wiebalck | 1) the NVRAM is updated with the EFI partition from the image, i.e. /dev/md126p15 inside the MD device (which the BIOS is most likely not able to read), it should use the EFI partitions on the holder devices instead; however: | 17:51 |
| arne_wiebalck | 2) the EFI partitions on the holder devices are not prepared, i.e. the bootloaders are not relocated, as this happens in _install_grub2 (which we is what we do not want to call :-) | 17:51 |
| rpittau | good night! o/ | 17:53 |
| TheJulia | arne_wiebalck: so the bios should be able to read past the 32 bit boundary.... I think. | 18:09 |
| arne_wiebalck | TheJulia: will the bios is able to find a partition on an md device which is made on top of partitions? | 18:11 |
| TheJulia | oh, on the MD device, no unlikely | 18:11 |
| arne_wiebalck | yeah | 18:11 |
| arne_wiebalck | this is why we move the loaders from the image into the prepared ESPs | 18:11 |
| arne_wiebalck | on the holder disks | 18:11 |
| TheJulia | I thought we were doing this on UEFI | 18:11 |
| arne_wiebalck | yes | 18:11 |
| arne_wiebalck | inside grub2-install | 18:11 |
| arne_wiebalck | which we do not call when we move to efibootmgr :) | 18:12 |
| arne_wiebalck | we have the building blocks | 18:12 |
| arne_wiebalck | I will try to re-use them | 18:13 |
| arne_wiebalck | I _think_ all we need to do is move the loaders and call efibootmgr for all holders/ESPs | 18:13 |
| arne_wiebalck | meanwhile there is a patch to deal with the problem at hand: https://review.opendev.org/c/openstack/ironic/+/825311 | 18:14 |
| arne_wiebalck | anyway, sth for tomorrow :) | 18:14 |
| arne_wiebalck | bye everyone o/ | 18:15 |
| TheJulia | arne_wiebalck: yes, likely need to. On a plus side, it should fail and recover without any issues should a member be replaced :) | 18:22 |
| TheJulia | just need to add the new member on the new disk | 18:22 |
| iurygregory | bye everyone o/ see you tomorrow | 18:48 |
| opendevreview | Rafael Castillo proposed openstack/ironic-python-agent-builder stable/wallaby: [DNM] testing dib-centos9 experimental job https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/824136 | 19:04 |
| opendevreview | Rafael Castillo proposed openstack/ironic-python-agent-builder stable/wallaby: [DNM] testing dib-centos9 experimental job https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/824136 | 19:09 |
| arne_wiebalck | TheJulia: Yeah, I'll see if I can try this tomorrow. (Keeping these devices long-term comes with some subtleties, in particular in the context of replacements: https://techblog.web.cern.ch/techblog/post/longterm-maintenance-of-bootable-raid-devices/ ) | 20:34 |
| opendevreview | Steve Baker proposed openstack/ironic-python-agent-builder master: Move centos9 jobs from experimental to check https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/825233 | 20:58 |
| opendevreview | Steve Baker proposed openstack/ironic-python-agent-builder master: Do not depend on element pip-and-virtualenv https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/825400 | 20:58 |
| opendevreview | Steve Baker proposed openstack/ironic-python-agent-builder master: Remove centos7 jobs, python2 support https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/825400 | 22:49 |
| opendevreview | Steve Baker proposed openstack/ironic-python-agent-builder master: Move centos9 jobs from experimental to check https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/825233 | 22:49 |
| stevebaker[m] | JayF: hey, my pip-and-virtualenv change has ... evolved https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/825400 | 22:52 |
| JayF | rloo might have opinions on that; I added her to the patch | 22:55 |
| JayF | she is one of the deployers I mentioned you might wanna talk to | 22:55 |
| stevebaker[m] | JayF, rloo : I don't mention it in the change, but the centos7 and centos7-python3 jobs were *actually* using python2, even though python2 support was removed in 2019 | 22:56 |
| rloo | my opinion is to leave it as-is because yahoo is using it and i have no time for the foreseeable future to even think about replacing its use! (thx JayF for noticing). i'll comment in the patch. | 23:07 |
| JayF | lmao | 23:07 |
| stevebaker[m] | oh, I didn't realise "python3 -m venv" installs a pip inside the venv even when no system pip is installed | 23:08 |
| JayF | I feel that in my bones | 23:08 |
| rloo | (sorry, have to go make dinner, thx for the ping!) | 23:08 |
| stevebaker[m] | rloo: the problem is adding support for centos9 is difficult because virtualenv is so deprecated now, which makes pip-and-virtualenv more fraught | 23:09 |
