[00:11] I followed a guide and made a .deb package. How can I submit it to be included in the Ubuntu repositories? [00:11] I think it's a hot mess! But I tested it and it does appear to work right now. [00:14] running ubuntu 18.04, trying the 2nd answer located here, python script in /bin/myscript.py isn't running. does it also need to be executable even if i call "python /bin/myscript.py" ? https://askubuntu.com/questions/814/how-to-run-scripts-on-start-up [00:15] ekaj: what errors do you get? [00:15] deadmund: You may want to start with searching and finding documentation that readily available: https://wiki.ubuntu.com/Upstream [00:16] well i'm used to centos and there being a /var/log/cron log, but i don't see that file. also, i don't see the output in ps [00:16] so no errors that i've found yet sarnold [00:16] pragmaticenigma: thanks [00:16] ekaj: aha, cron usually mails errors to people, look in /var/spool/mail/root or similar [00:17] hold on [00:17] ekaj: which python version is your script targeting? Also, /bin is not where you should be placing your own stuff [00:17] in syslog i do see "CRON[413]: (root) CMD (python /bin/myscript.py &)" [00:17] 2.7 [00:17] it's currently in bin because i was following a different guide that didn't work [00:17] it's usually in /home/user [00:18] also nothing in /var/spool/mail [00:18] i went ahead and moved it back to /home and updated crontab [00:18] ekaj: try taking that & off [00:19] ekaj: You should be able to run a python script from your home folder. If I'm planning on something being more perm, I use /opt [00:19] rebooting, 1s [00:20] ekaj: Yes, sarnold is right... don't launch your script with "&" at the end until you have confirmed it is actually working, and for sure do not use & in a cron tass [00:20] *task [00:20] does anyone know how can i get the ip of a android phone connected thru kde connect from my computer? [00:21] no change removing the & [00:21] mattfly: Is the device connected via USB? [00:21] wifi [00:21] ekaj: does the script have any print statements? [00:21] ekaj: I think the systemd solution is more modern, maybe give it a try [00:21] pragmaticenigma yes [00:22] but it prints to a file as well [00:22] Ryvius: that's a nice suggestions, but we have to get the script executing first [00:22] ekaj: Is the script a direct copy from the CentOS box? [00:22] It executes if he runs it manually, yes? [00:22] Ryvius: no, it's not [00:23] pragmaticenigma no, no centos boxes were used. it was made on the box i'm trying to get it to run on boot [00:23] and yes it runs manually [00:23] So he's trying to autorun something that doesn't work at all? [00:23] ekaj: is it possible to post the script to paste.ubuntu.com? [00:23] the script runs and works fine [00:23] ekaj: or explain what the script does? [00:23] I didn't love how *one* line on crontab turned into multiple systemd files [00:23] essentially records video when a door is opened, uploads the script via scp when door closes, repeats [00:24] plus a bunch of error checking [00:24] so i noticed it's breaking when it tries to establish an ssh connection [00:25] ekaj: try adding something like writing the date to a file in /var/log/sarnold in the first or second line of the script -- maybe it's executing, but just not working when run through cron [00:25] it is executing, it's failing when it tries to establish an ssh connection [00:25] aha there we go :) [00:25] ekaj: what are you using to establish the SSH connection? A library or system call? [00:26] before i do the try/catch for my ssh connections i have an "Attempting ssh connection.." line, but nothing is written after that [00:26] paramiko [00:27] worst case scenario i can just autologin the user i think? But I'd prefer not to [00:28] ekaj: It might not be so much a difference in platforms as much as the progression of certain elements moving into userspace and attempts to ensure the system is secured when no user is active [00:29] ekaj: I'm not familiar with paramiko... I typically use subprocess to make an outgoing connection [00:30] i had assumed it was something userspace related === coffeeguy is now known as covfefe-guy === covfefe-guy is now known as coffeeguy [00:30] is there a way to just auto-login the user then? [00:31] ekaj: You can auto login a user... should be in the settings control panel, Users [00:31] no gui :p [00:31] oh [00:31] well... that's a bit trickier [00:31] intention is for this device (when it works) to just sit somewhere forever and do this [00:31] so didn't bother installing one [00:33] ekaj: I can understand that philosophy... the intention is for it to run as a system service though? [00:33] or "like" a system service [00:33] i don't know, i was just going to have the user run the process [00:33] doesn't matter which user [00:33] all of the paths in the file are hardcoded [00:34] ekaj: another possibility is to use incron or systemd units: https://www.freedesktop.org/software/systemd/man/systemd.path.html [00:34] i had just stumbled on this as well https://selivan.github.io/2017/12/27/console-autologin-systemd-ubuntu.html [00:35] ekaj: that way you could get the files a lot faster than waiting a full minute, *and* it wouldn't run the script every minute, 1440 times a day, just when needed [00:35] ekaj: in truth, if you could get that running, I suspect you could get your python script going [00:35] it doesn't run every minute, it runs a while true: loop and constantly runs [00:36] eek. that might be way worse, depending on what you're doing :) [00:36] it has to catch a door as soon as it opens, so if i had it on a timer i could potentially miss people [00:37] a reed switch is tied to some of the gpio pins on the pi [00:37] pragmaticenigma were you referring to the autologin? [00:38] ywa [00:38] yes [00:38] shoulda just bought a damn camera lol [00:38] :) [00:39] it's been fun though, learned quite a bit [00:39] ekaj: or look into the project "motion" which does a lot of that stuff [00:39] it doesn't need motion detection though.. i woulda used that one if it did [00:40] i use "motion" for my cameras to detect and record when it sense motion. I believe a camera can be configured to trigger with switches as well [00:40] the reason it's tied to a reed switch is that the only time people use this particular door is when something is being stolen, or an emergency [00:40] but traffic passes by it on a regular basis [00:42] kind of sounds like if it was working before... why change anything with the OS? [00:44] i had to manually run it before [00:44] also, this link worked, which is odd because i didn't set a password https://selivan.github.io/2017/12/27/console-autologin-systemd-ubuntu.html [00:48] well script runs but now there's a new error yay. [00:50] yay :D [00:53] interesting, receiving two errors, seems to maybe be some type of race condition [00:54] but anyways, my ubuntu specific part is solved, thanks all for the help [00:55] oh fixed it.. either the network service hasn't started yet is one cause, other cause is because i didnt' remove the script in root's crontab, so root made the file in /home/user/logfile.log, so it said permission denied and failed [00:56] when user tried to run [00:59] I recently installed the email client seamonkey and noticed that the icon does not remain in the dock after I close out of the app. I right-clicked the icon in the dock while seamonkey was running but there was no option to "add to favorites". Any ideas? [01:02] For context, this is SeaMonkey 2.49.4 on Ubuntu 19.04 x86_64 [01:13] The documentation is bit outdated. It mentions GNOME panel. [01:15] "gnome panel" it's been a while since I heard that one [01:16] elPuma, You're trying to add the launcher to the bar? [01:17] , uRock yes exactly [01:17] elPuma, This might give you a new gray hair or two. https://linuxconfig.org/how-to-create-desktop-shortcut-launcher-on-ubuntu-18-04-bionic-beaver-linux [01:19] , uRock word thanks. [01:20] elPuma, anytime [01:23] does the foo.desktop file have to be saved in `~/Desktop`? [01:25] whoops, ignore backticks. I ask because I try to remove or hide superfluous default directories so they don't clutter my home. But [01:28] elPuma, I'm looking. I am pretty sure they go somewhere else [01:28] this seems necessary. === zbenjamin_ is now known as zbenjamin [01:32] elPuma, should be able to drop them in /usr/share/applications/ [01:32] That's where the rest of them are. [01:34] The .desktop files for the launcher icons are typically saved in /usr/share/applications/ ? [01:39] Update: Partial success. The icon is in the applications menu. So I can access it with the super button now. Still can't add it to the dock, but I'm good with this. I just wanted an easy option to launch the app in case I forgot the command I created for in .bash_aliases [01:39] elPuma: On xfce4 here - and yes /usr/share/applications/ :) [01:40] The logo is pretty ugly anyways lol [01:46] Wait nevermind. I can add to favorites from the super/applications menu. Everything is right again. It's weird how so much of what I learn about unix related tools stems from a desire to change some minor cosmetic thing, or to save a few keystrokes in bash === Vilegent is now known as kk4ewt [02:03] hunter [02:09] tomreyn any development for https://bugs.launchpad.net/bugs/1833829 ? [02:09] Launchpad bug 1833829 in xorg-server (Ubuntu) "High memory usage by Xorg with Kubuntu 18.04, plasma-desktop, nvidia 390" [Undecided,New] === [b0b] is now known as b0b [02:44] I'm trying to activate the BFQ scheduler via grub boot options, for kernel 4.21>, but it isn't working on ubuntu. [02:46] maxcell_: do you get any error messages? [02:47] sarnold, no it boots fine but when i type the command to show what's going on i only have MQ-Deadline and None activated. [02:47] I need bfq-mq or bfq for my rotate disk [02:47] https://wiki.ubuntu.com/Kernel/Reference/IOSchedulers [02:47] For Ubuntu 19.10 (Linux 4.20) onwards, multiqueue is enabled by default providing the bfq, kyber, mq-deadline and none I/O schedulers. [02:48] i don't have bfq here, only mq-deadline and none OerHeks [02:48] let me show you. [02:48] maxcell_: try "modprobe bfq" then try setting your scheduler again [02:48] sarnold, OerHeks https://paste.ubuntu.com/p/7PpZw6dPzt/ [02:48] sarnold, ok [02:49] yay worked [02:49] woot [02:49] how i can make bfq goes up on the boot? [02:49] thank you very much sarnold [02:50] hehe [02:50] maxcell_: sticking a file in /etc/modules-load.d/ with the module name will be enough to get it loaded, but I'm not sure if there's a similarly polite way to change it for all your devices automatically or not [02:51] sarnold, that is https://community.chakralinux.org/t/how-to-enable-the-bfq-i-o-scheduler-on-kernel-4-12/6418 an udev rule on this website to do that [02:52] maxcell_: oh cool! [02:52] i've used many times it works [02:53] you only need to make the code right by changing the deadline to mq-deadline and cfq to bfq [02:53] also the first bfq is actually mq-deadline [02:55] gonna reboot to test [02:58] sarnold, https://paste.ubuntu.com/p/Y6zSTkGjj6/ [02:58] huzzah! [02:58] o/ [02:58] thanks maxcell_ :) [02:58] i thank you haha [03:10] there is any way to activate top hot corner on ubuntu? [03:11] that's way outside my experience, but there's some extensions on https://extensions.gnome.org/# that mention hot corners and hot dogs [03:12] time to go grab some dinner, have fun :) [03:12] ok ok [03:12] have a good one [03:12] gnome-tweaks [03:12] https://www.fosslinux.com/4184/how-to-enable-hot-corners-in-ubuntu-18-04.htm [03:13] OerHeks, i'm on this website lol, but can we install gnome-extensions from outside on this new gnome? [03:13] i thought it was blocked (only way by downloading via apt-get or so) [03:14] OerHeks, i'm going to try [03:14] ‘custom hot corners’ in softwarecenter [03:15] and then enable / tweak it with gnome-tweaks [03:15] nice [03:16] worked [03:16] thanks dude [03:17] i love how friendly ubuntu are [03:17] and things just works the way you want to [03:36] On ubuntu18.04 do scripts in /etc/init.d will get run after a reboot? [03:37] yes [03:38] metacontent where in the boot sequence scripts in init.d get triggered? [03:44] mallu0987: I'm not sure, but check out `man systemd` for more information [03:44] ok thanks [04:09] I am trying to run a program thru crontab with - @reboot /bin/sleep 300 && /usr/bin/evebox -v -D /tmp --datastore sqlite --input /var/log/suricata/eve.json &>/dev/null but it didn't work, is the crontab syntax correct? [04:34] let me reboot and check modified crontab [04:58] uhh it didn't work [04:58] hey. is there a way i can make firefox NOT quit after the last window closes? very annoying. then it always opens the last tab i already closed when i open something else.... (ubuntu 19.04) [04:58] created a startup script from KDE and that failed too [04:58] script is working from console [05:01] looks like I need to create a service file and execute the script from there [05:03] running a script from crontab didn't work? [05:03] like... use a text editor, like vi [05:06] thanks, I finally managed to got it working [05:07] it's because I entered the wrong user name [05:07] thanks [05:10] Gallomimia: yeah, it didn't [05:12] howdy [05:13] Is System Profile reliable for benchmark tests when testing different system loads? [05:38] Guys I have just set up Ubuntu 18.04 dual boot alongside Windows10, everything was smooth when i was loging in Ubuntu after the last time I logged in Windows now when i try Ubuntu I am unable to reach the login screen [05:38] This is the error loghttps://ibb.co/Bg8YFwK [05:41] Why do people still recommend dual-booting? It's always safer to use a separate drive. [05:42] whenever I ssh into my Ubuntu server, the console prints "Computer reboot required", can I check why does it need to reboot? [05:42] CountryfiedLinux: