[00:01] hi, where do I get libncurses++.so on ubuntu? [00:02] medfly: try apt-cache search ncurses [00:02] I have a lot of the ncurses packages [00:02] medfly: ... and search for appropriate libs [00:02] !find libncurses++.so [00:02] Found: W:, W:, W:, W:, W:, W:, W:, W:, W:, W: (and 373 others) http://packages.ubuntu.com/search?keywords=libncurses%2B%2B.so&searchon=names&suite=bionic§ion=all [00:02] !help [00:02] Please don't ask to ask a question, simply ask the question (all on ONE line and in the channel, so that others can read and follow it easily). If anyone knows the answer they will most likely reply. :-) See also !patience [00:02] !patience [00:02] Don't feel ignored and repeat your question quickly; if nobody knows your answer, nobody will answer you. While you wait, try searching https://help.ubuntu.com or https://ubuntuforums.org or https://askubuntu.com/ [00:06] hello [00:06] i have trouble enabling the xen hypervisor on ubuntu 18.04 with secure boot enabled [00:07] booting the os itself seems find [00:07] s/find/fine/ [00:08] when i installed grub-xen(or related) some options related to xen have been added to the boot menu [00:09] but when trying to boot those options, i'm directed back to the boot menu [00:09] without actually booting the xen kernel [00:09] so what's the problem? [00:10] booting xen means booting another OS that your dom0 manages to live alongside, guess they didn't get that signed [00:10] it's quite a mutation [00:11] medfly: yes, quite weird then [00:11] medfly: you mean that they didn't get the xen kernel signed? [00:11] medfly: that's likely [00:12] a-l-r: That particular configuration does not appear to be supported at this time. Xen itself can provide en EFI environment for the instances it hosts, but the currently provided packages in ubuntu do not include support for Xen to operate on a system with secure boot enabled. [00:12] pragmaticenigma: th [00:12] pragmaticenigma: thx [00:13] pragmaticenigma: got it, it's unsupported [00:15] hi,does anyone know how to specify a custom keymap in an aptitude config file? I have put the following in /etc/apt/apt.conf.d/mykeymap: [00:15] Aptitude::UI::Keybindings "CyclePrev" "C-h"; [00:15] but it doesn't work [00:15] C-h just opens help [00:18] so if i enabled secure boot, won't dkms work? [00:19] actually MOKManager can enroll something like custom keys, so dkms kernel modules that is custom built can be verified [00:19] figured out my problem, it's a thing failing to build, its not expecting the system to have it [00:19] however, virtualbox won't work either [00:23] Hello [00:24] Anyone can help with AppArmor? [00:26] nightowl1: what's your actual question? If someone knows the answer, they'll respond :) [00:27] I have ESET , which is being denied by AppArmor [00:27] I just can't configure it right [00:27] There is a command where you would choose to Monitor things by AppArmor and then you Allow them, but still ESET is denied [00:28] Why do I use ESET , It's because I send files to Windows Machines so sometimes I need to scan before I do [00:28] 01/16/2019 01:37:33 AM Preload library access control Cannot connect to /tmp/esets.sock: Permission denied [00:28] nightowl1: unfortunately, I don't believe ESET is supported here. There are AVs for Windows viruses that are, however, like clamav [00:29] nightowl1: otherwise, you may need to check with the ESET devs themselves [00:29] Their suggestion was to disable AppArmor which sounds weird. [00:29] ClamAV unfortunately do have a low detection rate [00:30] nightowl1: that hasn't been my experience. Do you have anything that has hard numbers to support that? [00:31] What do you mean SlidingHorn ? [00:31] nightowl1: you said that clamav has a low detection rate...I'm wondering where you got that information [00:32] It lacks behind companies like Kaspersky and ESET [00:32] [citation needed] [00:32] not by much [00:32] I don't have the websites where did I read that information but I did read it [00:32] Could be [00:32] I am not 100% sure [00:35] Either way, I'd probably stick with an AV that is open source (ESET is proprietary), that doesn't recommend disabling APPARMOR (like ESET apparently does), is free as in beer (ESET isn't), and whose site doesn't have XSS attempts. [00:39] I see , thanks SlidingHorn , I will give ClamAV a test [00:40] nightowl1: just one thing to note: ClamAV isn't a real-time AV - scans are run on a schedule or done manually. Just FYI [00:41] Ok as on demand scanner and not real time [00:41] What abot the XSS attempts? [00:42] My dell 7490 laptop has a usbC dock with one displayport and one hdmi output. In Windows I can use two external monitors in this setup. So far in linux I have only gotten one at a time (either one!) to work. When I have both plugged in, one doesn't even show up on `xrandr --listmonitors`. What are my next troubleshooting steps? [00:42] nightowl1, what is your end goal [00:42] nightowl1: the ESET website had multiple Cross Site Scripting attempts which can be used as injection attacks on a browser/web application [00:42] oops I understand [00:43] in xrandr the internal display is eDP-1 and the external displays are DP-1-1 and DP-1-2 [00:43] kk4ewt, I wanted to fix ESET , now I will try ClamAV [00:43] what is your reason for needing either [00:44] Scanning files before sending them to Windows machines because they might be infected or something [00:44] More like double checking [00:44] then clamav is fine for that [00:44] Will try it , thanks. [00:44] you are running a samba share so set clamav to run every 30 minutes [00:45] sparr: It could be an ACPI issue; We've had a couple of instances of the firmware not enabling GPU outputs when running Linux. [00:47] could that be the case while each output works alone? [00:49] Hello everyone [00:50] does anyone use Ardour (DAW) here? [00:50] apparently libjack is a dependency of ardour, but ardour seems to be compiled without jack support :O [00:51] i have pulseaudio running as a jack client (very nice setup to use both at once) and ardour won't even run in "alsa-over-pulse" mode :/ [00:51] with alsa it needs direct control of its audio interface [00:53] sparr: Unless you can show some log messages with clues we can only take wild guesses [00:53] Ok, I have a stupid question. [00:54] delt: Yes, that is correct. You can also have Ardour use jack [00:54] OneM_Industries: those aren't allowed here. NEXT! :P [00:55] TJ-: that's what im trying to figure out how :) [00:55] I have a system running 14.04 (Yes it's outdated), and I stuck a drive from a windows 10 machine into it to alter some partition sizes to be able to install ubuntu as a dual boot. Gparted is showing the entire disk as unallocated. [00:55] TJ-: someone in #ardour just mentioned that the ardour package in ubuntu repos might be outdated vs. the ones on ardour.org...? [00:56] OneM_Industries, possible the old gparted doesn't know about GPT [00:57] rfm: It also gave this error, so it's possible. https://pastebin.com/aTbNuLyp [00:58] OneM_Industries: 14.04 gparted knows about gpt [00:58] (I just got this laptop, so I have no clue. [00:58] ) [00:58] EriC^^: Ah, ok. [00:58] OneM_Industries: what does sudo parted -ls show? [00:58] usually the disk shows empty if its msdos with residue of gpt there, gpt stray data [00:59] sudo fixparts /dev/sdb should fix that [00:59] delt: Indeed, I installed from a PPA or built it myself. I use it for recordings for a Blind Society and its' really good. I can hop over a VPN and check if you'd like [00:59] https://pastebin.com/62YvX1Av [00:59] TJ-: they have a PPA on ardour.org? [00:59] OneM_Industries: is sda the disk in question or sdb? [00:59] * delt checking... [00:59] sdb [01:00] OneM_Industries: oh ok, it's empty right? [01:00] Nope. [01:00] Windows 10 install. [01:00] oh [01:00] (Not my choice, but I need it for a few programs.) [01:00] that's alright.. [01:00] (Thus why I would like to install ubuntu alongside it.) [01:01] OneM_Industries: what does "sudo gdisk -l /dev/sdb" show? [01:01] TJ-: from syslog https://gist.github.com/sparr/f8822c741f897569549442ff002f0860 [01:01] not sure what other logs might be useful [01:01] is it off of a uefi machine OneM_Industries ? [01:01] https://pastebin.com/PGxe1hpv [01:01] unfortunately the two monitors are identical, so it's somewhat difficult to tell the difference between them by resolution/size/etc [01:02] EriC^^: Let me find out... [01:02] OneM_Industries: seems it has mbr and gpt signatures [01:02] (Thinkpad 540p, if that