| JayF | I think there's room to fix both | 23:09 |
| JayF | but maybe not time to do so | 23:09 |
| JayF | I thought I had a change in DIB to add a knob that was turned at my old job | 23:11 |
| JayF | but I can't find it. | 23:11 |
| stevebaker[m] | I mean, python2 is EOL. Its a pretty big liability at this point, pip-and-virtualenv only existed to ease 2/3 issues with pip and venvs | 23:12 |
| JayF | I think your change is 100% right, I just also think it's one of those sneaky changes that can break deployers without them even realizing something had changed | 23:12 |
| JayF | to be clear; that's what my feedback was mostly about: not that we have to maintain backwards compatability with something for ~forever, but that we should make sure to be loud and deliberate about breaking it | 23:13 |
| TheJulia | Another aspect to likely consider, is even *if* right now someone (somehow) builds an image, with how things are so broken right now no python based elements such as IPA are going to work in that resulting image... at least AIUI | 23:14 |
| TheJulia | That is, at least my $0.02 | 23:14 |
| JayF | My comment was basically that the change probably needed some loud noises made about it | 23:14 |
| JayF | I am the resident expert on loud noises | 23:15 |
| JayF | so... :D | 23:15 |
| * TheJulia has flashbacks to many many years ago | 23:19 | |
| TheJulia | err, well, a singular flashback | 23:19 |
| stevebaker[m] | JayF: ah, I can totally add a release note | 23:19 |
| JayF | TheJulia: If you only have one memory of me being excessively loud, you should go see a doc | 23:20 |
| JayF | lol | 23:20 |
| TheJulia | stevebaker[m]: Yeah, that would be the most reasonable... thing to do I think | 23:20 |
| JayF | IPA-B is always weird when thinking about api surfaces | 23:20 |
| TheJulia | JayF: numerous, but I think it was the first | 23:20 |
| JayF | because it kinda is an API surface, but also not really... one of those things I suspect most folks don't think about until their modification has broken | 23:20 |
| * TheJulia wonders where her brain went | 23:21 | |
| TheJulia | One more meeting, and then I go make lasagna | 23:23 |
| JayF | I always thought a benefit of lasagna was being able to make it in advance | 23:24 |
| JayF | I'm not sure I've ever eaten a lasagna made at home that didn't at least spend an overnight in the fridge or freezer | 23:24 |
| JayF | Now I want lasagna. | 23:24 |
| TheJulia | heh | 23:24 |
| TheJulia | It tends to take me about two hours start to finish including baking | 23:25 |
| TheJulia | but don't like to cook when I'm on calls and such so... | 23:25 |
| TheJulia | holding off until the last meeting is over | 23:25 |
| TheJulia | And we do like our fresh lasagna | 23:28 |
| * TheJulia goes back to this "baremetal cookbook idea | 23:28 | |
| TheJulia | so, pizza, lasagna, cookies... what else? | 23:30 |
| TheJulia | "Recipies which will take far longer than deploying physical machines" | 23:30 |
| stevebaker[m] | bread | 23:31 |
| JayF | You can make any recipe take a long time if you provide the right amount of incompetence. | 23:31 |
| stevebaker[m] | gerrit is down noooooo | 23:31 |
| TheJulia | I guess we could do some breads | 23:31 |
| TheJulia | nooooooooo | 23:31 |
| TheJulia | JayF: or insufficent yeast | 23:32 |
| TheJulia | I tend to under-yeast my pizza crust, but I use a fancy imported pizza dough yeast because the flavor is so much better | 23:32 |
| JayF | Vanessa found a nice pizza dough recipe using an Italian gluten-free flour that is super delicious. We made a neopolitan and a grandma pie outta them and it turned out pretty OK | 23:33 |
| JayF | you can make killer gluten free detroit style pizza | 23:33 |
| TheJulia | So who is starting the google doc? | 23:34 |
| TheJulia | or shall we use etherpad? | 23:34 |
| TheJulia | to start | 23:34 |
| TheJulia | "Hi, we'd like an ironic namespace in gerrit to begin coordinating recipes for a cook book" | 23:35 |
| stevebaker[m] | heh | 23:37 |
| TheJulia | Manicotti | 23:38 |
| TheJulia | Which reminds me, next grocery order needs to go to someplace which can yield manicotti tubes | 23:39 |
| stevebaker[m] | does anyone know why the ipa centos jobs use the 'centos' element instead of 'centos-minimal'? | 23:43 |
| TheJulia | i do not | 23:43 |
| stevebaker[m] | the docs use 'centos-minimal' in 4 out of 5 examples https://docs.openstack.org/ironic-python-agent-builder/latest/admin/dib.html | 23:45 |
| TheJulia | https://tenor.com/view/whatever-shrug-love-lucy-okay-idk-gif-16447471 | 23:46 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!