Maybe coz they are scared of fucking up with linux and want smething to work with in times of crisis [05:43] *log https://ibb.co/Bg8YFwK [05:43] I've been dual-burned in the past and would never recommend dual-booting to anyone. [05:45] well now it works via systemd [05:50] @aneon how does it change the boot process? Does it make dual boot difficult to run? === Roy_Mustang is now known as A_D [05:51] lately, the dual boot stuff doesn't work [05:51] you need to delete secure boot keys and set it to "other os" from bios [05:51] secure boot is garbage [05:53] @aneon not relly my pc you just set it from hp keys to custom key [05:53] some people think adding "secure" to anything and then breaking the stuff is the right way [05:53] you dont need to deleted anything [05:53] gogeta: what should be done [05:53] ? [05:53] as a precaution, on my tower I removed keys [05:54] @sid21gyou bios under securty should have the abilty to set it to use custom keys [05:54] then it will booot under efi [05:54] relly depends on the pc [05:54] some you have to use legicy bios mode [05:55] anyway I get back to my tasks [05:55] also the usb stick should have been set to gpt efi [05:57] @aneon should not do that as that stores your windows key if you every wanna go back or dual boot [05:57] I'm confused with PC benchmark info. Which part is higher is better, or lower is better? Should I assume if an atom processor has a lower number in a benchmark than my i7 that higher is better, or vice-versa? [05:57] It shows my info and other hardware below mine in each category. [05:58] @CountryfiedLinux depends on the softwhere and yes the i7 is better lol [05:59] gogeta, I know it's better, but I'm trying to figure out how to read these benchmarks as to which benchmark parts are higher is better and which are lower is better. If the weaker processor has the lower number does that mean that higher is better, and if the weaker processor has the higher number does that mean that higher is better? [05:59] we dont take polls here CountryfiedLinux [05:59] That second part should be lower is better [05:59] focus on actual ubuntu support questions please [06:00] lotuspsychje, I'm not asking for polling, just trying to figure out how to read these benchmarks. It doesn't specify each part being higher is better or lower is better. [06:00] CountryfiedLinux: hardware & benchmarking there are other channels for, like ##hardware [06:01] @gogeta i7 is faster. "better" depends upon your criteria. @CountryfiedLinux it depends . some benchmark graphics have lower is better, others have higher is better. can you link us to an concrete example? [06:02] any ideas on how to kill a daemon & zombies, my program doesn't have exit or shutdown option [06:03] Sweedish, https://pastebin.com/faBpwdF6 I don't know which ones represent higher is better and which ones represent lower is better. [06:03] I assume that a lower powered processor with a higher number would indicate that lower is better, and if it has a lower number that would indicate higher is better. [06:03] !offtopic | Sweedish CountryfiedLinux [06:03] Sweedish CountryfiedLinux: #ubuntu is the Ubuntu support channel, for all Ubuntu-related support questions. Please register with NickServ (see /msg ubottu !register) and use #ubuntu-offtopic for other topics (though our !guidelines apply there too). Thanks! [06:04] move to the right channels please [06:04] I'm almost done lotuspsychje just give me a minute please. It's not very busy here right now. Not likely to get help elsewhere atm. [06:04] aneon: zombies are already "dead" - they just need to be reaped by their parent process (typically by calling wait/waitpid) - otherwise they'll keep occupying a PID [06:04] okay, I will try /bin/kill -9 $MAINPID [06:04] aneon: what's the program in question? is it a service/daemon? [06:05] CountryfiedLinux: its still not ubuntu related, if you want to continue, pm or meet each other in better channel [06:05] aneon: yeah no, don't do that [06:05] ayekat: dragonfly-mle [06:05] aneon: how do you start it? [06:06] ayekat: /usr/local/dragonfly-mle/bin/dragonfly-mle -r /usr/local/dragonfly-mle -p -v [06:06] aneon: does it background itself? or why do you need to kill it (i.e. does ^C not just quit it?) [06:06] kill it [06:07] doesn't bg [06:07] so where's the problem? [06:07] I cant start it later since it chroots [06:08] I can start it but it doesn't work well [06:09] that sentence about chroot doesn't make much sense to me - and what do you mean with "doesn't work well"? [06:10] when I try to start again it errors saying can not enter chroot, there are zombies [06:10] so I kill zombies first [06:10] I am trying to automate as much as possible [06:10] ... sounds like badly written software to me [06:11] primarily BSD-ish [06:11] not everyone get it working on Linux [06:12] there is a docker build but I am trying to avoid it since it destroys config files [06:14] does anyone checked installation current stable on laptops with eMMc already? [06:14] Hello aneon did what we discussed here but still no login screen for Ubuntu [06:14] uh... if something destroyed config files (or any files for that matter), I probably wouldn't decide to run it directly on my host rather than containerised or virtualised... [06:15] sid21g: do you get a GRUB menu when the system starts? [06:15] sid21g: (or if not, do you know how to make it appear?) [06:15] yes! problem starts after I choose ubuntu [06:16] sid21g: try hitting `e` (I believe? it's marked at the bottom of the screen) to get into the boot arguments editor, and pass systemd.unit=multiuser.target to the kernel command line [06:17] this way at least you get a TTY login, and can try to diagnose things from the command line (instead of staring at a frozen boot screen) [06:17] sid21g: wait, not [06:17] it's systemd.unit=multi-user.target (with a dash) [06:18] ayekat: I want it to not create directories in wrong places, when I change the config and initiate docker build, some strange stuff creates directories in /opt and causes confusion. Anyway, I am not using docker; it doesn't appeal to me [06:20] aneon: I see - but that program looks like quite a complex and fragile beast to me, so I'd consider using docker the lesser evil [06:20] but anyway, zombie processes are annoying, and I don't think you can simply kill them away [06:23] ayekat: this way at least you get a TTY login, and can try to diagnose things from the command line, i wish i could debug [06:23] lets see, I am tryin to figure out machine learning in security. It is shallow but I am trying to see what it does exactly [06:24] sid21g: well, do you get a login screen at least? or does it get stuck again? [06:25] Let me reboot and test [06:25] sid21g: you could look at the journal from the previous boot (i.e. `journalctl -b -1`), and maybe it tells you what happened when it reached the graphical stage [06:26] aneon: well, I have experience in neither one nor the other, so I'm afraid I won't be able to help further than "not waiting for child processes is bad" :-/ [06:27] hopefully I will fix it [06:28] hello , can i upgrade ubuntu 16.04 to ubutu 18 ? [06:29] yes [06:30] ayekat, what is the best way? [06:31] I'm not sufficiently proficient with ubuntu to give a good answer to that [06:32] ayekat, ok thanks [06:32] bryanroderyck: I know that there is the graphical upgrade tool, and there's a command line tool (`do-upgrade` or something), and they both seem to work fine [06:32] I mean... "fine" - modulo all the issues that may arise with an ubuntu upgrade, that is [06:33] bryanroderyck: ah, `do-release-upgrade`, apparently - https://help.ubuntu.com/lts/serverguide/installing-upgrading.html [06:52] can someone please help me on how to chanche hostname on Ubunti 18.04.2 [06:54] nothing seems to work [06:54] I've tried both using hostnamectl set-hostname and also manually editing /etc/hostname and /etc/hosts, but when I reboot the hostname is back to the previous one [06:55] anonymip: so after editing /etc/hostname and rebooting, the content of /etc/hostname is back to the old one...? [06:55] yes [06:55] that... doesn't sound right [06:55] isn't that strange? [06:57] @ayekat failed to type anything [06:57] also when I do: sudo service hostname restart, I get: Failed to restart hostname.service: Unit hostname.service is masked. [07:00] sid21g: where? do you get a TTY login screen or not? [07:00] I did not [07:01] https://ibb.co/Bg8YFwK [07:01] When a user locks the computer screen, how would that affect rsync running in a terminal window (assuming computer is not set to suspend automatically)? [07:01] this is what i see [07:03] sid21g: that's what you see when you attempt to boot into the graphical session (at least that's the same picture you posted above) [07:03] sooo... someone posted a solution for a bug with 18.04 and Kodi https://bugs.launchpad.net/ubuntu/+source/kodi/+bug/1836828/comments/9 can anyone elaborate on this? It says I need to download a file and install it... How do I install it? [07:03] ... but I assume you see the same thing this time [07:03] Launchpad bug 1836828 in kodi (Ubuntu) "Kodi crashes when trying to play any video" [Undecided,Confirmed] [07:04] ayekat: I am unable to do anything [07:04] sid21g: yes, if you get stuck at the same place, you can't do anything [07:04] then whats the way forward? [07:04] I'm out of ideas - not being able to boot to the TTY is pretty limiting [07:05] hmm, I found this thread: https://ubuntuforums.org/showthread.php?t=2389098&page=2 , and removing ubuntu cloud packages did the trick, now th new host name sticks. [07:06] sid21g: what graphics card do you have? [07:07] AMD Radeon R7 M360 and Intel HD Graphics 520 [07:07] ayekat: [07:08] oh, two... [07:08] Intel HD Graphics is the built in one [07:09] I'm a bit confused that it stays stuck like this, though [07:09] but it does stop at diferent stages [07:09] I'd expect it to either just turn blank (and then it's a graphics issue, and booting to a TTY *should* work), or that systemd starts the timeout counter for whatever service is blocking [07:09] i sent the similar one coz i thought it makes no difference [07:10] but that it just stays blocked in the boot screen like that... then again, I'm really not familiar at all with those dual-graphics-card setups and their quirks (mode changes, etc.) [07:10] let me sent you the latest [07:10] I also think its the timeout counter [07:11] https://ibb.co/1RwBqT0 [07:11] @ayekat this is how it looked this time [07:11] https://ibb.co/1RwBqT0 [07:12] unless someone else has a better idea, I'd probably also try passing `nomodeset` to the kernel command line - but that's just a random suggestion [07:13] but where can I do that? [07:13] I cannot type anything [07:13] ayekat: [07:14] sid21g: it's a kernel command line option - you put it the same place where you put systemd.unit=multi-user.target [07:14] but i was unable to put that too [07:14] AH! [07:14] so "boot to TTY" was actually never really attempted [07:15] I am so sorry if I wasn't clear at any stage! [07:15] sid21g: do you get a GRUB boot menu when you boot? [07:15] yes I do get that and now I know i should open it in recovery mode [07:15] uh... "recovery mode"? [07:15] Ok am quite, you tell. [07:16] grub is visible I have option to open Ubuntu/Windows [07:16] anyway, at the bottom of the menu, there's typically some text that tells you which key allows you to edit the kernel command line (I think it's `e`, or maybe F10, or whatever) [07:17] does it say anything there? if yes, press that, and you should be put into some minimalistic text editor interface [07:17] give me a few I will try that right away [07:17] tell me when you're there [07:22] hi everybody [07:22] does somebody have experience using cubic for live iso customization? [07:41] how to purge docker logs when using journald as log driver ? [07:41] https://ibb.co/qFwgyfR [07:41] https://ibb.co/qFwgyfR [07:43] i don't know what i'm doing wrong, but i'm having terrible performance in my ubuntu desktop [07:43] Output after running the journal comand https://ibb.co/jrymJNM [07:44] abbiya: truncate -s 0 /var/log/docker [07:44] and i seem to have the same problem with all my installations.. on the system i'm working on now loading a particular web page (probably with some js code) renders my computer totally useless.. the mouse cursor starts to freeze up and if i'm not quick to do ctrl + w to close the tab, the computer will spiral into a complete halt.. [07:45] /var/log$ ls docker [07:45] ls: cannot access 'docker': No such file or directory [07:45] it is somewhere /var/lib/ i think [07:45] search it [07:46] Output after running the journal comand https://ibb.co/jrymJNM aneon [07:46] sid21g: I a not familiar with secure boot [07:47] var/lib/docker is a directory full of different things [07:48] when I used to dualboot linux stuff on a single disk I used to install uEFI on MBR and other OS in legacy mode with their grub on / [07:48] abbiya: you need to run find and truncate [07:49] LOGDIR="/var/lib/docker" [07:49] sid21g: GDM failing means you're still booting into graphical mode - did you add the systemd.unit=… option correctly? (see /proc/cmdline) [07:49] sid21g: but anyway, now that you're in the journal, you can scroll down (with G you get to the end) - do you see anything useful there? [07:50] its a 20k line output [07:50] and later make a loop [07:50] sid21g: yeah, but the interesting part is at the end [07:50] I doubt there is anyway to create an output file of it [07:50] or try this = find /var/lib/docker -type f -iname '*.log' -print0 | xargs -0 truncate -s0 [07:51] sid21g: you can `journalctl -b -1 >boot.log`, and then paste that [07:51] @ayekat this is the end [07:51] https://ibb.co/qFwgyfR [07:51] @ayekat but where would that log saved and would it be accessible via windows? [07:52] sid21g: no, but you can just use a paste service from ubuntu [07:52] !paste [07:52] For posting multi-line texts into the channel, please use https://paste.ubuntu.com | To post !screenshots use https://imgur.com/ !pastebinit to paste directly from command line | Make sure you give us the URL for your paste - see also the channel topic. [07:52] hm... how does this bot work [07:53] sid21g: journalctl -b -1 | curl -F 'f:1=<-' http://ix.io/ [07:53] someone posted a solution for a bug I have with with my 18.04 install and Kodi https://bugs.launchpad.net/ubuntu/+source/kodi/+bug/1836828/comments/9 it says to download and install libdrm-amdgpu1_2.4.95-1~18.04.1_amd64.deb but in the link provided I don't see that file, how do I get that file and how do I install it? [07:53] Launchpad bug 1836828 in kodi (Ubuntu) "Kodi crashes when trying to play any video" [Undecided,Confirmed] [07:53] wow did not know that @ayekat this is life changing [07:54] sid21g: there's dozens of paste services like those [07:55] ... oh, they left [07:56] * ayekat hopes they're not actually using IRC on the same machine that they're also debugging [07:57] hi [07:58] Hello [07:59] do you use ubuntu? [08:01] yes [08:03] ZaZaGX: what did you break? [08:03] what didn't he break? [08:05] can't find the place where "Try Ubuntu before installing" menu entry is defined on the live CD and cubic does not seem to create the correct auto boot entry for 19.04 [08:06] i broke my dd command [08:06] tuxinator: cubic isnt an official supported package on ubuntu [08:06] of ubuntu mate for the raspberry pi [08:06] it won't load [08:07] !arm | ZaZaGX for PI support [08:07] ZaZaGX for PI support: ARM is a specific (RISC) processor architecture used in a variety of applications such as handhelds and networkdevices. For more information see https://wiki.ubuntu.com/ARM . For ARM specific support, stop by the #ubuntu-arm channel. [08:07] tuxinator: maybe you can contact the cubic maintainer/forum or seek a channel with !alis [08:08] i don't think i'm suppose to do sudo sync afterwards [08:08] ZaZaGX: type /join #ubuntu-arm [08:09] Mead: i was wondering yesterday, are you using official kodi from apt, or did you add a ppa? [08:10] lotuspsychje: any other officialy supported method? or any other idea for a bootable live cd which is easily customizable for a machine which will only be used by guests and should reset on reboot (some kind of kiosk mode) [08:12] tuxinator: we hear good things on cubic, think its a pro package, but we just cant support it here, i hope you understand [08:13] see also https://tutorials.ubuntu.com/tutorial/secure-ubuntu-kiosk#0 tuxinator [08:28] lotuspsychje: I'm using 17.6 from the offical Kodi ppa [08:28] Mead: we dont support ppa's, this should really be tested from the 17.6 apt version on bionic only [08:29] This is how the output looks like http://ix.io/SP20 [08:29] of the journal command [08:29] it worked fine until the update yesterday... [08:30] Sorry ix.io/10SP [08:30] http://ix.io/10SP [08:35] sid21g: that doesn't look like the right output (kodi on windows...?) [08:36] dont know where did i make the mistake [08:36] let me try again [08:37] lotuspsychje: The bug report on launch pad says I need to install a specific library, but I can't find it in the link provided and even then I'm not sure how to manually install a library [08:41] Mead: there are 2 ways of testing this, the official ubuntu way with kodi from apt, removing your ppa, or asking the kodi maintainer whats wrong with 17.6 version on the ppa, probably broken and reccomended to use latest [08:43] I cannot add a openvpn connection in network-manager, I've already installed network-managager-openvpn. [08:43] the openvpn command line utility works. [08:52] sorry, was missing the ....openvpn-gnome piece. [08:52] exit [09:02] hello. recently i cannot seem to ping 192.168.1.1 anymore. but it still works from a windows pc on same lan [09:02] anyone else experienced this issue? === mathisen is now known as Mathisen [09:07] I can't resolve domain names. Please help me thanks [09:18] i have command /usr/local/bin/my-awesome-shell ... it's work like charm when i typing on shell, but when i add to cron with @reboot timing not work. both tested on root user and root cron [09:19] it's just command to reconfigure my firewall using iptables, ufw and ipset and fail2ban [09:20] any alternative to restart point for run special shell scrip [09:20] ? [09:22] sweb: not sure about all the others, but iptables rules are not persistent across boots, so if you change them before a reboot, they won't stick around [09:23] ayekat: any good alternative just i need after all initialize run sample scrip run same env as root shell level. i dont need to create service for that just simple shell after complete init [09:24] sweb: what is your goal? do you want the script to be run when the system starts up? [09:25] ayekat: yeah, after all initialze steps, latest one, run with same env when i normally run the script with user. user is root [09:25] why not put it in bash_profile ? [09:25] Mathisen: user may be not loggin [09:25] Mathisen: because that would require them to log in as root [09:25] Event is reboot not loggin [09:25] sweb: create a systemd service [09:26] it's not service [09:26] it's not anti pattern ? [09:26] sweb: no, but you want your script to run automatically at startup, which is done using a systemd service unit [09:27] and no, it's not an anti pattern [09:28] consider "the Firewall is up and functional" your "service" - and your script simply "starts" that "service" - makes sense to me [09:28] Hi. Anyone know why the ubuntu 5.2.1 and 5.2 builds are failing for amd64? [09:28] https://kernel.ubuntu.com/~kernel-ppa/mainline/v5.2.1/ [09:29] ayekat: what about Type of systemd ? simple, exec, forking, oneshot, dbus, notify or idle [09:29] is there any sample for such my usecase ? [09:29] Shibe: have a look at the build logs [09:30] ikonia: '/home/kernel/COD/linux/include/linux/uuid.h:62:1: error: '-mindirect-branch' and '-fcf-protection' are not compatible' [09:30] hmm [09:30] seems like a bad compiler option? [09:30] are the devs aware or is there any place to contact them? [09:30] sweb: read their descriptions - but I guess you probably want `oneshot` there [09:35] sweb: if you want to do it cleanly, you could also add a script (or an option to your script) (or some other way) to tear down/reset your firewall setup, and have ExecStop= run that [09:35] see http://ix.io/1OTd for an example for how nftables.service does it === katnip` is now known as katnp`[chc [09:42] hi [09:49] Is there ubuntu for rpi 4? [10:02] ayekat: thanks a lot, you save my day [10:02] you're welcome ^^ [10:15] is anyone using fluentd, elastic & kibana on host? [10:26] Hi guys, quick question [10:26] If I use sudo update-alternatives --set php, this sets only the PHP CLI version, right? [10:26] Not the Apache version as well? === imsurit_ofc is now known as imsurit [10:27] Everyone gives conflicting answers it seems. [10:27] systemwide [10:27] So that includes Apache? [10:28] yeah [10:29] Thanks. [10:31] I think there is a slave provision, check documentation [10:32] http://sprunge.us/zvqji2 This is the response to the journalctl query [10:33] I set up dual boot on my system two days back and the ubuntu has stopped entering into the log screen mode, this is the output generated after I ran the journalctl query [10:34] http://sprunge.us/zvqji2 [10:43] hey folks [10:47] sid21g: /usr/bin/gnome-shell: symbol lookup error: /usr/lib/x86_64-linux-gnu/libmutter-2.so.0: undefined symbol: gbm_surface_create_with_modifiers [10:47] no idea if that is a fatal error, but it might be related [10:48] also, don't you have a second machine or any other way to connect to IRC, so you don't need to reboot each time for doing something? you've spend 4 hours on something that's normally just a matter of minutes [10:49] also: /usr/lib/xorg/Xorg: symbol lookup error: /usr/lib/xorg/modules/drivers/modesetting_drv.so: undefined symbol: drmModeListLessees [10:50] sid21g: are you using any special repositories for the graphics driver? have you recently updated your system? [10:52] ... also, for some reason this keeps spinning infinitely - I can see dozens of attempts for starting GDM, each failing the same way (AIGLX/amdgpu errors and then out) [10:54] Hello ayekat: I did try setting this up https://www.amd.com/en/support/kb/faq/gpu-635 [10:55] ugh... [10:57] ayekat: Now i do have second machine [10:59] sid21g: looking at https://help.ubuntu.com/community/AMDGPU-Driver, did the xserver-xorg-video-amdgpu package itself not work? [10:59] or why do you need the pro driver? [11:00] I was playing around, I had an external 4 GB graphic card, so thought why not set it up [11:00] oh sheesh just read the page you shared [11:00] I'll take a look at the script they provide - maybe they're doing sane things, and the issue is perhaps elsewhere [11:00] ok [11:00] thanks a lot, I will try to find something in it too [11:01] I have raised it up here too https://superuser.com/questions/1461723/boot-process-for-ubuntu-18-04-dual-booted-alongside-windows10-halts-before-log [11:01] hi alll [11:01] sid21g: what model have you got? (or which driver did you download?) [11:02] I have Lenovo ideapad-500-15ISK [11:02] attempted to set up this AMDGPU-PRO Driver Version 17.40 for Ubuntu 16.04.3 [11:10] ... aaand they left [11:10] hello. I downloaded the "nvidia" ubuntu image from "https://developer.nvidia.com/embedded/learn/get-started-jetson-nano-devkit#write" Well I know I cannot look for technical support on that image here but maybe one can send me in the right direction [11:10] I can mount that FS locally on my PC but how can I enable sshd? [11:11] my installation is a headless one [11:11] ayekat: excuse me if you said something, net got disconnected [11:12] I see the links in ./etc/rc4.d/S01ssh and also rc2 and rc5 [11:12] this means sshd will start? or not [11:14] sid21g: the install script looks sane, but it appears to install packages it ships on its own, instead of repo packages [11:15] sid21g: not sure if this may cause issues when interacting with the kernel, especially after upgrades [11:16] although, I would expect kernel modules to remain compatible throughout the lifetime of an ubuntu release (that's the entire point of having stable releases, after all), so... I don't see anything that would cause issues there [11:17] and ssh is not enabled, OR it's 100% firewalled [11:17] ssh: connect to host 192.168.1.253 port 22: Connection refused [11:19] alesan: I'm not familiar with the debian/ubuntu abstraction/compatibility layers for sysv-style initscripts that are still lying around, but normally enabling a service happens by creating a symlink to the service under /etc/systemd/... [11:20] (i.e. /etc/rc{something}.d shouldn't really matter anymore) [11:20] ayekat: that is what is disturbing. I have no clue how a fresh was ripped by minimal changes [11:20] sid21g: did it actually ever work at some point? [11:21] yes it worked alright the couple of times I used to set it up [11:21] ayekat, do you think it can be done by mounting the FS on the SD card on a different PC? [11:21] ayekat: but when I opened Windows and the next time i opened ubuntu the situation arise [11:22] alesan: yeah, just mount the root filesystem and create the symlink that would normally be created by `systemctl enable ...` [11:22] sid21g: no upgrade or anything inbetween? === imsurit_ofc is now known as imsurit [11:22] no upgrade I set up 18.04 directly [11:23] ayekat, I have no experience at all with this systemctl, is that systemd thing? [11:23] alesan: yes [11:24] alesan: daemons are expressed through service files that live under /usr/lib/systemd/... (sshd is probably /usr/lib/systemd/system/sshd.service) [11:24] OJ [11:24] I'll try [11:25] ./etc/systemd/system/multi-user.target.wants/ssh.service [11:25] ./etc/systemd/system/sshd.service [11:25] these already exists [11:25] alesan: ah yes, so it should already be enabled [11:25] why then :( [11:25] not sure what the /etc/systemd/system/sshd.service does there, though... but ¯\(°_o)/¯ [11:26] the first one looks correct (and it should usually be a symlink to the file under /usr/lib/systemd/...) [11:27] yeah [11:27] the IP address I'm trying *is* right [11:28] but yeah, there's a couple of other factors that may influence this - multi-user.target not being reached, sshd failing, firewall, sshd running on a different port, ... [11:28] now am scared of setting up ubuntu from scratch again [11:28] If only I knew the problem is with graphic update or the dual boot itself [11:29] sid21g: I'm not sure if that's relevant, but the driver package there is for 16.04 - then again, if it worked at least once, that's still weird [11:30] is there any way windows drives' might be interfaring? Though I didn't find it possible [11:30] alesan: you could try looking at the journal on the filesystem with `journalctl -D /path/to/mount/var/log/journal/{large_id}` [11:31] ayekat, now I have reinserted it in the original device I'll try again, do nmap etc [11:31] Hi guys I have dual boot Debian and Windows10 and I would like remove completely Debian and install Ubuntu. When I remove from live cd the partition of Debian with Gparted, how should I remove the Debian installed grub and install the new Ubuntu one? I have UEFI GPT and in the path /boot/efi/EFI/debian/ there is grubx64.efi [11:32] alesan: otherwise, out of curiosity, what is that /etc/systemd/system/sshd.service file? is it a symlink? (if yes, where?) otherwise, what is its content? [11:33] ayekat, I'll try that in a few minutes [11:34] # nmap 192.168.1.253 .... All 1000 scanned ports on tegra-ubuntu (192.168.1.253) are closed [11:34] MAC Address: 00:04:4B:E6:17:29 (Nvidia) [11:34] caldarella: I'd say the Ubuntu