helps.) [01:02] OneM_Industries: try choosing to list using MBR, i think it doesnt write anything it just lists the partitions found there [01:02] Yes, uefi. [01:03] oh, cause windows needs gpt for uefi if you were using that mode [01:03] https://pastebin.com/SXfd3c4n [01:03] Selected MBR. [01:03] OneM_Industries: try listing both partitions, if there's an efi partition it'd be a little clearer on what it's using [01:03] Ok. [01:03] aha [01:04] what about gpt? [01:04] Shows the disk as empty. [01:04] So, guessing that it's MBR. ;P [01:05] aha yeah i think so [01:05] Hunh. [01:05] the mbr partitions seem right, no efi partition, try mounting /dev/sdb2 perhaps [01:05] or just run sudo fixparts /dev/sdb and zap the stray gpt [01:06] OneM_Industries: if it cant find the partition, you could do "sudo mount -o offset=$((1126400 * 512)) /mnt" [01:06] Oh, there's part of it. Windows 10's hibernate when told to shut down feature again... [01:06] delt: I cannot recall if it was PPA or self-built now! And Sod's Law - I'm on kernel 5.0-rc2 right now and the VPN module isn't built! [01:06] aha mount -o ro,offset should let you mount it [01:08] TJ-: haha :D thanks anyway kind sir! i'm dl'ing the binary from ardour.org, getting a $1 / month subscription [01:10] EriC^^: Brilliant! Gparted can now see the partitions. [01:10] OneM_Industries: great [01:14] delt: I use it with a Delta 1010 I/O and UC-33e controller - really nice [01:16] TJ-: m-audio oxygen 88 here, and tons of guitars and guitar related stuff :D [01:24] delt: just got the VPN sorted; we're using the version in the Ubuntu 18.04 archive without problems [01:25] Hey, everyone, I'm trying to run a django command through cron but (think) I am encountering issues with my environment. If I `printenv` and output to a file, I see many fewer variables than if I run printenv from the shell. So I've tried adding `PATH=$HOME/.profile` to my script but see the same error which is that there's an error reading `/home/app/.env` from environ.py. [01:26] trying to do `xrandr --output DP-1-2 --mode 2560x1440` errors out with "xrandr: Configure crtc 2 failed" [01:27] mcaesa: not to say you won't possibly get an answer here, but I'd recommend asking in ##python, as it's more their specialty [01:27] I will ask there. Thank you. [01:28] sparr: are you using debian or ubuntu? Please stick to the channel for whichever distro you're using [01:29] Is there any tool that supports to repartition/shrink the size of the "ISO partition" once a hybrid ISO has been written to a USB stick? [01:29] Would like a separate partition following the written ISO, but not sure if it's possible. [01:30] ddybing: i guess you could modify the mbr etc so that it encompasses the whole disk and then create a partition after the iso's 2nd partition [01:30] ddybing: shrinking would be problematic since you'd need to recalc/rewrite some of the ISO9660 file-system metadata. For hybrid images though, you can tack more partitions on via the MBR and/or GPT that are embedded [01:31] SlidingHorn: feel free to coninue not helping me just as much as you were already not helping previously, and I will stick to distro-specific channels when I encounter a distro-specific problem. [01:34] sparr: I'll be willing to help when I find out which distro you're actually using. If you're using Debian, it's not supported here, and rude to cross post. Vice versa if you're using Ubuntu and asking in Debian's channel [01:35] I see, thanks TJ and EriC. Would there be a way to specify the size of the target without creating a partition? Thinking maybe dd-ing the exact size of the ISO to a specified block of the same size on the target device. [01:35] Your attitude makes me doubt your ability or willingness to help. I suspect you just have a copy paste rant prepared every time you see someone ask something in both channels (I also asked in #linux eventually). [01:36] I'll "cross post" in another channel after waiting about another hour, and no that is not rude at all. [01:36] Can't you have this fight elsewhere? [01:36] ddybing: if only [01:38] ddybing: it shouldnt write anything more than the iso's size if you dont specify anything [01:38] remember to use sync after dd or conv=fsync [01:39] TJ-: with jack? [01:40] delt: Yes, although later we switched to ALSA since it is easier to configure :) [01:40] TJ-: here ubuntu (mint 19.1) version info: Ardour5.12.0 (built using 1:5.12.0-3 and GCC version 7.3.0) [01:40] TJ-: can't use jack with this version [01:40] hi folks, i have got a problem with vlan using ifupdown in 18.04 lts. [01:41] delt: ardour/bionic,now 1:5.12.0-3 amd64 [installed] [01:42] my host has two interfaces. if i disable one nic and use the other with vlan, no problem at all. [01:43] as soon as i enable the other nic in /etc/network/interfaces, the one with vlan stops working. [01:44] any ideas? [01:45] partlycloudy: "stops working" doesn't really tell us anything. Please be more specific [01:47] TJ-: certainly. basically, i can no longer ping the ip on the vlan interface. [01:47] partlycloudy: can you show us the configuration? [01:47] SlidingHorn: if you insist on an ubuntu-specific question... due to the underlying xrandr failure mode I am pursuing here, the Displays settings panel exhibits misbehavior. Where/how might I file that as a bug report? [01:48] sparr: use ubuntu-bug [01:48] how can I find out the package name for the main Settings program? [01:49] (or specifically the Displays part, if they come in separate plugin-like packages) [01:49] sparr: which Desktop Environment? [01:49] default on 18.10 [01:49] TJ-: sure. thanks for helping out. here is the /etc/network/interfaces when eno1.100 works: https://pastebin.com/8mFT5mzq [01:51] sparr: (looking - haven't used Gnome in quite a while, so I'm a little rusty with its applications) [01:52] TJ-: and here is the version when the ip 10.0.0.11 on eno1.100 can no longer be reached: https://pastebin.com/bw7vP9cL [01:53] sparr: I believe that uses gnome-control-center [01:54] partlycloudy: what are these NICs connected to physically? the same switch, separate switches, router, another device? [01:54] TJ-: both NICs are connected to the trunk ports on the same switch. [01:55] partlycloudy: and is there any config for enp2s0 being done via scripts in /etc/network/if-up.d ? [01:56] Suppose I'm a upstream package maintainer and I've issued a new release of the software I maintain. How do I get the package updated in ubuntu? [01:57] Tried posting in #python and #django but no one seems sure about my issue. [01:57] TJ-: i didn't add any extra scripts in /etc/network/if-up.d. only default scripts are present. [01:57] More generally, how can I import a users env into a cron script? [01:58] sparr, maybe this is your fix, `xrandr --output eDP-1-2 --mode 2560x1440` # this page gave me a clue https://doc.ubuntu-fr.org/xrandr#ajouter_et_appliquer_une_resolution_non_listee [01:59] mind the e > eDP-1-2 [01:59] Just tried: `$HOME/.bashrc;` and then `printenv` after and the variables I need are not there. [01:59] partlycloudy: It *sounds* as if, when the 2nd interface is UP, the packets may be being routed out of the wron interface. So check "ip route show" for starters. If that doesn't reveal anything obvious, then use "sudo tcpdump -ni eno1.100 icmp" and on the remote host start pinging 10.0.0.11 - you should see the ping packet arrive and a reply sent out. if you see both then the problem is on the switch or other [01:59] host. If you don't see a reply leave, switch tcpdump to eno1 (the untagged if) and see if the replies are going out there. If not, check the enp2s0 ! [02:00] else i like to see 'xrandr' output please [02:00] treehug88: you generally won't get a new version into an existing release - unless it's a snap [02:01] treehug88: get it into the in-development release before Feature Freeze [02:01] treehug88: once there, it *may* qualify for an SRU to older releases (Stable Release Update) [02:01] TJ-: thanks TJ. working on it. [02:02] OerHeks: https://gist.github.com/sparr/13717d0d5f8a0885dd199d25c29e9100 (note it's DP not eDP) [02:02] partlycloudy: I've got a similar but different issue here with a switch/router/AP. Wired>Wifi or Wifi>Wired works fine, but wired>wired loses ARP/ND replies - the hardware literally eats them! [02:03] TJ- how do I get it into the in-development release? [02:04] treehug88: presuming the package is already in Debian/Ubuntu, best bet is to get it up to date in Debian before the last archive sync is done in the current