[00:42] rbasak, nacc: Here you go: https://code.launchpad.net/~tsimonq2/usd-importer/+git/add-lubuntu/+merge/345799 [04:21] realtek wireless drivers suck [04:21] i just patched the 8822bu driver to work with 4.15+ [09:17] when is popcon.ubuntu.com going to be fixed? [09:18] Did someone file an issue with rt? I've done that in the past to get it fixed. [09:18] where's the rt? i'll file it [09:18] TheMaster: past? being > 1 year? [09:19] Wed Jun 22 15:09:05 2016 UTC last update [09:19] 2 years almost [09:19] RETARDED, there's bugs for it on launchpad.net [09:19] Yeah I quit filing them after it kept on breaking, it was fixed every time I filed it. [09:19] tarzeau: rt@ubuntu.com [09:19] how many more bug reporting places do they want? they already also have bug reports in gnome-software appstore [09:19] TheMaster: where can i look the bug up, after filing it? [09:20] do you have your last mail that succeeded filing it? i'll just copy paste your's [09:20] can you send it to alex@aiei.ch ? [09:20] I do not have it, no. I used the web interface to file. [09:23] TheMaster: sent [09:35] tarzeau: Looks like I filed it 5 years ago, also had gotten someone else to file one about 4 months later too. [09:35] i filed it got an auto reply to see further discussion in another channeld [09:36] #canonical-sysadmin [09:36] but they say it's not their business [09:37] I would put money on a stale lock file. [09:37] it's another people, and they don't read rt@u-d [09:37] tarzeau, hey, just curious but do you see value in those popcon result/want to use them? like would it be let down if the site was removed rather fixed? [09:37] not saying that it's going to [09:37] xnox: how does systemd decide if the system is 'running' do you know? [09:37] seb128: absolutely, YES! [09:38] I ABSOLUTELY ENJOY THE DATA YES [09:38] but the major thing is, either inform people on the url, it's not there anymore AND REMOVE POPCON-deb package [09:38] OR fix it [09:38] yeah, agreed either it should be fixed or removed [09:39] unsure it provides useful data though [09:39] not have the url work with half assed outdated data. making the poeople go crazy reporting it to gnome-software center review, launchpad and rt@u-d [09:39] it's strongly biased toward what is installed by default [09:39] seb128: maybe not useful for you. but maybe check the webserver access logfiles? of the time when it was up to date [09:39] "people" being you? ;) [09:39] it'll have the answer [09:39] That's true, but can still be useful nevertheless. [09:39] there's this other guy: [09:39] TheMaster: SAY SOMETHING! [09:40] i don't know how much work it needs kept running, if it was way too complicated, i'd also just get rid of it - but i have no idea, i was only user of it [09:40] It was mentioned in the other channel that it's only useful to show 'installed', but that entirely ignores the 'vote' data.. [09:41] i'm only interested in how many machines have it installed [09:41] but then now that ubuntu 18.04 has snap, flatpak support, they should probably integrate/add that to popcon too: [09:41] http://www.aiei.ch/linux/sw [09:41] Eh, I couldn't really care about that info. [09:41] i did write this littly thing for me to get an overview on cluster nodes and foreign-computers [09:41] TheMaster: what info are you looking for? [09:42] I'm only interested in things actually packaged. [09:42] TheMaster: i want to know of specific deb packages i maintain how many users do i have in debian, how many in ubuntu (registered popcon users) [09:42] TheMaster: anything you miss? or work on? [09:42] xnox: oh right, I guess if some units are still activating [09:42] TheMaster: ah about non-deb packages. likewise here [09:43] TheMaster: so my sw script is to make sure there's no snap packages [09:43] TheMaster: for the at my @work place [09:43] tarzeau: Yeah, I maintain a few packages in Debian too, but not usually the only things I'm interested in. [09:43] the biggest issue now is pip packages, and getting them updated (or not, if people link stuff against cython stuff) not packaged, but hand installed [09:48] xnox: so on artful & bionic in lxd it seems like we timeout waiting for dev-getty, do you know why that might be? [09:48] that makes the time to running be 90 seconds [09:49] Laney, need a copy of your container i guess. or how do recreate it locally? [09:49] Laney, we possibly have bugs in the config of the default container. [09:51] xnox: just lxc launch images:ubuntu/bionic/amd64 should do it I think [09:51] see if it is 'starting' or 