installer will do this itself. [11:35] caldarella: if i get it right, just launch ubuntu setup, overwrite debian partition by formating and you should be fine [11:35] vlt: same intention :D [11:38] alesan, I have to ask the obvious, is ssh-server installed on the remote machine? [11:38] tuxinator vlt removing the debian partition is not a problem, my doubt is that in the efi partition these folders are present: Boot, debian, Microsoft and inside the Boot folder there is the bootx64.efi file which has the same MD5 checksum as the present file in the debian directory: grubx64.efi [11:39] BluesKaj, well, some files are there... including the binary [11:41] alesan: so yeah, have a look at the journal - because debugging that as a blackbox is not very efficient IMHO [11:45] Is it possible to install Ubuntu with support for BOTH UEFI and CSM boot? Trying to create a portable USB-stick but I can't be sure which type of hardware will be present [11:46] ayekat, exactly [11:49] Any idea if 'du' and 'df' will be updated to treat snap mount packages as some sort of special device / non-standard, because I'm sick to death of my terminal flooding out every time I want to check either, because of snap mounts. Going to get far worse too when this obsession with snap keeps taking over everything [11:50] I know I can alias off du / df with things like 'du -h | grep -v snap', but it shouldn't have to be done like that [11:52] inflex: I've got a snap-less 16.04 system (not sure if that's still possible with 18.04), so I'd say the cleanest approach if you don't like snap is to simply get rid of all the snap packages [11:52] (and with "packages", I obviously mean "distribution packages") [11:53] !ot | inflex [11:53] inflex: #ubuntu is the Ubuntu support channel, for all Ubuntu-related support questions. Please register with NickServ (see /msg ubottu !register) and use #ubuntu-offtopic for other topics (though our !guidelines apply there too). Thanks! === westor{off} is now known as westor [12:03] ayekat, ./etc/systemd/system/sshd.service -> /lib/systemd/system/ssh.service [12:04] alesan: weird... so there's two symlinks to the same file? [12:04] ayekat, there is NO /var/log/journal directory [12:05] I have the impression I need to attach a monitor to this thing and perform some installation procvedure [12:05] ah yes, that's because - alright, create the /var/log/journal directory, then try again [12:06] (I forgot that debian-based distributions don't provide that directory by default, so the journal is not saved to disk by default) [12:06] regular attributes? [12:06] drwxr-sr-x+ 4 root systemd-journal 4.0K 2017-09-14 15:58 /var/log/journal/ [12:07] it needs to be set like that? [12:07] but that may be different on ubuntu - otherwise edit /etc/systemd/journald.conf and set Storage=persistent [12:07] I think editing journald.conf is safer [12:08] yeah [12:08] should I delete the directory now I created? [12:09] probably better, yes [12:09] starting up... I'll give it 5 min [12:09] with Storage=persistent, journald will create the directory itself [12:10] yeah I figured [12:14] OK now the directory is there [12:16] now I got the logs... [12:16] nothing about openssh or ssh [12:16] I'll just wait for the cable to connect it to a monitor [12:18] maybe it simply doesn't reach multi-user.target... [12:18] yeah [12:19] I'll delay all of this to tomorrow when I get the cable [12:19] thank's anyway! your help was very nice [12:19] np - good luck! === westor{off} is now known as westor [12:33] so i have a ubuntu 14.04 and on apt-get update i get a few repo's and then 100% [Working] what can i do? [12:35] !14.04 | ws2k3 [12:35] ws2k3: Ubuntu 14.04 LTS (Trusty Tahr) was the 20th release of Ubuntu. !End-of-life was April 25th, 2019. Paid support (ESM) is available. See also !esm, !eol, !eolupgrade [12:36] hehe nice [12:36] i know. but that wasnt the question. [12:36] ws2k3, and what is ? [12:37] see the topic for versions we support here [12:38] if you need assistence upgrading to a supported release (rather than doing a fresh install), please say so. [12:41] nevermind =) [12:49] Getting issues with missing cryptsetup / crypto modules in my initramfs recently - had to boot an earlier kernel to bring system back online and think my current kernel may also be impacted after the initramfs script ran while installing updates. Any suggestions? (Running 19.04). Relevant config excerpts here: https://paste.ubuntu.com/p/45mzTcRdXs/ [12:55] aikema: the incorrect resume device is configured in /etc/initramfs-tools/conf.d/ [12:55] aikema: The initramfs build process seems to be very picky about /etc/crypttab. [12:56] aikema: To not lock you out you can force to include the cryptsetup binary by `CRYPTSETUP=y update-initramfs ...`. [12:57] either replace the resume device by your swap UUID or remove the file [12:58] removing it will make update-initramfs guess the resume device, and document its guess as a warning when it runs [12:58] s/document/report/ [12:59] is your /boot encrypted, too? [13:11] tomreyn: had seen that but swapping the resume device for the one in /etc/crypttab just left me with no "RESUME error" but still the missing binaries noted if I do an "update-initramfs -u -k all" [13:13] vlt: with "CRYPTSETUP=y update-initramfs -u -k all" still getting hte missing cryptsetup / crypto binaries error [13:14] tomreyn: /boot is not encrypted [13:15] aikema: can you quote said error message? [13:17] tomreyn: https://paste.ubuntu.com/p/vZnm99cS95/ [13:18] err... now it's gone it looks like. Removing the resume file worked... updating it hadn't [13:18] solve the remaining issues, post your crypttab if you need more assistence [13:21] tomreyn: /etc/crypttab contents were in my original paste (https://paste.ubuntu.com/p/45mzTcRdXs/). Now down to warnings. Am a bit confused about the reference to a keyfile - only thing I see there is /dev/urandom where googling seemed to suggest this just replaces key with each reboot [13:23] hello i have an dell n5010 and i have issue installing ubuntu 19 64 bit... https://dpaste.de/NmBR [13:23] tomreyn: is size necessary for the resume? Not seeing it in the crypttab eexample in https://wiki.archlinux.org/index.php/Dm-crypt/System_configuration (for arch not ubuntu but should be similar) and strangely the the file the warning points me to doesn't exist on my system (may need to install a docs package or something like that I guess) [13:23] bryanroderyck: this morning you asked about an 16.04 to 18.04 lts upgrade? [13:25] lotuspsychje, yes i did i want to other version of ubuntu than ubuntu 16 [13:25] bryanroderyck: you changed your mind, want clean install 19.04 now? [13:26] cause i want to install blender 2.8 and other softwares [13:26] aikema: README files are usually compressed, you should have /usr/share/doc/cryptsetup/README.initramfs.gz as well as the crypttab(5) man page (run: man 5 crypttab) [13:26] !man | aikema [13:26] aikema: The "man" command brings up the Linux manual pages for the command you're interested in. Try "man intro" at the command line, or see https://help.ubuntu.com/community/UsingTheTerminal | Manpages online: http://manpages.ubuntu.com/ [13:27] bryanroderyck: please describe your issue to the channel, whats happening exactly? [13:27] lotuspsychje, this ubuntu is only weeki have installed it but it doesnt support all software [13:27] tomreyn, I found /usr/share/doc/cryptsetup-initramfs/README.initramfs.gz via apt-file. Strangely the swap line suggested in there for crypttab also doesn't include a size field - which is what I'm getting a warning about now having defined [13:28] is my laptop a 64 bit comatible processor? https://dpaste.de/NmBR [13:28] hi @lotuspsychje [13:29] aikema: yes, size is needed in your case, it is documented in crypttab(5) [13:29] bryanroderyck: yes. [13:29] bryanroderyck: i3, yes 64bit capable [13:29] tomreyn, I see this as the example swap in "man crypttab" - "cswap /dev/sda6 /dev/urandom cipher=aes-xts-plain64,size=256,hash=sha1,swap". No size field there [13:30] but why i want to install the ubuntu 64 it doesnt boot ? [13:30] bryanroderyck: elaborate 'doesnt boot' please, what happens exactly? [13:31] aikema: hmm, right, maybe it's just warning you that it'll go with the default then [13:31] oh, it just does not boot, no "it doesnt support all software" [13:31] it switches off .. [13:32] tomreyn, err... or I can't read. At second look seeing a size=256 I glossed over the first time there. [13:32] i dont see on the web people talking about dell n5010 having problem with ubuntu 64 bit? [13:33] bryanroderyck, did you check the iso? [13:33] !md5sum [13:33] To verify your Ubuntu ISO image (or other files for which an MD5 checksum is provided), see https://help.ubuntu.com/community/HowToMD5SUM or http://www.linuxquestions.org/linux/answers/LQ_ISO/Checking_the_md5sum_in_Windows [13:34] aikema: make that you and me ;) [13:35] i should try to install it from usb key ...are there setting to be done on the cmos bios to install ubuntu 64 bit? [13:35] aikema: note also this quote from cryptsetup(8): For XTS mode you can optionally set a key size of 512 bits with the -s option. Key size for XTS mode is twice that for other modes for the same security level. [13:36] OerHeks, no i didnt check md5sum [13:38] tomreyn, hopefully the default will work. Been seeing 256 elsewhere which is probably good enough here for the moment. Any tips as to how to get back online if a test reboot fails? Would that be to comment out the swap in /etc/fstab and /etc/crypttab or something in addition (after booting via a USB key or some such thing to gain access)? [13:39] bryanroderyck: if you want volunteers to help, its adviced you explain your steps so volunteers can try to help you [13:40] bryanroderyck: you connect your ubuntu usb in your computer, then what? [13:40] aikema: i think you can just resume from the initramfs (busybox) shell if it fails to boot again. after all, the only thing you are encrypting is your on-disk swap partition. [13:41] lotuspsychje, im explaining ...i crashes ..it dosent boot the usb [13:41] aikema: you'd boot without swap, but i guess that's tolerable [13:42] bryanroderyck: what crashes at wich point? [13:42] tomreyn, thanks for the help! Looking at https://bugs.launchpad.net/ubuntu/+source/cryptsetup/+bug/1802617 sounds like if i hit the initramfs shell i just hit "exit" and boot should continue minus the swap. Crossing my fingers and rebooting! [13:42] Launchpad bug 1802617 in cryptsetup (Ubuntu) "After upgrade to 18.10 boot fails on waiting for encrypted source device" [Undecided,Confirmed] [13:43] hi all. just trying linux again after a long time. is compiz still the way to get compositing (in dingo)? installed it, but lost the unity launchers/bars etc. [13:43] lotuspsychje, the os doesnt boot ...anyway i have to do the experience again and record all the parameter. [13:45] testing [13:45] working [13:46] bryanroderyck: please take a picture of your screen, at wich point you get stuck [13:46] Hello, I have installed kubuntu-desktop package, how do I uninstall Gnome safely now without breaking things? [13:47] lotuspsychje, yes i will try ...thanks for your help [13:47] arka500: whats your base ubuntu install? [13:47] arka500: there is no way to easily un-install Gnome (to say, ubuntu-desktop) [13:48] arka500: just live with both. The only real option is to re-install Ubuntu [13:49] tomreyn, on OS X at the moment which gives you an idea how well that reboot went. Hit ESC and the boot is hanging at "A start job is running for /dev/mapper/cryptswap1" which shows as time elapsed / no limit. How do I abort the swap start if ctrl-c seems to have no effect? [13:49] What is the point of clicking on the title bar and change window? [13:50] This is the behavior on the new desktop environment of 18.04. [13:50] Many times I want to close the tab with middle click but instead cycling through the windows a bit shocking at first. [13:50] aikema: hmm 'no limit' on the timeout, i hadn't considered this, sorry. [13:51] leeijaw, for what program? [13:51] All of them [13:51] uRock: all of them. [13:51] aikema: i need to read up on the available options [13:51] aikema: maybe try booting to recovery for now [13:52] This is a system behavior. [13:52] Try it in 18.04 [13:52] tomreyn, ctrl-alt-del'ed my way to a grub menu. as guessing that that's where I need to do some changes. Not sure if basic recovery mode works or if I need to add some flags to abort the swap mount [13:53] If I want to switch apps, I will use alt+tab. What is the point of middle clicking on the title bar? [13:53] leeijaw, middle click doesn't do that for me. I can scroll between tabs in the web browsers by hovering over the title bar, but not in other apps. [13:53] aikema: me neither, give it a try. [13:53] that has been the case for several years. [13:54] I mean the topmost section of a window. [13:54] so, what's the deal with compiz [13:54] youurayy, specifics? [13:55] The bar that has minimize, maximize, and close button, but the bar below it. [13:56] tomreyn, gets to me the "recovery menu"... can get to a root prompt or a bunch of other things but not quite sure how to fix the swap issue. Wondering if my error might have been triggered by my swap getting corrupted somehow such that it'd need to be rebuilt before it could be remounted. Should be safe to blow the current swap contents away after a clean shutdown IIRC? [13:57] tomreyn, or can you fsck a swap partition? [13:57] uRock, i'm trying to set it up on the latest Ubuntu Dingo -- but I lose the Unity Activities/Top bar -- just want my smooth desktop back [13:57] uRock, these things used to work perfectly a decade ago [13:58] youurayy: 16.04 is not 18.04 anymore [13:59] aikema: i think you'd just mkswap again if it was inconsistent for some reason. [13:59] youurayy, Ubuntu