development cycle. If it is an Ubuntu-only package then prod the maintainer to sync with upstream [02:05] TJ- debian, of course. The package is in sid, what's the process to get it updated in Debian dev? [02:05] treehug88: I'm not entirely sure aside from a bug report requesting the maintainer to update to upstream [02:06] TJ-: tcpdump on eno1.100 shows incoming traffic (echo request), but no echo reply being sent out. [02:06] ok thanks TJ- [02:07] treehug88: Here's an example: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=887465 [02:07] Debian bug 887465 in guacamole "Please update to new upstream version" [Important,Open] [02:07] partlycloudy: right, so now see if you can find it on another interface on that host === TheRedQueen is now known as Drone [02:07] partlycloudy: might be worth checking the 'lo' interface as well if all others draw a blank [02:10] TJ-: thanks TJ. just checked. no echo reply from enp2s0 or lo. [02:10] partlycloudy: did you test eno1 itself as well [02:10] partlycloudy: I should have asked earlier, but are any firewall rules being configured ? [02:11] partlycloudy: also, what subnet is the remote pinging client on? also 10.0.0.0/24 ? [02:11] TJ-: got echo reply from eno1. ufw is disabled. iptables is empty. [02:12] partlycloudy: aha! so the reply is being routed out on eno1. That would suggest the remote client is on the 192.168.0.0/24 subnet [02:14] TJ-: actually, i replaced the public ip with 10.0.0.11 in my post. the remote client is on a different public ip network. [02:14] partlycloudy: also, you've got *two* gateways configured, so check which one is the default (maybe their metrics are different). It very much depends on the actual client IP address [02:15] TJ-: there is something i dun quite understand from the ip r s. [02:15] sparr, DP1-2 else i have no clue [02:15] TJ-: the top two lines shows as: [02:16] default via 192.168.0.1 dev eno1 onlink [02:16] default via 192.168.0.1 dev eno1 proto static [02:16] partlycloudy: I suspect there may be an ordering issue that surfaces when you've got the enp2s0 i/f active that causes the default gateway order to change [02:16] * not without 'xrandr' output [02:17] partlycloudy: I'd need to see the entire routing table to make sense of it [02:18] TJ-: i see. you meant the full ip r s? [02:18] partlycloudy: yes; this feels like a routing issue but without the real data it's impossible to know [02:20] Are there security issues associated with allowing ~/.bashrc to be executed in non-interactive env? This would be for a user only used to serve a webapp. [02:22] TJ-: certainly. i changed the public ip (/27) with 10.0.0.64/27. the rest remains the same. [02:22] default via 192.168.0.1 dev eno1 onlink [02:22] default via 192.168.0.1 dev eno1 proto static [02:22] 192.168.0.0/24 dev eno1 proto kernel scope link src 192.168.0.11 [02:22] 10.0.0.64/27 dev eno1.100 proto kernel scope link src 10.0.0.67 [02:23] the ip on eno1.100 would be 10.0.0.67 [02:24] partlycloudy: so in your 2nd pastebin, there was a "gateway 10.0.0.1" ... I don't see that above [02:26] partlycloudy: Use the config without the enp2s0 i/f and see if the routing table is the same as above [02:26] I have to run. Thanks for trying to help! [02:26] partlycloudy: because I'd have expected to see another default route via 10.0.0.1 [02:28] TJ-: working on it… [02:31] TJ-: without enp2s0, it've changed to: [02:31] default via 10.0.0.65 dev eno1.100 onlink [02:31] default via 192.168.0.1 dev eno1 proto static [02:31] 192.168.0.0/24 dev eno1 proto kernel scope link src 192.168.0.11 [02:31] 10.0.0.64/27 dev eno1.100 proto kernel scope link src 10.0.0.67 [02:31] partlycloudy: Aha! [02:32] partlycloudy: so, for some reason, *something* is removing the default via eno1.100 when you add the enp2s0 and replacing it with that 'onlink' - there must be some scripts acting on that you've not noticed [02:33] hey question about building packages that have build dependencies that conflict with installed packages [02:33] namely: how do you do it [02:34] i obviously need to put the build dependency somewhere while the thing compiles but installing it means wrecking my system [02:34] partlycloudy: is it possible you've got systemd-networkd and ifupdown fighting here? [02:35] dorian, start looking at chroots [02:35] i hope ifupdown wins [02:35] partlycloudy: default for 18.04 server is systemd-network configs often via /etc/netplan/* YAML [02:36] TJ-: turned out it is… the systemd-networkd is still running… [02:36] partlycloudy: doh! [02:36] TJ-: so if i want to use ifupdown, it should disable the systemd-networkd, right? [02:36] *i should [02:37] partlycloudy: yes. although I'd use systemd-network given the choice :) [02:37] ugh chroot huh [02:37] dorian: see "sbuild" and friends [02:38] dorian: see http://packaging.ubuntu.com/html/setting-up-sbuild.html [02:39] looking [02:39] TJ-: i was following the documentation for openstack-ansible. it uses /etc/network/interface instead of netplan. some of the settings have no equivalent in netplan. [02:40] TJ-: so i switched to ifupdown. [02:41] partlycloudy: netplan is only really needed for orchestrated containers; for regular installs a static config is better I think ( /etc/systemd/network/ ) [02:42] TJ-: a million thanks TJ! :-D [02:42] partlycloudy: I still think you have a problem in your config though - there should not be two default gateways [02:43] TJ-: this looks straightforward enough [02:43] partlycloudy: the entire point is the default gw is used for addresses that don't match the known sub-nets [02:45] TJ-: i guess i'll remove the gw on 192.168.0.0/24. it was used for the initial setup via a router gw. [02:46] partlycloudy: if you need specific routes for other non-public sub-nets add those specifically, to avoid problems [02:48] TJ-: thank you so much TJ. i'd have never be able to solve it without your help. === mkv is now known as m4v [03:06] This is the htop screenshot after resuming from hibernation, how to debug this? https://imgur.com/a/hNRSEoV [03:13] why didn't you take a screenshot [03:13] gnome-screenshot -a [03:14] agio, system was unresponsive (look at the cpu usage) [03:14] :( [03:14] ah ok [03:16] marcanuy: holy damn, I've never seen a load avg so high. Im surprised you haven't cooked your cpu [03:16] how many cores does you CPU have? [03:19] agio, two cores [03:20] what does 'top' say, it shows zombies too [03:20] is your system still unresponsive? [03:21] agio, had to restart to avoid getting burn ;) [03:22] i cant make my ethernet on my dell e7440 laptop work with ubuntu 18.04 [03:22] OerHeks, htop should show them too in the status column as Z [03:23] lspci shows an ethernet controller, ifconfig doesnt, network manager shows no devices for ethernet [03:23] OerHeks, I'm also displaying user and kernel threads [03:31] what is "disco" [03:31] Emcy: disco is the upcoming 19.04 release of Ubuntu [03:32] oh [03:32] its listing packages already [03:34] Emcy: what do you mean? === mkv is now known as m4v [04:53] So, another stupid question. I can't get the wifi to work. [04:54] It's an intel 7260 chipset, and I can't seem to get it to connect to the wifi. It asks for a password, thinks for a bit, and asks for the password again, then the process repeats. === avelardi_alt is now known as avelardi [05:12] Has anyone had any luck following the openstack workstation deployment on ubuntu 18.10 server? [05:12] Having issues connecting to LXD after ubuntu install, and can't run lxd.migrate [05:22] chilli0: try asking in #ubuntu-server, too [05:31] @tomreyn, will do thanks [06:19] any idea why the menu in my editor "atom" is hardly visible, but only under ubuntu gnome (xubuntu it works fine) : https://discuss.atom.io/t/hardly-visible-menu-on-linux/61816/4 [06:24] pragomer: it uses electron, which handles font rendering in a somewhat different manner to regular GTK. [06:25] pragomer: perhaps you need to set some font settings it relies upon that are only set in xubuntu. [06:27] mm, you mean setting fonts in the app itself (atom) ? [06:27] or in a css file of atom? [06:28] do people still use iptables? [06:28] I known its a "hackable" editor ;-) but I am more a user and do unfortunatly not know where to hack this font problem. do you have an idea? [06:34] anyone here use iptables [06:34] pi0: only passivley with fail2ban [06:35] pi0: or on the webserver to save log lines: wp-login.php:/sbin/iptables -A INPUT -p tcp --destination-port 80 -s $REMOTE_ADDR -j DROP [06:36] apart from iptables what does