'running' after 5 seconds or so [09:52] Laney, ack. I can look into it. note this is not the image that foundations produces, but it is stgraber's edition. [09:52] yeah [09:53] # systemctl list-jobs [09:53] JOB UNIT TYPE STATE [09:53] 1 graphical.target start waiting [09:53] 64 serial-getty@getty.service start waiting [09:53] 2 multi-user.target start waiting [09:53] 58 getty.target start waiting [09:53] 82 systemd-update-utmp-runlevel.service start waiting [09:53] 66 dev-getty.device start running [09:53] yeah, not sure why serial-getty@ is there. [10:00] yeah, who put that there? [10:01] Laney, # systemctl disable serial-getty@.service [10:01] Laney, makes the reboot of said container faster. It is an odd thing: [10:01] # systemctl enable serial-getty@.service [10:01] Created symlink /etc/systemd/system/getty.target.wants/serial-getty@.service → /lib/systemd/system/serial-getty@.service. [10:02] who's enabling it? [10:02] given that no instance is specified, thus it's confusing as to what device it is picked to operate on. Imho that shouldn't execute at all. [10:02] something in the image build I guess [10:02] Laney, most likely overlay metadata of the container. [10:02] Laney, let me try digging as to where the images are built. I've done that once before. [10:03] https://jenkins.linuxcontainers.org/view/Images/job/image-ubuntu/architecture=amd64,release=bionic,restrict=lxc-priv,variant=default/ [10:04] Laney, i wonder if this is systemd regression. cause it looks like serial-getty@.service "choose" to use bar part of the foo-bar@.service as the instance name. [10:04] when in fact instance name is empty.... [10:04] (normally foo-bar@instance.service) [10:04] laney@nightingale> ls etc/systemd/system/getty.target.wants ~/Downloads/s [10:04] 'getty@tty1.service'@ [10:05] hmm [10:05] xnox: the ubuntu: images don't have this unit enabled [10:05] just getty@tty1 but not serial-getty@ [10:12] Laney, hehe =) i hope you feel special now [10:12] I do feel confused [10:12] but that's not special, quite normal :'( [10:12] Laney, i guess my email is faster than yours ;-) [10:15] Laney, back to image - i've tracked down the image rootfs in /var/lib/lxd/images by hash from image list, mounted it.... and it doesn't have said thing enabled in the image [10:15] it's magic as to where it comes from [10:16] xnox: I only get emails every 5 minutes, and even then there's no notification thank you very much [10:16] * Laney goes to see what you're talking about :-) [10:17] and yes indeed, that rootfs I linked doesn't have it enabled in there [10:17] Laney, i get push pop-up on the phone & chrome browser on the desktop =) [10:17] I reckon some runtime thing [10:17] eww [10:17] my phone is also on silent and turned face down [10:18] haha, thanks, what a nice email [10:22] Laney, stgraber - 1) no idea where /etc/systemd/system/getty.target.wants/serial-getty@.service comes from 2) no idea why that runs at all, given no instance specified... [10:25] xnox: it's actually serial-getty@getty that gets instantiated [10:25] this seems weird [10:25] Laney, yeah, that's also one more no-idea [10:25] there's no DefaultInstance in there (or whatever it is) [10:31] omg [10:31] something is broken in systemd. [10:31] Laney, so i created a service called 'foo-bar@.service' which does ExecStart=/bin/echo %i [10:31] impossible to start it. [10:31] Laney, made it "wanted by" getty.target, and I got this.... [10:32] # ls -latr /etc/systemd/system/getty.target.wants/foo-bar\@.service [10:32] lrwxrwxrwx 1 root root 19 May 18 10:30 /etc/systemd/system/getty.target.wants/foo-bar@.service -> ../foo-bar@.service [10:32] May 18 10:31:00 prepared-pony systemd[1]: Started foo-bar@getty.service. [10:32] May 18 10:31:00 prepared-pony systemd[92]: foo-bar@getty.service: Executing: /bin/echo getty [10:32] haha [10:32] getty is a special case? [10:34] (hmm, there's a systemd-getty-generator) [10:36] yeah, i'm looking into getty-generator.c and it appears to be mangling stuff in add_symlink [10:45] xnox: this is just what systemd does... [10:45] May 18 11:45:26 nightingale systemd[3946]: Starting Why laney... [10:45] May 18 11:45:26 nightingale echo[9380]: laney [10:46] laney@.target.wants/laney@.service -> ../laney@.service; systemctl --user start laney.target [10:46] you get @targetname apparently [10:46] O_o [10:47] * xnox is very puzzled. [11:56] xnox: it's got to be that detect_container() <= 0 hasn't it? [11:56] and add_serial_getty() get an empty string [11:56] but we shouldn't get that far