has moved to gnome. I think you can log out and select Unity [13:59] aikema: but keep in mind that the swap file system is on top of the crypto container which you effectively regenerate on every boot. [13:59] Not sure how well that works, though [13:59] lotuspsychje ubuntu version? i'm using latest 19.04 [14:00] I have gotten to the point of not liking Unity anymore [14:00] aikema: well at least you give it a new key file on every boot, using /dev/urandom [14:00] uRock yeah i just want whatever comes w/ the install to also work with compiz, but maybe that's not the idea [14:01] compiz is still a "thing" ? [14:01] aikema: so your swap file system does not survive reboots anyways, which is the entire purpose of this configuration you chose to have there [14:01] BluesKaj i guess it is if you want smoothness [14:02] tomreyn, so I dropped to a root shell, and found that "free -m" reports active swap so guessing that means my swap mounted properly (dmesg | grep swap only reported the swap being added / no errors) . Am confused about the boot hanging though [14:02] youurayy: can you give us details of how you installed unity? [14:02] youurayy, glad I'm a kde/plasma guy :-) [14:03] lotuspsychje sorry - i confused the terms - i didn't install unity - just went with the out of box Unbuntu manager -- and tried to add compiz later [14:04] youurayy: you want to use compiz on gnome? [14:05] yeah, we used to compiz on anything [14:05] *to use [14:05] aikema: try if you can complete the boot without modifications. if you cannot, boot to the recovery again, enable networking, drop to root shell again [14:12] ok i see, "gnome-sheel uses mutter, and mutter is compositing" ? [14:12] tomreyn, resume normal boot seems to have gotten me to my desktop (albeit guessing with crappier graphics drivers than usual). What would you suggest I attempt now? Am combing log files to see what I can find [14:13] aikema: i'd like to see all configurations in their current state again [14:13] youurayy: That would be correct [14:14] youurayy: Mutter is being used because it works with both X-server and Wayland. Gnome-Shell has also been tightly integrated with Mutter which can make it difficult to use a different compositor tool. [14:15] aikema: yu probably wont find logs on how the initramfs failed to enable swap during the last but one mount, since log files are not available at this time of booting. [14:15] (possibly there was something in dmesg, though.) [14:16] thx for the explanation pragmaticenigma - any way to make mutter smoother when dragging windows? / tweaking it [14:17] aikema: the reason your graphics are now worse is that recovery boots with the !nomodeset boot parameter. [14:18] youurayy: wich graphics card and driver are you using exactly? [14:18] youurayy: That's something I do not have any experience in. I typically just use my machine as is, and haven't noticed any issues with jutter or tearing. I make sure I'm running up-to-date graphics drivers and that's about it. [14:19] lotuspsychje GeForce GTX 1060 and i believe the i use the proprietary nvidia driver -- how to make sure? [14:19] tomreyn, /var/log/boot.log seems to have some relevant info. Repost of the relevant bits of config (plus boot.log excerpt): https://paste.ubuntu.com/p/8PnDnQ7Xgf/ [14:19] lspci -nnk | grep -i vga -A3 | grep 'in use' [14:20] youurayy: What OerHeks just wrote [14:20] lspci -nnk | grep -i vga -A3 | grep 'in use' [14:20] sorry, Kernel driver in use: nvidia === BrianG61UK_ is now known as BrianG61UK [14:20] is `nvidia` the proprietary? [14:20] yes [14:20] I am on 18.04.2 x86_64 system from Dell [14:20] yes, else you would read nouveau [14:20] No proprietary drivers installed [14:21] I did a fsck -y /dev/sdX on my seagate expansion 1TB disk but it took 2 days and still didn't complete the repair [14:21] The disk has no useful data [14:21] I want to restore the entire disk to a clean slate [14:21] gparted doesn't fix errors [14:21] mkfs.ext4 doesn't fix errors [14:22] so yeah window dragging isn't as silky smooth as in compiz (or win10 / mac) [14:22] I get a READ-ONLY filesystem error while copying [14:22] saurabhdhandare, what does SMART say about the disk ? [14:22] saurabhdhandare, sounds like you do have hardware issues, bad disk? [14:22] it is a brand new one. [14:22] i know i'm nit picking but we had this [14:22] tomreyn, Found someone else reporting an error like mine (though they say they've hit it intermittently - wonder if the reason for it work in recovery mode is just that I got lucky): https://askubuntu.com/questions/1153319/starting-problems-a-start-job-is-running [14:22] saurabhdhandare, have you chowned the directory yet? [14:23] youurayy: what about your other system specs, cpu, ram, hd/ssd? [14:23] yes I did using chown saurabh: /media/saurabh/XXXXXXX [14:23] it had exFAT so I changed to ext4 [14:23] now it doesn't work [14:23] https://help.ubuntu.com/community/FilesystemTroubleshooting [14:23] lotuspsychje it's a beefy PC, if i run compiz the window moving is smooth [14:24] thanks [14:24] aikema: this post lack context, it's pretty useless really. [14:24] *lackS [14:25] aikema: can you show us "pastebinit <( sytemctl status cryptswap1.mount; journalctl -b )" [14:25] saurabhdhandare, chown users:user /path to directory [14:25] aikema:oops, typo! [14:25] youurayy: details of 'beefy' please? [14:25] aikema: can you show us "pastebinit <( systemctl status cryptswap1.mount; journalctl -b )" [14:25] so that and also the font rendering isn't as good as win/mac [14:26] uRock, yeah [14:26] saurabhdhandare, chown saurabh:saurabh /media/saurabh/XXXXXXX [14:26] lotuspsychje: Intel(R) Core(TM) i3-8100 CPU @ 3.60GHz, M2 SSD, GeForce GTX 1060, 32 GB RAM [14:26] i did. I have permissions to write to the disk [14:27] youurayy: if those specs cant run 19.04 then something really wrong in your system [14:27] It is when I copy more than 15 GB that the disk becomes READ-only [14:27] youurayy: can you pastebin us your dmesg please? [14:27] oh they can run it, just not as smoothly as win/mac (nothing suspicious in dmesg) [14:28] youurayy: let us have a look? [14:28] what's the pastebin cmd? [14:28] uRock, can I clean the disk with gparted to a fresh ext4 state?? [14:28] !pastebinit | youurayy [14:28] youurayy: pastebinit is the command-line equivalent of !pastebin - Command output, or other text can be redirected to pastebinit, which then reports an URL containing the output - To use pastebinit, install the « pastebinit » package from a package manager - Simple usage: command | pastebinit [14:28] saurabhdhandare, yes, but you'll have to chown again [14:29] uRock, ok [14:29] uRock, [14:29] After pairing my phone using bluetooth it says "Connection failed -No usable services on this device" ...How to fix this ? [14:29] will it fix the ext4 errors [14:29] lotuspsychje http://paste.ubuntu.com/p/zHQWsnGzZ2/ [14:30] saurabhdhandare, it should. I've never had that issue with extra drives [14:30] uRock, ok thanks. Doing that now... [14:30] saurabhdhandare, ext4 errors on a blank disk are likely disk errors, hardware failing, check with disks > s.m.a.r.t. [14:31] OerHeks, okay i'll [14:31] youurayy: you have ROG STRIX Z370-I GAMING, BIOS 1002 07/02/2018. you should have https://www.asus.com/us/Motherboards/ROG-STRIX-Z370-I-GAMING/HelpDesk_BIOS/ [14:32] thanks for spotting that, i'll update [14:33] youurayy: nvidia: module verification failed: signature and/or required key missing - tainting kernel [14:33] youurayy: are you dualbooting or single boot ubuntu? [14:33] tomreyn, I agree it's not very complete but at least matches my symptoms [14:34] single boot, giving it all [14:34] ^ lotuspsychje [14:34] TJ-, https://paste.ubuntu.com/p/HWXMBkSNMh/ [14:34] lotuspsychje does that mean the driver isn't active? [14:34] aikema: right, it *could* be related. [14:35] youurayy: please try a higher nvidia driver from the ubuntu graphics ppa [14:35] lotuspsychje: That message might be from installing the nvidia driver from the graphics driver's team's PPA [14:35] !nvidia | youurayy [14:35] youurayy: For nvidia and matrox graphics cards, see https://help.ubuntu.com/community/VideoDriverHowto . For AMD/ATI graphics cards, see « /msg ubottu ati » and « /msg ubottu fglrxmissing » For the latest Nvidia drivers see https://launchpad.net/~graphics-drivers/+archive/ubuntu/ppa [14:35] tomreyn, it's the right version of Ubuntu and reported around when I started having these problems [14:35] aikema: the timing is a good hint, i agree. [14:37] youurayy: try also to install preload and haveged and reboot after switching to your new nvidia driver [14:39] TJ-, also poked around for other swap-related systemctl things after that first one didn't amount to anything: https://paste.ubuntu.com/p/69z4gx2Yhb/ [14:39] aikema: was that boot.log you pasted earlier from the same boot session as this log? Because in this session cryptswap came up immediately - "Jul 19 15:54:03 nuc systemd[1]: Activated swap /dev/mapper/cryptswap1." [14:40] OerHeks, uRock I believe I have a hardware issue [14:41] TJ-: i think journalctl -b -1 currently points to where the swap partition would not mount and the boot got stuck indefinitely waiting for this to happen. [14:41] SMART test doesn't work on my expansion drive [14:41] GUI option is not available [14:42] something has to do with 'uas' module in the kernel [14:42] I think 18.04.2 has the stable kernel with fixed bugs [14:42] aikema: also note you have an unrelated USB device failure recurring constantly: "usb 1-4.1: USB disconnect, device number 10" [14:42] it is true, not all ext hdd controllers support smart control, so be it [14:43] saurabhdhandare, then check the health on a windows machine [14:43] OerHeks, https://askubuntu.com/questions/637450/cannot-perform-smart-data-and-self-test-on-external-hard-drive/692892#692892 [14:43] aikema: as tomreyn points out, was boot before this one, the one where cryptswap1 failed? [14:43] OerHeks, [14:43] yeah [14:43] Windows :( [14:44] ok using nvidia-driver-418 (proprietary) now [14:44] TJ-, https://paste.ubuntu.com/p/ZRPgTxZGN3/ is the "journalctl -b -1" which I think is the failed boot which misses a bunch of those swap messages. [14:45] think it's smoother.. 4k youtube video in chrome not silky smooth still, but i'll try that bios update [14:46] it also offers "nvidia-driver-430 (open-source)" not "(prorietary)" - but it doesn't say it's the "nouveau" -- should i activate it? [14:47] TJ-, if you have a fix for that USB device failure, that'd be awesome. It's a UPS that fairly regularly doesn't seem to init properly (but usually seems happy if you restart nut-server and/or nut-client). I sort of gave up on a permanent fix after having previously poked at it for a while [14:47] aikema: i recommend you remove this "INNO TECH USB to Serial" device you have connected to usb there for now. [14:48] youurayy: better not use nouveau on a big GTX card :p [14:49] tomreyn, permanently detach or just disable for debugging purposes here? [14:49] youurayy: that's actually the proprietary driver, too, it's just listed incorrectly. this is neither a recommendation to use or not to use it. [14:50] aikema: for now, while we try to fix your boot [14:50] lotuspsychje what is the logic there -- is it as good as the proprietary? [14:51] tomreyn thanks, these things are confusing for novices [14:51] tomreyn, OK. It's not unplugged. Living in an area with power grid problems, so hopefully I'll remember to plug it back in in the end [14:51] err... now unplugged I mean [14:51] youurayy: depending your needs of course, if you want the full use of your nvidia card= use the nvidia driver [14:52] youurayy: normally you would not run into this, it's just due to using the PPA. a bug has been filed about it, too, i think, so chances are it will get fixed at some point. [14:53] So seems that "Found device /dev/mapper/cryptswap1" is the start of the chunk of logs in the working boot that's missing in the failed boot [14:54] thanks for your help guys. i'll check the bios update later tonight, hopefully the 4k video will be smooth (may also play with chrome://flags a bit) [14:55] also what's the easiest way to get a full win/mac-like font rendering (patents notwithstanding lol) into ubuntu? [14:55] i think we used to compile freetype in the old days [14:59] aikema: those USB errors are logic-level on the USB, which means the remote device is causing the problem [15:00] aikema: is the UPS running well before the PC boots? [15:00] aikema: those errors make it very hard to analyse the logs [15:00] TJ-, am curious why the nut-client and nut-server restarts seem to resolve the issue once they happen - i.e. think this is an issue also implicating drivers - probably a race condition somewhere [15:02] TJ-, am going to try adding that size=256 to the crypt tab though as https://bugs.launchpad.net/ubuntu/+source/cryptsetup/+bug/1802617 seems to suggest it might have been a problem in the past. That said, the logs both report key size = 256 for the device [15:02] Launchpad bug 1802617 in cryptsetup (Ubuntu) "After upgrade to 18.10 boot fails on waiting for encrypted source device" [Undecided,Confirmed] [15:06] aikema: 'size' only applies when a swap *file* (/swapfile) is used; your system has a swap partition that isn't needed [15:06] aikema / TJ: so, as aikema said, the issue seems indeed to be that for an unknown reason systemd just does not 'find' the cryptswap1 device when it fails: https://paste.ubuntu.com/p/kdtzKV33bF/ [15:07] TJ-: according to the crypttab man page: size=: Encryption key size (ignored for LUKS and TCRYPT devices). See cryptsetup -s. [15:07] (but then "cryptsetup -s" just returns an error message) [15:08] TJ-, am wondering why cryptsetup would be giving me warnings about not seeing the size param then (and also the other error re use of a key file