ubuntu come with [06:36] iptables and firewall the same? [06:40] pi0: there is (g)ufw [06:41] those are frontends to iptables. there are many others. [06:45] ufw is front end for iptables [06:45] sorry but the default firewall for ubuntu is ufw? [06:45] new to ubuntu <--- [06:47] i think with 18.04 gufw is no longer installed by default, but i think it was in earlier releases. [06:47] by default there is no friewall. [06:56] i installed 18.10 [06:57] on my laptop man its nice [06:57] minimal gnome install [07:09] pragomer: No, not CSS. desktop font rendering settings. not the fonts. [07:09] pragomer: it may respond to older methods of configuring things like subpixel rendering. [07:37] Hello ! Is there a way I can monitor database connection(using python) to an mssql server ? I am on ubuntu 18.04 [07:38] Also I tried netstat -nat | grep while I was doing the query with no results [07:45] Hi, anyone using discord? I cant start it today, its nagging me about downloading an update (0.8).. but I thought it was supposed to be auto updated with snap? snap says I have 0.6 [08:14] sliddjur: for snap support, contact the maintainer of the snap please [08:16] sliddjur: i also dont see a 0.8 version of discord between the snaps? [08:17] Is there a good PDF reader for Ubuntu that allows me to change the default background colour? [08:17] Kira: inkscape can edit pdf's, but not sure about the background [08:17] reading long PDF documents in black-on-white is really tiring. :( [08:19] Kira: i think the original pdf always will show how its created [08:19] hey quick question, my laptop battery suddenly "moved" from bat0 to bat1 - any idea why that is? [08:20] lotuspsychje: when I start discord, it forces me to download update in a .deb package [08:20] sliddjur: where did you get discord in the first place? [08:21] lotuspsychje: snap [08:22] sliddjur: snap info discord, to see wich maintainer plz? [08:23] publisher: Snapcrafters [08:23] lotuspsychje, okular allows you to change the paper color [08:23] already much better than Evince. [08:24] I have a problem when I try to login as a user, I see the background for a second and then I get the login screen again. This happened after a recent base update. [08:24] sliddjur: that version shows 0.0.5 on my side, but please contact the snapcrafters for support [08:24] hi, does launchpad work for any of you guys with google chrome? since forever i get this error: https://cloud.neobits.nl/s/kwYyTDTJxDDAsTH i've disabled adblock, referer-block and the disconnect addon. but i always get this error (at least for 4 years). anyone else got this problem? and got it fixed? [08:25] alt-f2 is no longer bringing up a terminal. How do I get around this screen? [08:33] AquaL1te: for me it works, however i've got chromium [08:33] AquaL1te: but i also remember surfed it with google chrome and having never seen the error of your screenshot [08:35] How to solve assert.h not found problem even after installing build-essential and libc6-dev packages ? [08:35] AquaL1te: think ive seen such an error before, but i think it was when a user reported a bug with ubuntu-bug with chrome as default [08:36] tarzeau: i will check this, hopefully it works https://ubuntuforums.org/showthread.php?t=2355447 [08:36] lotuspsychje: i'm actually on fedora and want to file a bugreport for raspbian :) but launchpad is from canonical, so i came here [08:38] AquaL1te: but what i would like to know is, do you browse directly to launchpad or did you use ubuntu-bug? [08:41] AquaL1te: or you currently not on ubuntu? [08:44] lotuspsychje: i'm on fedora, no ubuntu software is running on my machine :) i created a new account, that account logs in fine... so i guess it's an internal problem of launchpad, that got reported everytime, but wasn't fixed. [08:46] try another browser AquaL1te as a test, we can also only support with ubuntu issues here [08:47] just tested with chromium and its also working fine like tarzeau says [08:47] lotuspsychje: browser is ruled out, it's a launchpad problem ;) i created a new account, which works fine [08:48] AquaL1te: why report a raspbian bug on launchpad? [08:48] ducasse: raspbian uses launchpad [08:48] huh, til [08:59] I have a problem when I try to login as a user, I see the background for a second and then I get the login screen again. This happened after a recent update. I can access and login to the terminal outside of the gui, booting to an older kernel does not help, moving .Xauthority is also not successful. [08:59] cYnIxX3: what kind of graphics card chipset & driver plz? [08:59] hum, laptop. [08:59] intel? [09:00] cYnIxX3: i dont know mate, we cant see that from here [09:00] thoughts of what i could check? [09:01] cYnIxX3: sudo lshw -C video or if you cant enter your system, lookup your laptop model? [09:01] one moment, let me bypass the gui and get that for you. [09:02] cYnIxX3: ubuntu version might also we usefull to know [09:02] *be [09:02] 18.10 [09:04] lotuspsychje, Intel HD Graphics 5500 driver=i915 [09:06] My appologies, 18.04.1 [09:06] is it possible to start a linux program with gtk2 instead of gtk3 ui ? [09:06] How would I attempt that? [09:07] or is that a general gui question unreleated to my inquiry? [09:08] cYnIxX3: as a test you could try !nomodeset to see if you can bypass login loop === lotuspsychje__ is now known as lotuspsychje [09:08] where do I set that? [09:08] !nomodeset | cYnIxX3 [09:08] cYnIxX3: A common kernel (boot)parameter is nomodeset, which is needed for some graphic cards that otherwise boot into a black screen or show corrupted splash screen. See http://ubuntuforums.org/showthread.php?t=1613132 on how to use this parameter [09:09] cYnIxX3: ill be afk for a bit, if it doesnt fix ask again in the channel ok [09:09] I have used this PC on linux for a few years. Its only after an update reboot that this started. [09:09] Thank you [09:09] cYnIxX3: maybe pastebin your recent dpkg log to the channel also, volunteers might see whats going on [09:32] hi [09:32] how do I create an archive which tar xvf file.tar.xz expands to several files instead of a single directory. [09:33] have all the files in the archive. [09:33] which cmd [09:33] nope all the directories as well, it should expand to right there. [09:34] tar options archive -C directory [09:34] not in file/ but the subdirs of file/ should be extracted to [09:51] Hi all. I've copied an Ubuntu BIOS installation to UEFI and installed grub-efi-amd64-signed and now it boots, but only with secureboot disabled in UEFI settings. What do I need to make it boot with secureboot enabled? [09:52] In other words, if anyone here boots with secure boot enabled, could you please run `dpkg -l '*signed*' | grep ^ii | nc termbin.com 9999` so that I see which package I'm lacking? [09:54] * alkisg tests installing shim-signed... [09:55] Oh great that was it :D [09:55] * alkisg waves [09:58] hello everyone-can anyone tell me if this "sudo apt-get -f install" will check for missing dependencies (ubuntu 18.04) [09:59] can paste uname if its relevent [10:01] all good sorted [10:06] it won't check, it will force install missing deps [10:18] BurekzFinezt, thanks man, is there a way to check to see if i have missing dependencies [10:22] not sure about that, but if you do, that command will solve it [10:22] jongi: aptitude will probably tell you [10:24] mkay so for that sudo apt-get update && sudo apt-get install aptitude -y (yes?) [10:25] if you don't have it [10:25] i'd not recommend aptitude these days. [10:25] then try: [10:25] aptitude search '?broken' [10:25] tomreyn, why not [10:25] agio, thank you [10:25] thank you both [10:26] jongi: it doesn't understand apt-mark's "hold" [10:26] tomreyn: does that prevent automatic upgrades ? [10:27] lol i dont either (can you explain that in noob) [10:27] agio: not from being attempted, but it could break dependency resolving. [10:27] you can hold a package at current version in aptitude [10:28] missing dependencies errors usually give 'fix with apt install -f' [10:28] im not sure im following what you're sayng? [10:29] what?! [10:29] -f == force? [10:29] some claim aptitude is superiour over apt, but i have seen no proof [10:29] -f fix [10:30] I prefer aptitude because it gives control over the depedency conflicts - I never run those automated fixes [10:30] yes, you can hold packages with aptitude, but that's not my point. there were traditionally two approaches to hold packages. aptitude understands only one of them, which can cause it to resolve dependencies