| opendevreview | Merged opendev/system-config master: Add OpenMetal to cloud-launcher https://review.opendev.org/c/opendev/system-config/+/921804 | 00:35 |
|---|---|---|
| fungi | Build succeeded (deploy pipeline). | 00:48 |
| fungi | oh, that was the other change, this hasn | 00:49 |
| fungi | 't deployed yet | 00:49 |
| fungi | deployment reported 5 minutes ago | 01:10 |
| Clark[m] | Looks like it succeeded so in theory we're good to start using the cloud now | 01:38 |
| opendevreview | Riccardo Pittau proposed openstack/diskimage-builder master: Disabling CentOS Stream 8 jobs https://review.opendev.org/c/openstack/diskimage-builder/+/921823 | 07:13 |
| rpittau | mmm probably need to disable the functional tests too | 07:14 |
| opendevreview | Riccardo Pittau proposed openstack/diskimage-builder master: Disabling CentOS Stream 8 jobs https://review.opendev.org/c/openstack/diskimage-builder/+/921823 | 07:16 |
| corvus | does anyone else's "modify attention set" field in the gerrit reply dialog (if you expand it) look like this? https://imgur.com/screenshot-oY7pAdI | 14:40 |
| fungi | corvus: i'm not super familiar with the gerrit webui myself, but if i use the global reply button on a change and then edit the attention set, that's what it looks like to me too | 14:45 |
| fungi | probably missing an icon there? | 14:45 |
| fungi | or some sort of rendering problem anyway, like a label bleeding into the content | 14:46 |
| fungi | looks like clicking it still does what it should, so just a cosmetic problem i guess | 14:46 |
| fungi | i'm guessing that the barred double-headed vertical arrow icon is all that's supposed to be shown, _*_UP is an internal identifier maybe, and COLLAPSE is supposed to be a hover tooltip? | 14:47 |
| corvus | yeah... i wonder if it's something with the fonts in our build? | 14:51 |
| fungi | oh, could it be related to the change to disable loading remote webfonts? | 14:52 |
| corvus | i think "Collapse" is supposed to be there as text, but _*_UP is supposed to be an icon "expand_circle_up" which... well, that is what is displayed now that i look at it | 14:57 |
| corvus | https://imgur.com/bx3okUc.png is what it is on gerrit's gerrit | 14:59 |
| corvus | the html looks the same, so i'm guessing something is missing or incorrect in the js bundle or font or something | 14:59 |
| clarkb | we didn't land any changes to modify fonts | 15:04 |
| clarkb | I decided to defer that as it wasn't a regression | 15:04 |
| clarkb | it could be a bug in 3.9 builds that is not present in master builds | 15:04 |
| corvus | ++ might just be worth keeping an eye out for the next time we build something | 15:05 |
| fungi | heading out to grab a quick lunch since the weather's nice enough that the tourists will be on the beach rather than mobbing the restaurants. bbiab | 15:31 |
| clarkb | ha thats a fun strategy | 15:34 |
| frickler | tonyb: my understanding of the status in https://bugs.launchpad.net/ubuntu/+source/openafs/+bug/2064628 is that it is waiting for someone to verify that the bug is actually fixed on noble. so if you could actually test that on some node by installing the pkgs from -proposed that would be great to help make progress | 15:43 |
| tonyb | frickler: Okay that's good to know. I'll look at that. | 16:01 |
| opendevreview | Tony Breeds proposed opendev/system-config master: [DNM] Switch mailman role to use docker compose v2 https://review.opendev.org/c/opendev/system-config/+/921764 | 16:33 |
| opendevreview | James E. Blair proposed opendev/git-review master: Stop setting a default topic on new changes https://review.opendev.org/c/opendev/git-review/+/921879 | 16:52 |
| fungi | i want to say someone (maybe me?) already has a change up for that | 16:53 |
| corvus | hah, i did look! | 16:53 |
| fungi | no, i guess i never did push one | 16:53 |
| corvus | i would be happy to abandon and review if there is | 16:53 |