which in my case I think is /dev/urandom) [15:08] i was also wondering whether or not this is correct. there is also a size mount option in fstab, which i think does refer to volume size [15:10] aikema: tomreyn I think we may have the wrong focus. If you compare differences in the log files, in the successful boot "Started udev Wait for Complete Device Initialization" occurs a long time after cryptswap1 is created and activated, whereas in the failed boot that happens at the same time as "Started Cryptography Setup for cryptswap1." - I think this is a race condition [15:11] a race condition seems plausible to me. [15:14] 16:09:01 is indeed a lot later [15:19] aikema, TJ-: what doi you think about just trying to boot again normally (after another sudo update-initramfs -k $(uname -r) -u in case we changed anything?) and see if it works out this time, without the UPS connected, or just out of luck? and if it gets stuck, just reboot normally a couple times more to see whether it's really racy? [15:19] aikema: tomreyn Please do some tests. keep the UPS's USB<>Serial connector unplugged until the system has booted fully. If you can reproduce the cryptswap1 issue like that we can look again at the logs. Currently I think the problem is if the udev "Wait for Complete Device Initialization" starts *before* cryptsetup has created /dev/mapper/cryptswap1, then the udev 'settle' action is being delayed because of [15:19] the USB serial device repeatedly disconnecting (that will be triggering udev jobs constantly) [15:20] aikema: that udev service is executing "/bin/udevadm settle" which "... Watches the udev event queue, and exits if all current events are handled...." [15:22] aikema: if you cannot reproduce with the UPS USB unplugged, you could, instead of unplugging, alter the cryptswap mount to declare a "Before=systemd-udev-settle.service" [15:23] tomreyn, TJ-: Rebooted 3x without the UPS USB-attached. First 2 reboots worked without problems and now the 3 reboot is hanging with same symptoms as before [15:24] hmm so this seems to confirm *some* race condition, and i do like TJ-'s theory. [15:27] aikema: OK, don't shut it down, let it time out the job [15:30] TJ-, already shutdown the hanging system. Unlike some other things I've had hang before this one reports "no limit" so I assume that means it'd hang indefinitely [15:33] TJ-, tomreyn : 4x boots of logs: https://paste.ubuntu.com/p/4ZVnzpZ7Tq/ [15:36] https://www.freedesktop.org/software/systemd/man/systemd.mount.html specifies x-systemd.device-timeout and x-systemd.mount-timeout which can be used in /etc/fstab [15:39] aikema: tomreyn if you look at line 8252 onwards you'll see, at shutdown, the swap/crypto device services were present and terminated cleanly ... so I think the message you're seeing is somehow misleading and something else is the true cause here. [15:40] aikema: line 8261 ends that group with "Deactivated swap /dev/mapper/cryptswap1." [15:46] meh, this is not fun. [15:46] TJ-, tomreyn, should I try instead setting TimeoutSec in the systemd mount setup. That way a failed reboot might be more likely to proceed even if TJ- is accurate in suspecting a different root cause? [15:48] tomreyn, thanks for helping though! [15:48] aikema: oh, you're welcome! [15:50] TJ-, tomreyn : I do find systemd rather confusing though. If I'm trying to limit the "no limit" service that prevents the reboot from continuing in event of a failure, which file should I edit? There seem a lot that are potentially related [15:50] aikema: i'm not really sure how to continue. i assume you'll want to develop a workaround next, which is understandable, and lowering the timeouts *might* provide that. on the other hand, identifying a root cause or at least filing a bug report would probably be very useful in general. [15:52] i would not modify the generic mount timeout, but just the one for the specific device in fstab, by adding those x- options i quoted. if you want to try this path. Or wait for TJ- to point out my flawed logic in this (please do!). ;-) [15:52] aikema: tomreyn OK, that group of messages was from a successful boot - I thought with it being a shutdown it was the failed boot. I found the failed boot at line 3796 [15:53] tomreyn, I know TJ- seemed to think seeing what happened if continuing past the "no limit" thing blocking the way. Poking around an active system with the problem seems as though it might potentially help isolate the problem === im0nde| is now known as im0nde [15:54] TJ-, tomreyn : would it help to reupload those logs as 4 separate pastes - one per boot? [15:54] aikema: yes, and to identify which ones ere successful and which were not [15:55] although we could do that together later [15:55] tomreyn: aikema sure, add a timeout, it should help. Currently the problem as far as I can tell is this: cryptsetup creates the device (/dev/mapper/cryptswap1), mkswap 'formats' it, the "Local Encrypted Volumes" is therefore reached (which tells us all the crypt side succeeded) *but* "swap.target" is never reached and this is because for some reason udev isn't reporting "Found device /dev/mapper/cryptswap1." [15:55] aikema: no, the logs are fine once I realised what I was looking at! [15:56] aikema: so no need to repost the logs then [15:57] aikema: Have you created/edited any systemd services/mounts, or installed 3rd party packages that use systemd service control? It *feels* to me like a race with dependency issues between other services/mounts - it is unusual to see this using just Ubuntu packages untouched [15:58] aikema: I think we need to see the complete "/etc/fstab" because there are other crypto mounts going on and related fscks that could be an issue here [16:03] TJ-, no other crypto is mounted at boot or exists in /etc/fstab (encrypted devices are mounted after user login). ("/" is ext4, /boot/efi is fat, and an nfs mount are the only other entries in /etc/fstab) [16:04] TJ-, note that the NFS mount has options set to defaults,nofail. (I probably should be adding _netdev) [16:06] aikema: the crux of the problem is definitely the "udevadm settle" [16:07] aikema: Add that - it might be the cause of udev failing since that mount isn't appearing [16:09] TJ-, even if nofail is set? If udevadm settle is the issue couldn't we just use Before/After to ensure ordering like you'd suggested for one of them before? [16:09] aikema: lets do this first to test the hypothesis [16:10] aikema: in theory waiting for the settle shouldn't hold up the appearance of cryptswap1 which would trigger the activating of the swap, but something is doing that so we need to isolate it [16:10] TJ-, _netdev in place. Trying a few restarts now [16:11] aikema: I don't think a netdev should cause settle to stall - usually it's caused by a local device which was why I recommended unplugging the UPS for now. [16:11] what app can I use to get rid of the artists name of an avi file? I renamed the file, but the artists metadata has not been changed. On vlc it still appears before the file's name. I tried mkvtoolnix, but the artists name is nowhere to be seen [16:12] TJ-, already first reboot is a fail (and still "no limit" as I hadn't figured out which systemd file to update [16:15] aikema: OK, that is helpful in it reduces the problem space [16:24] aikema: in case we don't solve this now and you involve others in trying to solve this, make sure they realise that /dev/mapper/cryptswap1 *does* exist, and you can prove it since we see in the logs "mkswap[943]: Setting up swapspace version 1, size = 31.9 GiB (34244915200 bytes)" and that is executed by the " systemd-cryptsetup@cryptswap1.service" unit which has "ExecStartPost=/sbin/mkswap [16:24] '/dev/mapper/cryptswap1'" <-- and the device is referred to by its device-mapper path [16:26] TJ-, for now had you been suggesting a Before or After modification to a systemd file as a possible workaround? (And might this also work for my UPS connection?) [16:26] aikema: I'm not so sure at the moment... I just had an idea I need to check [16:27] frad: ffmpeg -i input.avi -metadata artist='' -acodec copy -vcodec copy output.avi [16:30] aikema: thinking aloud here to confirm what we know. "Reached target Local Encrypted Volumes." is cryptsetup.target so that tells us all the crypt services have completed successfully, and that infers that "ExecStartPost=/sbin/mkswap '/dev/mapper/cryptswap1'" returned OK (I had been wondering if mkswap process had hung, but this proves it didn't [16:31] TJ-, to answer my last question (I think), you did mention that '"swap.target" is never reached' so I guess adding something like After=udev-settle to swap.target probably wouldn't help? [16:31] aikema: so, we're hunting for something else that is causing a race. I'm wondering if we can stop it happening simply by slowing down boot (as a test!) by causing systemd to do "debug" logging (it gets verbal diahorrea!) - at boot-time interrupt GRUB and edit the kernel command line, remove "quiet splash" and add "debug" and test [16:32] aikema: I agree it probably wouldn't help but lets see if "debug" solves it by slowing things down dramatically [16:34] thanks tomreyn [16:35] does mysqldump fail if there's a short network interruption or will it wait for connectivity and proceed with dumping when networking is restored? [16:47] TJ-, https://paste.ubuntu.com/p/YBN3VRjhGv/ is the debug log [16:47] becool: if transferring over tcp and running into a tcp transmit timeout, it will fail in the same way any tcp transfer would fail then. [16:51] tomreyn: it doesn't actually fail, though. the process counter just stops counting and the dump stops growing after a variable amount of gigabytes [16:51] aikema: did it fail to start the swap in that session? [16:53] Updates for squid and apache on 18.04 were installed today. When updates are installed like these does the update process restart the services so the new version is loaded? [16:54] becool: maybe ask in #mysql (or is it ##mysql ?). to me this sounds like it didn't actually run into a tcp transmit timeout, but packets were lost, and then apparently it fails to handle this. [16:54] becool: consider dumping locally, transmitting (over network link) separately. [16:55] tomreyn: no response there yet. i can't dump locally, it's an rds aurora instance in aws [16:55] i'm dumping to an ebs volume attached to an ec2 instance in aws [16:55] becool: then were's the relation to ubuntu? [16:56] the server i'm dumping to is running ubuntu [16:56] i see [16:56] i used the --compression switch for the mysqldump, so i'm hoping ubuntu supports network compression [16:57] maybe the DB server and compute node are in different locations? [16:57] they are [16:57] it's dumping "over the wire" [16:57] it'd also do that if they were in the same location, i guess, but the network links would most liklley be more stable. [16:58] *likely [16:58] TJ-, no failure that session. Thought you were expecting the slowdown to stop it happening ... or is this a case where you'd want N attempts again? In which case I'd probably need to figure out how better to access grub at startup (had to trigger a failed boot to get it to go) [16:58] really hoping mysqldump behaves more like an rsync rather than just a cp [16:59] aikema: I see a successful discovery at line 3567, is there 2 boot cycles there because I see what looks like another at line [16:59] puxavida: yes, network services which are started automatically when installed, should also be restarted automatically when updates are installed. [16:59] otherwise i'll never know whether or not i'm getting a dump with full integrity [16:59] aikema: sorry, yes, you'll need to do mutliple boot cycles to know if we've stopped it happening :) [17:00] aikema: accessing GRUB - tapping Esc key or you could temporarily edit /etc/default/grub and change things there, then run "sudo update-grub" [17:01] tomreyn, ok thanks [17:02] aikema: the fact we now have a detailed debug log of a succesful boot means we know what to expect to see [17:03] becool: there may be other (better?) data export options with aws rds. i don't actually know, but recommend you check the documentation on this. think of percona xtrabackup (innobackupex) for example. [17:03] tomreyn: i'll check it out. thanks [17:04] you're both welcome. ;) [17:05] becool: mysqldump's original design was more focused on writing to a local file, than over the network. Also, mysqldump is more like cp ... I am not aware of it having any feature that allows it to see what it has already done and what it needs to do. It always wants start from the beginning [17:06] pragmaticenigma: are there any more robust and open-source command line dumping tools to solve that problem? [17:07] becool: write locally, then rsync [17:07] i don't know if i can write locally since it's an amazon rds instance [17:07] i don't have direct access to the vm [17:07] becool: are you trying to do full backups? [17:08] yes, a full dump of a specific database [17:09] is there any hope left for my USB thumb drive? I read on SO that error -110 is about insufficient power supply, but I plugged it directly into several computer, even ones that did not have any other USB ports in use. http://paste.debian.net/plainh/fb798ecc [17:10] becool: I think I'm with tomreyn on their suggestion. I'd have to believe that the instance has to have some sort of snapshot or utility to help extract your data from the instance in an efficient manner [17:11] TJ-, updated the grub config to make debug the default and did an update-grub. It's survived 5 reboots so far which would seem to confirm the race condition as root cause. Setting it back to normal now / not quite sure how best to move forward (other than occasionally have to re-reboot) [17:11] Sven_vB: If it fails with the same message on multiple computers than it's not the computer that is the