incorrectly. [10:31] is that whats happening here '0 to upgrade, 0 to newly install, 0 to remove and 2 not to upgrade", from terminal after running sudo apt-get update [10:32] no, run sudo apt dist-upgrade in such event [10:32] !dist-upgrade [10:32] A dist-upgrade will install new dependencies for packages already installed and may remove packages if they are no longer needed. This will not bring you to a new release of Ubuntu, see !upgrade if that is your intention. [10:33] https://imgur.com/a/NrxxiWN [10:33] tomreyn: I see what you're saying. although an aptitude search '?broken' won't affect what packages get installed [10:34] roger [10:34] agio: i'm just saying it's not a good idea to recommend this as a general utility, and when recommended for a specific taks, it would be good to point out that apt(-get) should be preferred for standard tasks. [10:36] copy [10:36] still left with 2 not to upgrade https://imgur.com/a/52nbM2Z [10:37] how do i fix that [10:37] upgrade those two [10:38] (fyi not certain; or how to check, if those two are as a result of an install from .sh install of a git project clone/install [10:39] jongi: first of all it'd be good to know why they were kept back. which ubuntu release are you on (lsb_release -ds), which apt sources (apt update && apt-cache policy) do you have configured? [10:40] https://imgur.com/a/p6oVKva [10:40] not installed lsb that is [10:40] maybe a reboot needed, libsdl2 and libsdl2-dev kept back [10:41] jongi: you missed an underscore there [10:41] Ubuntu 18.04.1 LTS [10:41] sorry my bad [10:43] https://pastebin.com/4pzyHMxA [10:44] ran the other command re which apt policies [10:44] sorry apt sources [10:47] jongi: sudo apt-get --simulate -y -V full-upgrade [10:49] ls /var/run/reboot-required # will show if a reboot is required for the next updates [10:51] Hey, question about udev rules: Where do the attributes come from in the ATTR{} segment? [10:51] are they simply files located in the appropriate /sys/ node* [10:52] because the command "udevadm info /sys/" shows different parameters. [10:53] i.e. it lists ATTR{ID_VENDOR_ID} for a given tty node for a USB device, but the folder in /sys/ does not have a file named ID_VENDOR_ID; although some parent folder has idVendor file [10:55] tomreyn, OerHeks didnt work guys [10:56] https://imgur.com/a/sWNQeBg [10:56] so no reboot required, that is good [10:57] it kept the two back thou [10:57] gh [10:57] jongi: please run this: https://github.com/tomreyn/scripts#foreign_packages [10:57] +show output [10:58] do you want me to paste that into terminal or download install and then run [10:58] (clone) [10:58] jongi: you can download as the page i pointed you describes or you can git clone and run it, as you prefer [10:59] got it now had a read, will do [11:00] reading is generally recommended ;-) [11:06] Howdy folks [11:06] tomreyn, lols i know right-not sure if that went as planned but here goes [11:07] https://imgur.com/a/piDWahD [11:11] guys i have been ordered outta the house to get pizza by wifey, i will be back....and a big thanks for all your time [11:11] jongi: can you just show the output of: sudo /usr/bin/apt-show-versions | grep -Ev ' (uptodate|not installed|upgradeable to .*)$' | pastebinit [11:12] jongi: we can do this later, too [11:16] https://imgur.com/a/9VguBPV [11:17] its showing nothing i removed the pipe and got no output either [11:17] fyi i just installed pastebinit [11:18] says 'you are trying to send an empty document, exitiing [11:18] did i typo or is there another quick one you want pasted [11:18] in [11:24] What the heck is the difference between shift insert, and ctrl shift v (in this case when pasting into a terminal)? I get two different outputs. [11:26] Anthaas: there are two 'selections' - 'primary', which copies the selection (from the latest application with text selected), and pastes with middle mouse click, and 'clipboard' which requires explicit copy (ctrl-c) and paste (ctrl-v, in case of terminal shift-ctrl-v) [11:26] jongi: you did fine, i just wanted to make sure there's no output [11:27] hi all! I would like to know if I need to use bumblebee with my nvidia card. I have a GTX1060 and I don't know if bumblebee is preferred ? [11:30] caroga: bumblebee is outdated these days [11:30] !nvidia | caroga [11:30] caroga: 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 [11:31] no, bumblebee is old, prime is vurrent, it comes with your nvidia driver [11:31] c/vurrent/current [11:31] caroga: for gtx cards we reccomend adding the ubuntu graphics ppa [11:31] jongi: sudo apt-get install libsdl2-2.0-0 libsdl2-dev | pastebinit [11:31] jongi: actually: sudo apt-get install libsdl2-2.0-0 libsdl2-dev 2>&1 | pastebinit [11:32] he did update, so i wonder why the libsdl2 versions bite [11:32] me, too [11:33] thanks @OerHeks @lotuspsychje [11:33] hopefully the above will tell [11:33] @lotuspsychje, adding the nvidia driver from that ppa, does it use the dGPU stuff? [11:34] caroga: do you mean switch between powersaving vs performance mode? [11:36] @lotuspsychje, I think so. I remembered that you could set it to use the nouvaeu driver and have nvidia render certain apps, to save battery [11:36] caroga: yes, thats what OerHeks told you about, is called prime now and gets installed with the nvidia driver automaticly [11:37] caroga: after installing the nvidia driver, when you open your nvidia settings you can enable performance mode or powersaving mode [11:37] @lotuspsychje, will check this. thank you. [11:42] @lotuspsychje, thank you. I switched to powersaving mode. [11:42] I unplugged, let's see what will happen now. [11:42] caroga: wich driver do you have active now? [11:44] 1 proprietary driver - nvidia driver 390 [11:44] @lotuspsychje, ^ [11:45] the one from ubuntu or the one from ppa caroga ? [11:46] @lotuspsychje, ubuntu, I had already installed this from the additional drivers menu. [11:46] caroga: ok tnx for feedback [11:47] @lotuspsychje, is that good ? [11:47] Also, any other battery saving tools I can use? I remember something about TLP being a thing? [11:47] caroga, looks good, have fun with playing games [11:47] caroga: the choice of driver is good, when your card performs good [11:48] !info laptop-mode-tools | caroga it is [11:48] caroga it is: laptop-mode-tools (source: laptop-mode-tools): Tools for Power Savings based on battery/AC status. In component universe, is optional. Version 1.71-2ubuntu1 (bionic), package size 83 kB, installed size 378 kB [11:49] @OerHeks, thanks! I don't play games on it though. I'd rather switch to windows for that. It's more that it keeps heating up and wasting my battery life. [11:49] also it makes a lot of noise because of the fans. [11:50] Hey if Any of You Would Please Explain Why i'm Receiving adduser: Only root may add a user or group to the system. Error When Trying To Adduser As user Listed in /etc/sudoers [11:51] use sudo adduser [11:52] Nah i've Never Had a Problem Doing An adduser From a sudo LiSTED Path [11:52] OerHeks, [11:52] omg @lotuspsychje @OerHeks.... my laptop... it stopped making noise with the fans.... [11:53] ^. ש Not Winking [11:53] being listed in sudoers, does not mean you can drop sudo [11:53] enjoy it caroga [11:53] I am super grateful, thank you! [11:53] thats what we here for caroga np [11:53] Again @OerHeks Nah i've Never Had a Problem Doing An adduser From a sudo LiSTED Path [11:54] Never Had to sudo ir SU ROOT [11:54] iehusha, if you hacked that account to make it sudoless, i am not supporting that [11:54] @lotuspsychje, I would like to ask for your help on another issue I have. My mousepad is not working, never has (got this laptop 4 months ago) [11:54] iehusha, and you would know, would you not? [11:54] caroga: wich ubuntu version are you on? [11:55] I've tried installing synaptics [11:55] 18.04 [11:55] caroga: what exactly doesnt work on the touchpad? [11:55] @lotuspsychje, everything. [11:55] caroga: doesnt work at all ok? [11:55] correct [11:56] caroga: could you try this in wayland for us please? [11:56] Never Had to sudo ir SU ROOT"Not supporting it" [11:56] @lotuspsychje, sorry, I'm not up to date I think. What is wayland and how would I do this? [11:56] sudo su root ? [11:56] lolz [11:57] caroga: logout to gdm3 login window, and select ubuntu on wayland from there [11:57] caroga: not sure it will work on nvidia, but you can try as a test [11:57] OerHeks, " and you would know, would you not" What The Fuck is That [11:57] @lotuspsychje, okay. I would have to logout here as well. Be