| fungi | i think we just talked about it a lot a few years ago and it seemed like people weren't thrilled about breaking that feature | 16:53 |
| fungi | but i'm definitely in favor | 16:54 |
| corvus | anyway, i figure we should think about getting that in the pipeline, cause i think we might be sort of behind the times in terms of what gerrit expects now (now that they've changed the meaning of topic) | 16:54 |
| fungi | agreed. i guess the last time we discussed it heavily was around https://review.opendev.org/c/opendev/git-review/+/610573 | 16:56 |
| fungi | and then https://review.opendev.org/c/opendev/git-review/+/697448 brought us a little closer still | 16:57 |
| fungi | yeah, i commented on that one that it gave us a clearer path toward deprecating autotopic, but i guess we never followed up to do that | 16:58 |
| corvus | yeah, this feels like the last step in a series :) | 16:58 |
| fungi | https://review.opendev.org/q/project:opendev/git-review+message:topic shows a lot of the struggle around that feature, tbh | 17:00 |
| clarkb | I'm not sure if anyone else has this problem, but despite buying it by the case back when fry's still existed (and honestly long before they closed) its always a scavenger hunt to find a can of compressed air/co2/whatever goes into those cans | 17:00 |
| corvus | the only thing i didn't do in that change is let people turn autotopic back on. but at this point, if people are really attached to that, i think we should have a conversation about learning to love life without topics because of the way gerrit has changed their meaning. :) | 17:01 |
| fungi | some years back i bought a handheld "gun" that blows co2 through a fine nozzle, takes standard co2 cartridges | 17:01 |
| corvus | clarkb: i just went to costco for the first time in like 3 years and was very disappointed they had no compressed air! | 17:01 |
| corvus | what is even the point of costco without compressed air? | 17:02 |
| clarkb | ha | 17:02 |
| fungi | a box full of those cartridges takes a lot less room on the shelf than aerosol cans of "compressed air" | 17:02 |
| clarkb | fungi: neat, I have been keeping those little co2 cartridges around for emergency bike tire inflation. Maybe I should double up their utility | 17:02 |
| corvus | but yes, i too am on my last can. i guess i should look for fungi's thing. | 17:02 |
| fungi | i have no idea what it was called, but i'm sure with the right keywords a search will turn up lots of options | 17:03 |
| corvus | also i live in a small house with small closets, and the space taken up by the cans of compressed air is already in use by something else, so i guess i really only have space for the small co2 cartridges now. :) | 17:03 |
| clarkb | the naive google results return bb guns that operate on those catridges | 17:03 |
| corvus | fungi: yes, i will go google for air guns, wcpgw | 17:03 |
| fungi | it's super compact, barely larger than a co2 cartridge plus a trigger/valve and a stem for directing the blast where you want | 17:03 |
| fungi | as you can probably imagine, it doesn't need to be any more complicated than that | 17:04 |
| fungi | the only downside to the one i have is that it's a bit leaky, so if you leave it sitting around for 6mos with a full cartridge, it will depressurize and you'll need to refill | 17:04 |
| corvus | something like ? https://www.amazon.com/Gallo-Drain-Unthreaded-Cartridges-Condensate/dp/B07SVH5B78 | 17:05 |
| fungi | i also use the same cartridges in "humane" insta-kill rat traps | 17:05 |
| fungi | that looks fancier than the one i have, but yes basically | 17:05 |
| fungi | you might want something with a finer nozzle (or you might not, depends on your application) | 17:06 |
| clarkb | ac condenstate lines beware | 17:06 |
| fungi | the one i have has a thistle tube about the diameter of the ones you get with a can of wd-40, which is great for getting down inside keyboards | 17:07 |
| clarkb | when we had out heat pump installed I had them redo how our condensate line works and even though we had the coldest temps in a decade this last winter it didn't freeze over. I was very happy | 17:07 |