problem it's the device. Likely, a trace in the PCB is cracked and the drive is damaged. [17:12] aikema: arggh! I've had this challenge before and it is VERY frustrating.... you add debug options and the darned thing behaves. I think we sould next try telling udev to save a debug log [17:14] pragmaticenigma, does the error message indicate the broken trace is likely related to power supply? [17:15] aikema: try adding "udev.log_priority=debug" [17:15] Sven_vB: No, there was nothing to interpret with my message. Your USB Drive is failing, time to get whatever data that is on that device backed up someplace else and replace it. [17:16] pragmaticenigma, ok, thanks! [17:20] TJ-, does that mean that my options are udev being uber-verbose/slow vs. having to reboot an extra time every once in a while? Wondering in that case if it'd be easier just to reboot slightly more often. [17:21] TJ-, i.e. over the long haul system has only gotten rebooted 1x/month or so I think [17:22] aikema: What TJ- is asking you to do is for the short term, to try and figure out what is happening [17:22] aikema: well, from my perspective I'd always want to identify the issue and solve it. You may have a corner-case that affects others that has never been fully understood [17:25] aikema: I'm hoping that unlike systemd's 'debug' output slowing things down so much we avoid a race, that only having udev logging its action allows us to capture the problem in a combination of the udev log *and* the regular systemd/kernel boot logs, by looking at the timestamps of the log messages and understanding what is (not) being done when udev fails to report the device 'plugged' event for [17:25] /dev/mapper/cryptswap1 [17:26] TJ-, safe to pair that udev debug logging with the quiet nosplash? [17:28] aikema: leave off the "quiet" - that tells the kernel to do its lowest level of output, so we might learn something of interest without "quiet" [17:29] aikema: as for nosplash, yes, that is fine to have - in fact I prefer it since it prevents Plymouth from obscuring the kernel messages with a useless 'pretty' image! [17:36] aikema: also, it would be possible to automate the testing to force a reboot if the swap.target is reached, but have it not reboot if swap-target is not reached so you could leave it overnight [17:41] TJ-, interestingly with the udev.log_priority=debug added the system seems to always fail to boot. [17:41] TJ-, at least each of my tests was failing. Anyways, logs from a failed boot: https://paste.ubuntu.com/p/7dh9mRRkg9/ [17:57] TJ-, thanks for the help. Gotta run [18:00] Hello all. I have installed Ubuntu on an external USB hard drive. I installed a different kernel and am trying to sign it for secure boot. Now I am getting the following error at boot: [18:00] Failed to open \EFI\BOOT\mmx64.efi - Not FoundFailed to load image \EFI\BOOT\mmx64.efi: Not FoundFailed to start MokManager: Not FondSomething has gone seriously wrong: import_mok_state() failed [18:01] chrismo16: did you copy over the EFI partition too? [18:01] I might not have [18:02] In /boot/efi/ I have "EFI 'System Volume Information'" [18:03] in /boot/efi/EFI I have "BOOT" and "ubuntu" [18:04] in BOOT I have BOOTX64.EFI and fbx64.efi [18:04] is there a way in 18.04 to keep the screen blanking/monitor power off after 15 minutes, but stop it from forcing me to "swipe up" when i come back? i'd like it to just pop back up with the windows I had open before, like Windows and Mac does when I don't have a screen lock passcode enabled [18:05] in ubunti I have BOOTX64.CSV, grub.cfg, grubx64.efi, mmx64.efi, and shimx64.efi [18:05] chrismo16: here's what I've got on an EFI machine http://paste.ubuntu.com/p/ZcDf837YSX/ [18:05] chrismo16, what different kernel exacly? [18:06] is there any way to uninstall gnome notifications? [18:06] what happens if you tried? [18:06] OerHeks looks like we have the same structure [18:06] 4.19.13-041913-generic [18:07] that's an outdated kernel ppa build, i assume? [18:09] https://kernel.ubuntu.com/~kernel-ppa/mainline/v4.19.13/ vs https://kernel.ubuntu.com/~kernel-ppa/mainline/v4.19.59/ vs https://kernel.ubuntu.com/~kernel-ppa/mainline/v5.1.16/ [18:10] Im running 18.04 and followed a tutorial toget my touchscreen working on an HP Spectre x360 [18:10] Perhaps I should install a newer kernel? Would that mean I don't need to sign it for secure boot? [18:11] I'm following this tutorial to sign the kernal https://github.com/jakeday/linux-surface/blob/master/SIGNING.md [18:12] tomreyn yes. I used ukuu to install it [18:12] ha, `xset dpms 900 900 900` is my hero [18:13] instead of ubuntu's screen blanking option in power settings [18:13] this one turns off the monitor but doesn't bring up the swipe screen when i come back [18:14] !hwe | chrismo16 [18:14] chrismo16: The Ubuntu LTS enablement stacks provide newer kernel and X support for existing LTS releases, see https://wiki.ubuntu.com/Kernel/LTSEnablementStack [18:14] chrismo16: there are hwe and hwe-edge kernels, providing different kernel versions. i assume -hwe is sufficient for your needs if 4.19 is [18:15] tomreyn, isn't HWE 4.18? [18:15] that's vanilla [18:15] lordcirth: sorry, you'Re right [18:16] Yeah, I have HWE on all of mine. edge is 5.0, though, which should definitely cover it [18:16] so chrismo16 would need -hwe-edge which is at 5.0 [18:16] OK let me give that a try. However, have I messed up something with my EFI partition with this signing process? [18:16] tomreyn thanks [18:17] so linux-generic-hwe-18.04-edge [18:17] and xserver-xorg-hwe-18.04 if you don't already have it [18:18] i'm not sure whether hwe-edge are secure boot signed, though [18:19] I see a linux-signed-hwe-edge source package on my mirror [18:23] so I just ran "sudo apt-get install --install-recommends linux-generic-hwe-18.04 xserver-xorg-hwe-18.04 " and it says I already have the newest versions (4.18.0.25.74 and 1:7.7+19ubuntu8~18.04.2) [18:24] you missed -edge [18:24] got it [18:24] Should I also replace generic with signed? [18:24] those which don't say signed should be signed, those which are unsigned should say unsigned [18:25] ok thanks [18:26] "generic" is a different type of indicator, telling this kernel image apart from e.g. OEM images. [18:26] tomreyn: if you see aikema can you ask them to *disable* all snaps as a test? Looking at the udev log around line 877 there are several snap-related /etc/udev/rules.d/ custom rules files. I suspect they might be responsible. We could do with seeing their contents too. Udev log file is https://paste.ubuntu.com/p/7dh9mRRkg9/ [18:27] TJ-: recommending that people disable snaps is one of my specialties! (sure, i will try to remember to do so) [18:29] tomreyn: thanks... those are the only custom udev rules which is why I aim at those first. [18:30] TJ-: i'm not sure i could tell which ones are custom and which ones are not. could you tell by the file names? [18:32] actually if we can assume that all snap udev rules start with "70-snap." then i can tell that 70-snap.acestreamplayer.rules and 70-snap.dosbox-x.rules are custom [18:36] tomreyn: the path /etc/udev/rules.d/ vs /lib/udev/rules.d [18:37] tomreyn: ^^^ anything in /etc/udev/rules.d/ is suspect since it may not be well-tested, unlike the package-installed rules in /lib/udev/rules.d/ [18:37] I'm at 5.0.0-20-generic now [18:38] Unfortunately I still can't boot with secure boot enabled [18:39] TJ-: oh my, looks like i need a weekend. :) [18:39] tomreyn: plus of course I'd love to blame snaps :) [18:39] tomreyn: good news! :) [18:39] chrismo16: still mmx64? [18:41] tomreyn: ...although ... I do see a Debian patch in the UPS server systemd unit (for nut-server) which adds an "After=...systemd-udev-settle.service" see https://git.launchpad.net/ubuntu/+source/nut/tree/debian/patches/0004-fix-systemd-service.patch?h=ubuntu/disco [18:42] sarnold in my boot/efi/ubuntu folder? if so yes [18:42] chrismo16: that's a pity, let's see some details - what does it say when you try to? [18:43] It says it failed to authenticate [18:43] TJ-: but this rule should not trigger if the UPS wasn't connected, i assume? [18:44] tomreyn: it shouldn't *start* until systemd-udev-settle has reported success so in theory shouldn't affect this... but just seems rather conincidental when we consider all those UPS USB disconnect errors too === jhona-aima1 is now known as jhona-aima [18:45] tomreyn: so another experiment might be to temporarily 'systemctl mask nut-driver.service' [18:46] TJ-: but not nut-server.service ? [18:47] tomreyn: from what I can see it's the driver that relies on udev. comments in the -server unit indicates -driver isn't required for nut-server to start [18:48] tomreyn: see https://git.launchpad.net/ubuntu/+source/nut/tree/scripts/systemd/nut-server.service.in?h=ubuntu/disco [18:48] thanks. i'll try convey all of this should i get a chance. [18:50] tomreyn: my reasoning on the snaps... they're *all* loop **mounts* and therefore will trigger events for udevd, see e.g. line 1520 onwards at the failed boot log https://paste.ubuntu.com/p/ZRPgTxZGN3/ [18:50] i like to make people file bugs early. even if this turns out to be a useless or duplicate bug report in the end, you at least have a way to communicate with the later. [18:50] When I boot my machine I have 3 options: os not manager (uefi) - Ubuntu, os boot manager (uefi) - Windows, usb hard drive (eufi) [18:51] tomreyn: yes, if aikema hadn't left whilst I was at dinner I was going to ask they create a report [18:52] yes, those mounts have some overhead. [18:53] tomreyn: my thinking is, as sone of them include udev rules, there may be some kind of race between the snap mount and the snap-package's own udev rule [18:55] i see. [18:56] Sorry I keep going in and out of the channel. I'm using a second laptop now so I will stay on [18:58] So with secure boot enabled, if I choose the OS Boot Manager (EUFI) - Ubuntu I get "Selected boot image did not authenticate" [18:58] chrismo16: can you post sudo ls -lR /boot/efi/EFI | nc termbin.com 9999 [18:59] If I choose USB Hard Drive (UEFI) I get the not found errors [18:59] tomreyn ok [19:02] tomreyn https://termbin.com/89u7 [19:04] chrismo16: you must have run a different command? [19:04] i.e. i think you forgot -lR [19:06] tomreyn: chrismo16 also, "efibootmgr -v | nc termbin.com 9999" [19:06] chrismo16 R means 'recursive' (no harm) === mossman93_ is now known as mossman93 [19:08] Ok so I ran the ls command multiple times and just ran the efibootmgr command and keep getting the same thing however the last 4 characters changes every time [19:09] chrismo16: different urls, can you post them? [19:09] chrismo16 are you on livecd or chroot ? [19:10] 89u7, 3v0o,szgf,nfqw [19:10] chrismo16: what you do there is run some command, then have its output posted to the termbin website, which stores the output and lets us see it using these unique urls [19:10] I', only getting 1 at a time [19:10] lol got it [19:11] efibootmgr -v | nc termbin.com 9999 -> https://termbin.com/nfqw [19:11] Yes that is correct [19:12] sudo ls -lR /boot/efi/EFI | nc termbin.com 9999 -> https://termbin.com/szgf [19:12] ok [19:13] Yes that is correct also [19:13] chrismo16: so those files in /boot/efi/EFI/ubuntu were last written at 00:46 today, what time is it there now? [19:14] 314 [19:15] taiwan [19:15] 03:14 ? so then it's 2h30m ago. at this time you weren't even connected here [19:15] Also the time stamp may not be accurate. I was working on this yesterday and was having time issues between ubuntu and windows [19:16] what's the betting the mount /boot/efi/ is from the wrong EFI-SP - ssd (236970a6-5d57-4234-8b2d-e4b3cb6360fb) rather than the USB (7291f428-dc3d-4582-af19-981ec650962f [19:16] well its PM here now so that would be 15:14 [19:17] so the wrong partition is being mounted at boot/efi ? [19:18] chrismo16: show us "nc termbin.com 9999 < <( lsblk; cat /etc/fstab )" [19:21] I installed the latest Ubuntu (19.04), and it in general works great, but am having some problems with GRUB (UEFI). I multiboot lots of distros on my machine, so wanted the boot menu to say something more descriptive than "ubuntu". I renamed the directory, deleted the old entry with ufibootmgr, and ran update-grub. It seemed to work, but now I get a blank grub prompt and have to manually "configfile (hdx,y)/grub/grub.cfg" to get the actual menu. [19:21] Any idea how to fix that? [19:22] termbin.com/urt1 [19:23] chrismo16: tomreyn "/boot/efi was on /dev/nvme0n1p1 during installation" [19:24] chrismo16: show us "nc termbin.com 9999 < <( sudo blkid )" [19:24] good find, TJ- ! [19:25] tomreyn: it's only a comment note, but I'm wondering if there's an issue with the root-fs, need to check the UUIDs for all partitions [19:26] I tried moving grub to the external so I could make it portable (which I was able to boot it from a different machine) [19:27] termbin.com/wneh [19:27] speaking of /boot/efi [19:27] is it possible to save/restore specific XXXX entries? [19:27] efibootmgr i suppose, anyone know how? [19:28] chrismo16, usually you don't mv grub; you just install it on the external device [19:29] ioria right. I used the wrong terminology. I followed this to do it https://askubuntu.com/questions/740253/how-to-install-grub-in-an-external-hard-drive [19:30] ioria right I used the wrong terminology. I followed this : https://askubuntu.com/questions/740253/how-to-install-grub-in-an-external-hard-drive [19:30] chrismo16: everything there checks out in terms of it aiming at the correct partition UUIDS [19:30] chrismo16: so I don't get confused - the problem is Secure Boot from the USB does *not* start? [19:31] chrismo16: tomreyn I'm suspecting this is one of those UEFI's that requires the operator to *trust* the boot-loader file from the UEFI Setup options [19:32] chrismo16, you have now grub installed on sda1 (and i think is the