right back. [11:57] iehusha: please mind your language in the support channels [11:58] if you tweaked the way ubuntu user uses sudo, good luck [11:58] , SURE [12:00] Have Not one of You ever Added a User To sudoers File And Had The Full Root Accessies\ You Haven't And Why But The Matter is That its Never Complained Before [12:00] caroga: any luck? [12:00] @lotuspsychje, nope. [12:00] @lotuspsychje, I am guessing maybe the hardware isn't supported. [12:00] caroga: ok tnx for testing, what brand of laptop is that? [12:00] @lotuspsychje, it's a custom. Not a branded laptop. [12:01] being listed in sudoers, does not mean you can drop sudo [12:01] caroga: some brands of laptops have a little spot to touch, to enable touchpad/ or Fn button+ F key to enable/disable [12:01] caroga: do you notice anything like that? [12:02] OerHeks, YES IT DOES [12:02] @lotuspsychje, checking. [12:03] On Mine it Did Anyways [12:04] iehusha: it doesn't. [12:05] Thanks For the help OerHeks "and you would know, would you not" What The Fuck is That My Language Minds Me Sorry lotuspsychje [12:05] @lotuspsychje, I found what I think should be the button for it, but it's not responding [12:05] if you tweaked the way ubuntu user uses sudo, you would know? [12:06] caroga: maybe check your dmesg log, for acpi issues? [12:07] @lotuspsychje, dmesg shows unknown key pressed message after pressing the FN+F5 key [12:08] caroga: see any acpi issues in there relevant to touchpad? [12:09] @lotuspsychje, dmesg | grep acpi gives no specific touchpad related messages [12:09] ow [12:09] wait [12:10] caroga: does your touchpad show settings at the systemsettings/devices/mouse? [12:11] @lotuspsychje, no it doesnt. Also I run `dmesg | grep -i acpi` and it gave me more output [12:11] caroga: do you know if your bios version is up to date? [12:12] @lotuspsychje, I don't know that. [12:13] caroga: ok at this stage i would reccomend find out if your laptop has bios updates, and you could try a 18.10 liveusb as a test to see if touchpad works there, to exclude its a kernel issue [12:13] @lotuspsychje, https://gist.github.com/Caroga/b6eccfa24fee948bf4c5dd27567d6625 [12:13] @lotuspsychje, thanks, I will do so. [12:14] caroga: is your system up to date also? [12:15] @lotuspsychje, I should be yes. [12:15] allright, tnx for all the tests [12:15] caroga: if those doesnt work, come back and ask again here ok [12:15] I will. [12:15] thanks you so much @lotuspsychje. [12:18] caroga: installing inxi, then inxi -F might show you usefull info about your laptopbrand/bios version [12:19] thanks, i was looking for that name! I forgot that name of the package, was already searching through my installed list xD [12:19] @lotuspsychje, UEFI: American Megatrends v: N.0.05 date: 04/14/2018 <- was what I was looking for. [12:20] good job caroga [12:24] @lotuspsychje, I forgot to mention but this machine also has windows installed on it and the touchpad works flawlessly with it. Would this eliminate the bios as a bottleneck ? [12:24] caroga: its always wise to backup your data before you bios upgrade [12:24] caroga: also make sure its the right firmware version for your motherboard [12:26] @lotuspsychje, data is solid, no worries there. I have troubles finding the correct version so I've contacted the supplier. [12:26] @lotuspsychje, but, the touchpad working in windows, no indicator that it's not the bios? [12:27] caroga: ubuntu is kinda picky on acpi issues, so yes that still can be the case [12:28] !acpi | caroga [12:28] caroga: to debug ACPI issues on ubuntu make sure your bios is up to date and follow the procedure here: https://wiki.ubuntu.com/DebuggingACPI === daniel is now known as Guest84574 === TheHonorableKitt is now known as THKitten === Jubei_ is now known as Jubei [14:42] does ubuntu still do a livecd? [14:43] selphie_tilmitt: yes [14:44] @cryptodan is the livecd the ubuntu desktop 18.10, 1.9GB? [14:45] selphie_tilmitt: yes it is [14:45] ok, just making sure. thank you === daniel is now known as Guest97101 [14:56] i have a 18.04 server vm (kvm) htop shows its using 299MB ram but host shows its using 26% of host ram which is 32GB, how is this? [14:58] how to install php 7.0 on ubuntu 18.04? [15:00] ice9: https://linuxize.com/post/how-to-install-php-on-ubuntu-18-04/ # first result on google for "ubuntu 18.04 php" [15:00] ice9: if you're looking to install php 7.0 as opposed to 7.2, why? [15:02] sruli: how much total ram does htop shows ? [15:02] legreffier: in host 31.4GB [15:03] leftyfb, it's required to use 7.0 for application compatibility , and my question is about that specific version, not installing php in general [15:03] dom0 allocate all the requested space for the domU. [15:03] ice9: what application requires php 7.0 and doesn't work with 7.2? (a poorly written one you probably shouldn't be using) [15:04] that's the normal expected behavior. [15:04] ice9: https://tecadmin.net/install-php-7-on-ubuntu/ # first result on google for "ubuntu 18.04 php7.0" [15:05] ice9: ^ this is probably not in your best interest and might be unsupported here since it's packages from a PPA [15:05] sruli: you can tune the domU to use less ram, but its entire ram space will be preallocated. that's how it's supposed to work. [15:06] legreffier: it uses the entire allocated ram? i remember it only to do that for windows clients not linux clients [15:06] Hi I get Failed to start Snappy [15:06] !details | cswl [15:06] cswl: 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. [15:07] During bootup.. the systemd screen [15:07] cswl: what version of ubuntu? [15:07] 18.10 [15:08] cswl: what did you change that would have resulted in this? [15:09] Nothing. I just shut down normally [15:14] hi folks, i have multiple NICs and only one is assigned a gateway. still i have two "default" route in the routing table. is that right? [15:14] using 18.04 lts [15:14] partlycloudy: is one of them set to dhcp that's being given a default route? [15:15] leftyfb: nope. all are static. [15:16] i've disabled systemd-networkd and switched to ifupdown. [15:16] cswl: from what I understand, there's no snappy.service in ubuntu 18.10. Can you please explain in more detail your setup? I feel you're not telling us everything. [15:18] leftyfb: i use the same nic for native and vlan. the vlan is the one with gateway. [15:18] the two default routes are from the same nic. one is the vlan, the other is the native vlan. [15:20] partlycloudy: and both these configs are static and only one of them is set to have a gateway? Maybe confirm both configs are coming from ifupdown. Change up the ip to confirm. [15:22] leftyfb: both are configured using static. the native doesn't have a gateway, but its subnet still show as one of the two default routes. [15:24] Hey folks please visit my link [15:24] https://kidspaidmoney.com/r/c-sanchez === Class7_ is now known as Class7 [15:47] How can language screen be suppressed on Ubuntu 18.04 setup using preseeding. I was able to create a ISO file with 14.04 and make this work, but can't make it work for 18.04.1. [15:57] how do i open a nohup.out file from the command line? [15:57] tail -f nohup.out just shows me the end [15:58] reallymemorable: what is nohup.out? [15:58] it is a log of a process [16:00] reallymemorable: cat? vim? nano? less? more? gedit? [16:00] cat nohup.out? [16:00] reallymemorable: what exactly are you trying to accomplish? [16:00] cat worked [16:00] i just wanted to look at the text [16:01] reallymemorable: tail will also work. tail -f is follow and by default only shows the last 10 lines and any new ones [16:01] Anyone install Fuse in ubuntu server and use plexdrive [16:01] reallymemorable: less is better than cat for the purposes of scrolling up/down and searching [16:01] DexDeadly: that's not your question [16:04] how to make hexchat put nick into text box by doublelick? [16:05] vamadir: ask in #hexchat [16:08] Sorry, question was if anyone has had luck with it. Right now I believe I have fuse installed, apt-get install fuse and thats showing in the services as fuse3 [16:10] DexDeadly: so what is your actual question? Other peoples experience does not resolve your issue. Asking only prolongs your resolution as it's not getting to the root of your issue. [16:10] DexDeadly: so far nobody here knows what your issue is and cannot help you [16:12] The directions for plexdrive are rather vague, https://github.com/dweidenfeld/plexdrive/releases is where it says to get the package but no idea how to install it. Not a dpkg its not a tar