| corvus | https://www.americanrecorder.com/collections/gas-dusters | 17:08 |
| corvus | there we go | 17:08 |
| fungi | yep, that looks almost identical to what i'm using | 17:09 |
| corvus | can also stock up on reel-to-reel leader and splicing tape | 17:09 |
| fungi | how's that drive-in theater restoration project coming along? ;) | 17:10 |
| corvus | ooh also this has less poison than canned air | 17:10 |
| fungi | oh, never mind, that's magtape r2r splicing | 17:10 |
| fungi | so classic recording studio restoration project | 17:11 |
| fungi | corvus: yes, and the co2 cartridges should be more easily recyclable, in theory | 17:11 |
| corvus | i, um, did just repair my turntable with some ebay parts... so... i guess you got me there. :) | 17:11 |
| fungi | but the fact that there's no propellant is great | 17:11 |
| fungi | unless you like to huff aerosol propellant as a way to forget your troubles | 17:12 |
| fungi | but you can always switch to magic markers and rubber cement | 17:12 |
| corvus | i picked a bad week to stop sniffing canned air | 17:13 |
| fungi | roger, roger | 17:13 |
| clarkb | that site has some neat stuff in their catalog | 17:14 |
| clarkb | need random thread adaptors? they have you covered. Want to fill your car with nitrogen/argon for track day? covered. Want to set up a theater sounds system? covered | 17:15 |
| fungi | industrial supply catalogs are a weakness for me | 17:15 |
| clarkb | I wonder if I just found an apt-get bug. It either attempts to connect to ipv6 and then falls back to ipv4 on a host without global ipv6 addrs or its logging the ipv6 addr instead of the ipv4 addr when it connects to the mirror. I would have to strace it or similar to be more sure of what is happening there. But its definitely logging that it is connectiong to an ipv6 address on a | 17:21 |
| clarkb | host with no global inet6 address to route out via | 17:21 |
| clarkb | this is on my jammy home fileserver | 17:21 |
| tonyb | Yesterday I built a new CentOS-9-stream image after verifying that we have the newest kernel in our mirrors but it still got the older kernel | 17:25 |
| tonyb | https://nb02.opendev.org/centos-9-stream-149d2456452e49cab924ff85e4d178d4.log | 17:25 |
| tonyb | tony@thor:~$ (cd /afs/openstack.org/mirror/centos-stream/9-stream/ ; ls -l */x86_64/os/Packages/kernel-{,headers-}5.14.0-45*) | 17:25 |
| tonyb | -rw-r--r--. 1 10022 root 2681821 May 23 01:26 AppStream/x86_64/os/Packages/kernel-headers-5.14.0-452.el9.x86_64.rpm | 17:25 |
| tonyb | -rw-r--r--. 1 10022 root 2778445 Jun 4 19:23 AppStream/x86_64/os/Packages/kernel-headers-5.14.0-457.el9.x86_64.rpm | 17:25 |
| tonyb | -rw-r--r--. 1 10022 root 914741 May 23 01:25 BaseOS/x86_64/os/Packages/kernel-5.14.0-452.el9.x86_64.rpm | 17:25 |
| tonyb | -rw-r--r--. 1 10022 root 1012537 Jun 4 19:23 BaseOS/x86_64/os/Packages/kernel-5.14.0-457.el9.x86_64.rpm | 17:25 |
| tonyb | (also yay for kAFS on my laptop) | 17:26 |
| tonyb | the I guess I could verify the repodata | 17:27 |
| clarkb | ya could it be an index issue not pointing to the new file? | 17:29 |
| fungi | also possible dib isn't actually using our mirror | 17:29 |
| clarkb | fungi: oh it isn't? | 17:29 |
| fungi | i don't know, just thinking of possibilities | 17:29 |
| tonyb | Okay I'll poke more | 17:29 |
| clarkb | oh sorry you said its possible and I skipped right over that somehow | 17:30 |
| fungi | yeah, make sure it's actually hitting the mirror, and that the index it's pulling actually refers to the kernel package version we expect | 17:30 |
| jrosser | clarkb: I think I saw that same apt-get ipv6 behaviour also today | 17:55 |
| clarkb | jrosser: neat | 17:56 |
| fungi | it's somewhat sad to think about having actually worked on ipv6 routing for networks over two decades ago and still not being able to assume you'll have it on a new network | 18:06 |