external device, right ?) [19:32] ioria correct [19:32] chrismo16, and how do you think to boot ubuntu when it's disconnected ? [19:32] TJ- I read that somewhere as well but do not see that option in my bios [19:33] chrismo16: tomreyn *but* the 'trust' issue doesn't explain why grub-install is passing grubx64.efi to efibootmgr, rather than shimx64.efi [19:33] @iora ubuntu is installed on the external and windows on the internal [19:33] ok [19:34] After installing ubuntu I followed this if it helps https://askubuntu.com/questions/740253/how-to-install-grub-in-an-external-hard-drive [19:35] I have a computer for which the wireless card doesn't work after waking from sleep. [19:36] Then I installed a new kernel and then followed this up to step 5 to try to sign it: https://github.com/jakeday/linux-surface/blob/master/SIGNING.md [19:36] TJ- not sure if that addresses your question [19:37] chrismo16: Let's try adding another UEFI menu entry called "Ubuntu SB" and you can test that: "sudo efibootmgr -c -d /dev/sda -p 1 -w -L 'ubuntu-SB' -l \EFI\ubuntu\shimx64.efi" [19:37] chrismo16, then i guess it's an HP thing (alias: you are loading the operating system from cannot provide the information needed by security in order to boot) [19:39] How can I diagnose this? I have tried restarting NetworkManager service, reloading kernel modules (iwlwifi and iwlmvm), setting the interface up with "ip link set wlp63s0 up", but no chance. [19:39] TJ-done [19:39] Want me to try to boot from it? [19:39] chrismo16: now switch it into Secure Boot mode and test booting that "ubuntu-SB" entry [19:39] sim590: see https://iam.tj/prototype/enhancements/Windows-acpi_osi.html [19:39] OK [19:41] I don't see it as an option in my boot menu [19:42] TJ-I don't see it as an option in my boot menu [19:43] chrismo16: now that is strange! [19:44] TJ-: Thanks. I will try that! [19:44] chrismo16: this is starting to look like a problem in the firmware, what is the make/model? do "dmesg | grep DMI:" and tell us what it reports [19:47] TJ- termbin.com/zshq [19:52] chrismo16: HP's documentation isn't helpful unfortuntely; gives no indication of what the firmware setup options are [19:52] chrismo16: let me tell you my theory of what is happening/happened [19:52] chrismo16: originally I'm guessing you installed Ubuntu without Secure-Boot enabled, is that correct? [19:52] which log would i check to see if a remote file transfer failed? [19:53] or how a mysqldump process ended? [19:53] TJ- I believe that is correct, yes [19:59] TJ-: sorry, had a phone call there. [20:00] TJ- its ok. I'm not against reinstalling but I just tried booting to the installation usb and am getting the not found errors [20:01] With secure boot that is [20:02] hp has models which have a whitelist of EFI files names to boot from [20:03] it could be this [20:03] chrismo16: you had ubuntu secureboot working on this system though, didn't you? [20:03] tomreyn yes [20:04] wait no I didn't [20:04] oh ok [20:06] I'm beginning to confuse myself. I've been trying a lot of things. [20:07] tomreyn I do believe I had it working before trying to sign the kernel [20:10] My Meta-key in bash is bound to ESC, any idea how to change that to ALT? [20:10] Yes, I had it working but when I installed the newer kernel it would not work unless I selected advanced and selected the older kernel [20:10] chrismo16: tomreyn Sorry, was looking at an unrelated CVE. [20:12] no worries TJ- ;) [20:12] chrismo16: I'm now more certain that ever that the problem here is that the HP's firmware, when in SecureBoot mode, requires the operator to have pre-trusted the bootloader file in order to add it to its menu. The fact that the command I gave you didn't result in an entry seems to confirm that [20:12] TJ- so should I reinstall? [20:13] <_dbugger> In case anyone wants to repo in askubuntu.com here is a problem I have not managed to solve for quite some time... https://askubuntu.com/questions/1153422/18-04-screen-tearing-with-nvidia-gtx-970 [20:13] chrismo16: I don't think it'll help, I think the issue is that the firmware has hard-coded by-pass trust of the Microsoft Windows bootloader but for any other it pre-requires you, the operator, to indicate your 'trust' [20:14] chrismo16: would you be able to take a photograph of the HP's 'Security' setup page where you en/dis-able Secure Boot and upload it to imgur? [20:14] !paste chrismo16 [20:14] ok [20:15] chrismo16: Before we suggest a reinstall (a very 'Windows' thing to do, but not Linux!) lets be sure what the actual cause is :) [20:15] hmmm, looks like ubottu is asleep [20:15] oh! [20:16] !paste | chrismo16 [20:16] chrismo16: For posting multi-line texts into the channel, please use https://paste.ubuntu.com | To post !screenshots use https://imgur.com/ !pastebinit to paste directly from command line | Make sure you give us the URL for your paste - see also the channel topic. [20:16] * TJ- kicks self [20:16] can you define/set a variable you can use globally all the time ? [20:17] !details | netameta [20:17] netameta: Please elaborate; your question or issue may not seem clear or detailed enough for people to help you. Please give more detailed information; for example, we might need errors, steps, relevant configuration files, Ubuntu version, and hardware information. Use a !pastebin to avoid flooding the channel. === _dbugger is now known as dbugger [20:18] tomreyn, So say i want to commit something to github, and i want my password to be in a variable [20:18] netameta: eeeek!!! [20:18] netameta: use ssh keys [20:18] instead of typing it each time i could do something like $ghp [20:19] netameta: see https://help.github.com/en/articles/connecting-to-github-with-ssh [20:19] ssh keys will work on 2 fector authentication? [20:19] i don't know, haven't tried. but i don't see why not. [20:20] netameta: I use mutli-factor with github but I've never had it affect ssh key usage, only the web site [20:20] netameta: locally you'd have a password on the SSH key you create for github, and then have that cached in the ssh-agent so you do not need to type it every time [20:22] imgur.com/gallery/Sx7tMIP [20:22] well it effects normal login but i will check ssh [20:23] TJ- sory its imgur.com/gallery/Sx7tMlP [20:24] TJ- so you think it was working before because I was using the windows bootloader instead? [20:25] chrismo16: you were using the windows bootloader to boot ubuntu? [20:25] chrismo16: thanks; what is shown if you expand the "▶ OS Boot Manager" entry? [20:25] chrismo16: no, sorry, I didn't complete my theory did i!? [20:28] TJ- imgur.com/gallery/OClSFh9 [20:28] chrismo16: with SB disabled grubx64.efi can be installed, added to the UEFI boot manager, and booted at startup. But once you switch to SecureBoot as grubx64.efi is NOT signed by Microsoft UEFI key that entry will always fail, so the entry for Ubuntu should be updated to boot shimx64.efi, which IS signed by Microsoft. However, we just told efibootmgr to add such an entry and yet it didn't show in the menu at [20:28] boot-time. That indicates it is either a) in the menu but not allowed in Secureboot mode or 2) was silently refused but without returning an error so efibootmgr thought all was OK [20:29] tomreyn I'm just speculating based on TJs comments [20:29] I can't find the PPA url for this https://launchpad.net/ubuntu/+source/universal-ctags/0+git20181215-2 ? I'm not really familiar with ppa. How can I find the proper URL? [20:29] That is, the proper url for doing like "add-apt-repository URL" [20:29] chrismo16: oh okay ;) it'd actually be possible i think, so i thought i'd better ask [20:31] TJ- can we edit the one that exists to use shim? [20:31] !info universal-ctags | sim590: there is no PPA, that package is part of the archive [20:31] sim590: there is no PPA, that package is part of the archive: Package universal-ctags does not exist in bionic [20:31] hmmph! [20:31] !info universal-ctags disco [20:32] universal-ctags (source: universal-ctags): build tag file indexes of source code definitions. In component universe, is optional. Version 0+git20181215-2 (disco), package size 341 kB, installed size 952 kB [20:32] sim590: this is not a ppa, but a package source for building a binary package which could then be distributed in ubuntu proper. [20:32] sim590: which Ubuntu release are you using? [20:32] bionic [20:32] sim590: there you go then; the package wasn't available for 18.04 [20:33] I guess that I can backport it, no? [20:33] chrismo16: We could but I'd rather not because another issue here (since the firmware failed to add "ubuntu-SB") is the firmware is broken and if we managed to change the existing entry the system might fail to boot Ubuntu at all [20:34] sim590: checjk the package dependencies; you might be lucky and be able to (manually) download/install the .deb package from Disco [20:34] sim590: if it doesn't depend on a newer libc6 or libraries / library versions not present in bionic, then yes. [20:35] May be I can add the source to disco and do preference pinning like on debian, no ? [20:37] sim590: here's the depends: https://paste.ubuntu.com/p/fcH3SZcRQs/ [20:37] sim590: you could do this. it would break if dependencies are not resolveable in the future, [20:37] sim590: no, because you'd be adding ALL the packages from Disco :) [20:37] the direct dependencies are resolvable on bionic [20:38] chrismo16: experiment. boot the HP without Secure Boot to Ubuntu, then do "nc termbin.com 9999 < <( efibootmgr -v)" and lets see if that "ubuntu-SB" entry is there [20:40] sim590: found: https://launchpad.net/~hnakamur/+archive/ubuntu/universal-ctags for 18.04. [20:41] TJ-its still there [20:42] termbin.com/nj9c [20:42] TJ-: Why would I be pulling everything from disco? I just do some pinning in preferences.d so that all packages are prioritized on bionic and I explicitely use "apt install -t disco package" for some packages, no? [20:42] I do that on debian with testing and unstable. [20:43] chrismo16: haha, do you see the problem? the path has no /s in it "EFIubuntushimx64.efi)" [20:43] Bashing-om: Yeah. I found that one before, but on debian, the package maintainer is Alessandro Ghedini, so I was inclined to download his on ubuntu also. [20:44] sim590: well I guess you could but it just seems like a sledgehammer to crack a nut, especially for a single non-essential package [20:44] TJ- haha. How do we fix? [20:45] when using encrypted LVM, can I use arbitrarily weird python or bash scripts to produce the would-be-keyfile bytes? [20:45] ssh with 2 factor auth works ! [20:45] chrismo16: I'm trying to figure out how the command I gave you did that, because I tested it here and it worked.. is it possible you mis-typed the / characters ? [20:45] like submit this file to that server via HTTPS and XOR the reply with this other file [20:47] I copy pasted [20:47] chrismo16: let's start by deleting that entry: "sudo efibootmgr -v -b 0 -B" [20:48] chrismo16: good, so it must be something else then! [20:48] chrismo16: ha, my fault, I didn't surround the path in single-quotes! [20:49] How do we edit? [20:50] Or do we delete and recreate [20:50] chrismo16: let's try again with " sudo efibootmgr -c -d /dev/sda -p 1 -w -L 'ubuntu-SB' -l '\EFI\ubuntu\grubx64.efi' " and once you've done that, do "efibootmgr -v" and ensure the path has \s in it! [20:50] chrismo16: as above, delete entry then create a new [20:53] Is there a network troubleshooting guide I can look at off line, between my network and what ever is going on with spectrum its dropping randomly and I need to restart the router and modem to reconnect again. [20:57] TJ-ok so that worked. I have booted with secure enabled and am at the blue screen to enroll mok [20:58] So basically I am at step 5 of https://github.com/jakeday/linux-surface/blob/master/SIGNING.md [20:58] chrismo16: so making progress ? [20:58] yes === im0nde_ is now known as im0nde [20:59] salamanderrake: I would probably use the method described here: https://www.linuxjournal.com/content/downloading-entire-web-site-wget to grab the guide found here: http://www.linuxhomenetworking.com/wiki/index.php/Quick_HOWTO_:_Ch04_:_Simple_Network_Troubleshooting for offline viewing [20:59] chrismo16: can I step down now, the boot issue is now solved? [20:59] Looks like it is [20:59] Thank you [21:00] chrismo16: good. Still puzzled how the system got into that state but ... *runs off and hides from the UEFI gremlins* [21:00] !cookie | TJ- [21:00] TJ-: Wow! You're such a great helper, you deserve a cookie! [21:01] Bashing-om: I'm going to read the freezer for ice cream to cool my brain down :D [21:02] TJ-: Good thought :) Ice cream cures many ailments. [21:11] genii: does that cover the new network tools, or still using the old network tools? [21:29] salamanderrake: Actually it does use the older ifconfig instead of the newer ip commands. But you can also still install the older ones [21:30] ..apologies on lag, it's end of day time at work and there was running around to do [21:34] genii: don't worry, I'm lagging because my ISP can't stay up. [21:35] salamanderrake: But as far as all the other diagnostic methods and applications, all of them are available normally on *buntu except tracepath is instead of traceroute [21:36] Oh, ok. [21:39] genii: https://dougvitale.wordpress.com/2011/12/21/deprecated-linux-networking-commands-and-their-replacements/ === katnp`[chc is now known as katnip` === msalvatore_ is now known as msalvatore [23:35] during the 18.04 installation of server I am prompted to import an identity is there any way to run your own identity management for this? I only see no / launchpad and 1 other. [23:37] oddtod: interesting idea [23:38] oddtod: my guess is that it'll hand whatever you input to ssh-import-id(1) and that can probably be extended easily enough.. [23:40] ssh-import-id only supports launchpad and github