for me to unzip. [16:14] DexDeadly, probably an appimage ? [16:15] DexDeadly: it looks like the downloads are directly executable, if you look at the README.md on the code page "Installation" section === capella is now known as capella|away [16:23] I'm using server can I just execute it then? [16:24] question, I'm currently in the following situation: if I remove libfoo-dev, libfoo gets removed as well. Is there some way to prevent this from happening? [16:25] reinstall libfoo? [16:26] Amnesia, usually a -dev pkg does not remove the main pkg ....afaik [16:26] sounds kinda cumbersome:/ [16:26] ioria: it's probably cause it got installed with the -dev pkg [16:26] Amnesia, what package, specifically? Is it getting removed or merely marked for autoremove? [16:26] Amnesia, details matter [16:27] hm, I'm actually using apt-get autoremove --purge [16:27] and in this case it are libhiredis[-dev] [16:27] which are installed through (mkosi and then in turn debootstrap) [16:28] Amnesia, yes, it's a dependency [16:30] so I guess there's no way around it, right? [16:30] Amnesia, you just reinstall libhiredisx.xx (if you need it) [16:31] (apart from some performing some nasty procedures) === SimonNL is now known as SimonNL_Afk [16:45] Amnesia: you could have run "sudo apt-mark manual libhiredis" before removing libhiredis-dev [16:45] tomreyn: noice, ty! [16:46] yw [17:01] I have an issue, a ubuntu 18.04.1 laptop that i have used for years is now stuck in a login loop. On user login it gets to the desktop screen for a moment but then logs out. I can access a non GUI terminal. last thing to happen before failure was a generic ubuntu update of some libs and kernels. Things that did not help are Booting on an old kernel, rebuilding .Xauthority, or rebooting. [17:02] cYnIxX3: can you type "cat ~/.xsession-errors" and paste it [17:02] I dont really have a way to paste. [17:03] copy* [17:03] cYnIxX3: type "cat ~/.xsession-errors | nc termbin.com 9999" [17:03] and share the link [17:03] Thank you, will do. [17:03] I am trying to emit a signal using dbus-send in following way to mimic InterfaceAdded signal sent by UDisk2 using following way but it is not working. [17:03] dbus-send --system --dest=org.freedesktop.UDisks2 /org/freedesktop/UDisks2 org.freedesktop.DBus.ObjectManager.InterfacesAdded int32:47 string:'hello world' double:65.32 [17:03] cYnIxX3: or, in your irc client, type "/exec -o cat ~/.xsession-errors | nc termbin.com 9999" itll put the link here [17:04] EriC^^, stuck for a bit, gotta get back to prompt, will post as soon as I have the link. [17:06] cYnIxX3: you can open another tty with alt+F4 or so [17:12] so my ubuntu server install (18.04 for a while, upgraded several times) has been unstable in that it loses connectivity to services, including SSH, ftp, plex, and some other services I have installed for media purposes. A duct tape hack which seems to be working is restarting it via cron every 12 hours or so, but as soon as that is removed it has s [17:12] ame issue as before. Any ideas how I could go about debugging? I tried looking into recent journal entries but nothing really standing out which can cause this [17:15] EriC^^, here is the dpkg.log with the latest packages installed before this happened http://termbin.com/6bav [17:15] cYnIxX3: try from a tty, "DISPLAY=:1 startx xterm" and see what it says [17:16] that command gave me a white squre in the top corner with a prompt that does not respond. [17:16] oh, I have a mouse, and when clicked could use it. [17:16] so, I think that worked? [17:18] cYnIxX3: yeah [17:18] so it must be something not x related and more display manager stuff related [17:18] cYnIxX3: if you create a different account does it work? [17:19] will test. [17:19] (trying to see if its some user config if we're lucky) [17:20] yes, adding a new user works. [17:20] OerHeks TJ-: I have resolved my monitor problem from last night. The root cause was bandwidth limits on the USB3.1-over-USBC port leading to my dock, which capped out before allowing two external monitors to run in my target resolution. [17:20] sparr: just replied to you in ##linux :) [17:21] cYnIxX3: the account logs in and all good? [17:21] yes, gui for a new account works. So what kind of user configs should I look at? [17:23] cYnIxX3: start with ~/.config ~/.cache ~/.gconf [17:24] cYnIxX3: although i'm not sure, if that were the problem maybe x wouldn't start either [17:24] cYnIxX3: are there any display manager logs you can check in /var/log/gdm3 ? [17:24] moved those 3 to a gui backup directory and attempted to login, no go. [17:25] cYnIxX3: anyways worst case scenario you can always copy the stuff somewhere, delete and create the user again and copy them back, it might come to that [17:25] EriC^^, /var/log/gdm3 is an empty directory. [17:25] aha [17:25] EriC^^, Understood, I appreciate the assistance and information. This gives me some more to play with now. [17:26] alright, no problem [17:27] Thank you [17:27] cYnIxX3: also maybe ~/.compiz if you're using compiz [17:28] cYnIxX3: and ~/.dbus [17:28] try ~/.dbus ;) [17:29] EriC^^: has cYnIxX3 done a search for non-user-owned files in the directory? [17:30] TJ-: no, cYnIxX3 might want to run "sudo find ~ ! -user $USER" [17:30] EriC^^: assuming currently logged in as the problem user of course :) [17:31] EriC^^: it kind-of sounds like the root-owned files problem we sometimes see [17:31] *shrug* [17:32] I'm in dire need of help regarding GRUB [17:33] TJ-: good day! follow up on yesterday… i noticed i had to remove or rename the extension of the yaml file in /etc/netplan/ to fully disable systemd-networkd. otherwise, it would resume to running state after reboot, even after systemctl disable it. [17:33] surprise_inquisi: ask away === SimonNL_Afk is now known as SimonNL [17:33] I created a reddit post, which is where I explained the situation. Please see it here: https://www.reddit.com/r/linuxquestions/comments/aghgzh/urgent_upgraded_ubuntu_packages_now_i_cant_access/ [17:34] partlycloudy: the way netplan.io works is at *boot* it 'renders' a configuration to /run/systemd/network/ - you'd need to *mask* systemd-networkd service to prevent it running, with "sudo systemctl mask systemd-networkd" [17:34] partlycloudy: and to stop it whilst it is running, "sudo systemctl stop systemd-networkd" [17:36] TJ-: yes, i initially masked systemd-networkd. but later on another server i noticed i could still disable the systemd-networkd without using mask, and it survivied reboot. [17:37] TJ-: on that server, i renamed the file to *.bk in /etc/netplan/ . that was the only difference. [17:39] partlycloudy: unless a service is masked, *another* task can cause it to start if it is only 'disabled' - disabled only stops it being directly started as I understand it [17:40] surprise_inquisi: OK, looked at your bootrepair log. My first question is, what is in /dev/nvme0n1p2 ext* file-system. From the size I suspect it is a separate /boot/ file-system for that drive's OS install [17:40] TJ-: ok. i will mask it on all nodes. just to be safe. [17:42] EriC^^, .dbus/ is not owned by user === mikecmpb_ is now known as mikecmpbll [17:44] @TJ I'll get you some screenshots [17:44] surprise_inquisi: Looks to be like boot-repair messed up the nvme0n1p2 file-system, which I now notice is/was another EFI System Partition [17:45] surprise_inquisi: no need, it's making since the more I read. I can see what has happened. You have/had in OS install on nvme0n1 that has full-disk encryption including GRUB_ENABLE_CRYPTODISK=y so GRUB's /boot/ is encrypted (inside nvme0n1p3 I'm guessing). [17:46] surprise_inquisi: Boot-repair looks to have broken that by replacing the grubx64.efi that presumably was locally built and included the cryptodisk module with one that doesn't contain that. [17:47] surprise_inquisi: Currently, at boot-time if you tap Esc repeatedly are you able to get to the GRUB boot menu? [17:48] @TJ Be carefull, there are NSFW ads on this site [17:48] https://unsee.cc/5aab89d9/ [17:49] surprise_inquisi: see my question above ^ [17:49] That shows the 2 locations, one encrypted, and one for the boot stuff [17:49] Oh, it automatically goes to the grub menu, no input required [17:51] surprise_inquisi: that's the first good thing then! Does it present any custom entries, either on the main page or after choosing the 'Advanced' sub-menu? I ask because I note that bootrepair looks to have tried to add entries to the files it backed-up from nvme0n1p2 which will be the ones that