if detect_container works properly [12:11] Laney, but container=lxc is set as an environment variable... hence detect_container should always work right [12:11] Laney, also I hate that there is no logging from generators anywhere that I can see. [12:12] well it should, if that /run/systemd/container thing is written at the right time [12:13] root@test-laney-slow-boot:/run# rm systemd/container [12:13] root@test-laney-slow-boot:/run# /lib/systemd/system-generators/systemd-getty-generator [12:13] Found container virtualization lxc. [12:13] Automatically adding console shell. [12:13] container=lxc is set globally, as the environment =/ [12:13] i wonder if systemd clears that when launching generators [12:14] * xnox ponders how to create lxd container and mangle it before booting it for the first time [12:16] lxc init! [12:17] xnox: it only uses the getenv thing if you're pid 1 though, so not sure how that works [12:17] and yeah, when it works you get that console-shell thing which you don't get on normal first boot [12:18] maybe the ptrace thing works though to fish the environment out of pid1 from another process [12:21] Laney, well used init, and then modified things on disk, then started the container [12:21] diverted that generator to &> /run/getty.log [12:21] nice, I always forget about init [12:21] to get [12:21] # cat /run/getty-generator.log [12:21] Automatically adding console shell. [12:21] which is bad, as it doesn't say "detected container blah" [12:22] let me crank up the logging on said generator. [12:22] but console-shell isn't there? [12:22] /dev/console should be there. [12:22] and is mostly harmless. [12:22] the unit isn't generated [12:29] xnox: nope, I actually removed that generator completely and the units were still made :/ [13:06] Laney, i'm lost, strace lxd daemon?! [13:26] xnox: ha, no thanks, let's rope in stgraber when he's around [13:29] stgraber, in summary. we are confused and we need your expert guidance, cause it seems like the force is not with us anymore.... [14:53] am i the only one ? [14:53] Could not establish FTP connection to upload.ubuntu.com: timed out [14:53] that is just from 'dput' on cosmic [14:54] isn't this a policy violation? https://pastebin.ubuntu.com/p/JNtz6dZ7z9/ [14:54] having libcurl3 shipping libcurl4 instead of libcurl3 [14:58] smoser: logs seem to think it's up and handling connections, and I can connect to it with lftp (for testing). But of course FTP is often pretty firewall-sensitive, so if it's being problematic in your network environment then use SFTP. [15:02] well,. it worked 30 minutes ago [15:02] and then not 20 minutes ago [15:02] and now it just worked. [15:02] no indication of a problem in the logs [15:02] and just 'ftp upload.ubuntu.com' did get a login prompt when iut wasnt [15:02] odd. [15:02] but could be your network env. [15:03] yeah.. but i use the network sometimes [15:03] i would probably notice :) [15:04] ubiquity seem to use $RELEASE for at least the live installer. now, where does it get that from? :-) [15:04] ^-- xnox cyphermox: one of you might know? [15:04] I'm having problems using dput to Ubuntu now also [15:06] SFTP error uploading to upload.ubuntu.com: SSHException('Error reading SSH protocol banner',) [15:06] or [15:06] Could not establish FTP connection to upload.ubuntu.com: timed out [15:08] jbicha: i saw that and then it went away. [15:08] cjwatson: ^ [15:10] hmm, pepo is very loaded [15:10] top - 15:10:39 up 39 days, 20:34, 2 users, load average: 34.18, 27.68, 18.73 [15:11] maybe SAN sadness again [15:12] it was bad for a while and then better when its load dropped [15:12] ok, my upload worked now [15:12] xnox cyphermox: never mind. found it when I checked the right source code :-) [15:13] The state of the SAN is an ongoing problem there though, so if it's that then it's not something I can fix in a couple of hours before my EOW. wgrant ^- FYI [15:13] interesting, the load graph has a really obvious daily spike correlating with the apt-ftparchive cache vacuum [15:17] which I guess would make sense, so another reason to try to move that to weekly and at a more fixed time === caravena_ is now known as caravena [16:33] tseliot: What has changed between the old/new nvidia packaging? === shiftplusone_ is now known as shiftplusone [20:44] sarnold: Poke? [20:48] hey Unit193 :)P [20:48] sigh [20:48] please forgive typos, swapped keyboards again [20:48] I just thought your beard was on sideways. Mind a quick PM? [20:48] sure :) [20:48] * sarnold double-checks his beard