| clarkb | my isp finally started offering it but only to people on their 10gbe plans so far. Sounds like it should roll out globally for the rest of their customers eventually though | 18:07 |
| clarkb | and I'm too lazy to run a tunnel when that means coming up with routing rules for netflix et al as they don't like tunnels | 18:08 |
| fungi | yeah, i had a v6 tunnel from my firewall to hurricane electric for ages, and i maintained custom routing carve-outs for things that started balking at tunnel endpoints. but finally my broadband provider added it (and dhcp6 with prefix delegation!), so i'm pretty happy now. it's even reasonably stable these days | 18:19 |
| fungi | my firewall can request additional global network assignments and then dynamically assign them to internal interfaces | 18:19 |
| fungi | so i get network separation and global routing for things on different vlans in my home network | 18:20 |
| opendevreview | Jay Faulkner proposed zuul/zuul-jobs master: Block broken version of filelock from installing https://review.opendev.org/c/zuul/zuul-jobs/+/921885 | 18:39 |
| tonyb | Any objection to me running: # sudo /usr/launcher-venv/bin/launch-node --cloud opendevzuul-openmetal --region IAD3 --az reserved --flavor opendev-mirror --image 'Ubuntu 22.04 (Jammy)' mirror01.iad3.openmetal.opendev.org | 21:04 |
| tonyb | to make us a mirror node in the new cloud? | 21:04 |
| clarkb | tonyb: none from me. I was going to try and track down how to use the host aggregate and I guess we just schedule to the reserved az for that? that is easy. Note I uploaded a very recent ubuntu jammy image that you may want to use instead (it has opendev in the image name) | 21:05 |
| fungi | command lgtm, though i also don't know the extent to which availability zones influence host aggregate scheduling decisions | 21:07 |
| fungi | makes sense though | 21:07 |
| tonyb | clarkb: That's me guess WRT to the reserved AZ | 21:07 |
| clarkb | ya I think it is a correct guess | 21:08 |
| clarkb | I was just worried we would need to do something more complicated than that | 21:08 |
| tonyb | that az is limited to a single host BUT that host is also listed in the unnamed default az | 21:08 |
| tonyb | so I think we may want to keep an eye on the placement of nodes "just in case". | 21:09 |
| clarkb | ack | 21:09 |
| fungi | indeed, it was a problem on the previous iteration of that cloud | 21:09 |
| tonyb | clarkb: how much difference is the in the OpenDev vs Ubuntu Jammy images? | 21:09 |
| clarkb | tonyb: the image I uploaded is from liek a week ago and is using the kvm optimized kernels. I haven't booted the cloud supplied image to know if theirs is also kvm optimized. But I expect it to be a bit older (we'll update either way as part of the luanch node process) | 21:10 |
| tonyb | Okay Booting that one | 21:14 |
| opendevreview | Tony Breeds proposed opendev/zone-opendev.org master: Add new openmetal mirror https://review.opendev.org/c/opendev/zone-opendev.org/+/921895 | 21:25 |
| tonyb | Gah I should have added the LE stuff for those records | 21:35 |
| tonyb | I confused myself with the LE for the openmetal VIP | 21:36 |
| tonyb | I'll finish that off tomorrow | 21:36 |
| opendevreview | Tony Breeds proposed opendev/zone-opendev.org master: Add new openmetal mirror https://review.opendev.org/c/opendev/zone-opendev.org/+/921895 | 21:40 |
| opendevreview | Tony Breeds proposed opendev/system-config master: Add openmetal mirror node to the inventory https://review.opendev.org/c/opendev/system-config/+/921896 | 21:42 |
| corvus | that virtualenv lock thing is (unsurprisingly) having some wider fallout. here's a zuul image build job that failed due to it: https://zuul.opendev.org/t/zuul/build/f4fb2ccc7878406ab2ceddac1c89af90 | 23:42 |
| Clark[m] | They pulled the broken release so rerunning stuff should resolve it. | 23:43 |
| corvus | cool, i will recheck the huge stack then :) | 23:43 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