have the cryptodisk module in [17:52] I'm glad at least one of us is enthused! :) [17:52] surprise_inquisi: line 781 of your bootrepair pastebin shows it [17:52] boot-repair totally disabled the GRUB boot menu [17:52] I only got it back when I ran "sudo update-grub" [17:53] surprise_inquisi: line 787 also shows "Add /mnt/boot-sav/nvme0n1p2 efi entries in /etc/grub.d/25_custom" [17:54] surprise_inquisi: I think the best thing to do is talk you through how to boot the system into the LUKS encrypted OS one-time, from where you can use its own tooling to fix it. [17:54] surprise_inquisi: Are you able to chat on IRC from a different PC? [17:55] Yes, one second [17:55] And thank you, by the way, I really appreciate it [17:56] EriC^^, fyi: I have moved .cache, .compiz, .config, .dbus, .gconf, and .Xauthority, all users files are owned by user. Still stuck in the GUI login loop. [17:56] Once you're in channel from the other PC, reboot the problem PC to the GRUB menu and press 'c' to get to GRUB's command shell [17:56] surprise_inquisi: ^^^ last msg for you === Elon_Satoshi is now known as Copenhagen_Bram [17:57] Got it, morking on it now === daniel is now known as Guest61112 [17:59] what is the server address here? [18:00] surprise_inquisi: I'll summarise what we're going to attempt so you get an overview. From the working GRUB we're going to load its cryptodisk module and use the cryptomount command to unlock nvme0n1p3 which will show up in grub as (crypto/something-or-other) device, to which we can manually set 'root' and 'prefix', list the files in its /boot/ which will include the linux kernel (vmlinuz*) and initrd.img*, [18:00] write manual boot commands to load and execute those [18:00] surprise_inquisi: service address? for what, a regular IRC client, or the web interface? [18:01] Im trying to connect from another client, but it needs the server address for this channel [18:02] cYnIxX3: i see [18:02] surprise_inquisi: the server is chat.freenode.net, channel is #ubuntu [18:03] surprise_inquisi, use chat.freenode.org as server address, then type /join #ubuntu once the client connects [18:03] EriC^^, I appreciate the help and info none the less. [18:03] cYnIxX3: yw [18:04] Gotcha. back in 3 minutes [18:06] surprise_inquisi: try this, chat.freenode.net (204.225.96.251:6697) [18:06] cYnIxX3: EriC^^ I wonder if there's a problem with the /var/lib/ entries for the display-manager? I don't use gdm3, but lightdm has per-user entries there in /var/lib/lightdm-data/. Gdm3 may do something similar [18:06] Hello [18:07] My fstab using /dev/sdX etc as device names, and I just read that I should be using UUID as the kernel device numbers might change on upgrades etc. [18:07] Where can I get a list of my devices UUIDs [18:07] TJ-, I have a /var/lib/lightdm-data/USER directory, owned by user and lightdm group. [18:07] TJ-, however that file is empty. [18:07] ls -l /dev/disk/by-uuid/ is this the command? [18:08] I don't know enough to tell if this is correct, because isn't UUID a certain type of long number, but then some devices return a short number like so: 01CDD5CE8FF6FC40 -> ../../sdb1 [18:08] cYnIxX3: If the display-manager is gdm3 that directory is irrelevant, but I wondered if there might be a similar /var/lib/gdm3* directory/directories that may contain something affecting this [18:09] stoned: "sudo blkid" [18:09] stoned: blkid /dev/sdX1 [18:09] TJ-, how can I confirm which display-manager is used since I have both directories and they are both empty? [18:09] ok [18:09] cYnIxX3: I thought earlier you, or Eric, referred to gdm3. Maybe I'm remembering incorrectly though [18:10] stoned: example /dev/sda1: UUID="9d5d330f-d5d6-4a3c-8d1c-b47ce5eab65f" TYPE="ext4" PARTUUID="86688803-01" [18:10] Mine has shorter numbers [18:10] /dev/sda2: UUID="88622667622659E8" TYPE="ntfs" PARTUUID="d8a83730-02" [18:10] /dev/sda3: UUID="9dcac8e1-f557-4a4b-9ffa-fd6e1a554f39" TYPE="xfs" PARTUUID="d8a83730-03" [18:10] For example one is really long hexadecimal number, the other is a short number number [18:12] cYnIxX3, cat /etc/X11/default-display-manager [18:12] TJ-, I do have a /var/log/gdm3 directory, its empty and not user owned. [18:12] ioria, /usr/sbin/gdm3 Thank you [18:13] stoned: different filesystems, ntfs has a shorter UUID [18:13] cYnIxX3, if you have lightdm, then you upgraded from xenial or what ? [18:13] ok [18:13] ioria, yes, upgraded many times. [18:13] ok [18:13] What swapppiness do you use in kernel? [18:13] I have a 10 year old machine with only 8gb ram [18:13] thats a lot of ram. [18:14] for linux most of your ram is likely cached, if you use `htop` its easier to see whats used and whats cached in the bar. [18:14] stoned, ^^ [18:14] Well, 4GB of that 8GB is used as ZRAM compressed space for paging [18:14] You do the math. [18:15] still pleanty of space for ubuntu to run. ¯\_(ツ)_/¯ [18:17] swap_tendency = mapped_ratio/2 + distress + vm_swappiness; [18:17] The distress value is a measure of how much trouble the kernel is having freeing memory. The mapped_ratio value is an approximate percentage of how much of the system's total memory is mapped. And vm_swappiness is the swappiness parameter, set to 60 by default. [18:18] That makes more sense. cYnIxX3 ok thank you [18:18] If swap_tendency is below 100, the kernel will only reclaim page cache pages. Once it goes above that value, however, pages which are part of some process's address space will also be considered for reclaim. So, if the system's distress value is low and swappiness is set to 60, the system will not swap process memory until 80% of the total RAM in the system is allocated. [18:19] @TJ alright, And im back on a difference device, and at the GRUB CLI [18:19] I'm going to change this to 80 [18:21] some people study the use of the system before changing such values. [18:21] stoned: OK; give me a couple of moments. Start with "ls" to get the list of devices/partitions. Identify the device with the encrypted OS on (it only has 3 partitions I think) [18:22] TJ-: I got it. I was curious about diff sized UUIDs [18:22] I have 4 swap devices, I'm arranging the pririty right now. [18:23] TJ-, So I am using gdm3, why would the log directory be empty? [18:24] cYnIxX3: I'm no expert on Gnome; I've avoided it for 8 years! [18:24] intelligent. [18:25] do you use kdm? [18:25] cYnIxX3: xubuntu/xfce4/lightdm [18:26] One Wikipedia tab in chromium is taking 100MB of RAM [18:26] That is ridiculous. [18:26] Why did you take out the midori browser from 18.04 [18:26] stoned: Please don't fill the channel with observations. Ask questions, give answers, or keep the channel clear for others [18:27] surprise_inquisi: seems like I sent my last msg to the wrong nickname, grrr @ tab-complete [18:28] surprise_inquisi: Start with "ls" to get the list of devices/partitions. Identify the device with the encrypted OS on (it only has 3 partitions I think) [18:28] I just asked a question? [18:28] What is with the people in #ubuntu, does everyone like being a mall cop? [18:28] I literally JUST asked a question, and you're telling me to ask a qeustion. [18:28] I saw someone type a persons name right before they gave them advice, and someon botspammed him three times telling him to use a person's name before saying their sentence. [18:29] TJ: thats (hd1,gpt3) [18:29] Do you just have a habit of messing with people who are doing just fine? [18:29] Ignoring TJ now. [18:29] lol [18:30] yeah freenode staff is retarded [18:30] stoned: Why did you take out the midori browser from 18.04 [18:30] TJ- │ stoned: Please don't fill the channel with observations. Ask questions,... [18:30] Guy can't read. [18:30] Now I'm irritted by being trolled. Im going for a walk. [18:31] surprise_inquisi: OK, so now do "insmod cryptodisk" [18:31] Done [18:33] surprise_inquisi: now try "cryptomount hd1,gpt3" and you should get prompted for the pass-phrase. If you enter it correctly you'll see "slot X opened" [18:33] No immediately observable result [18:34] Should I try with parentheses? [18:34] I don't think anyone should use insmod or rmmod any longer. [18:34] Try modprobe instead. [18:35] Bacteria: it's not staff. Just some random. [18:36] modprobe intelligently looks for dependencies while loading a module. insmod won't load any dependencies. [18:36] surprise_inquisi: was there no response, it just returned to the shell ? [18:37] Correct [18:37] stoned: there is no modprobe, we're working in GRUB shell, not Linux === SimonNL is now known as SimonNL_Afk [18:37] surprise_inquisi: try "lsmod" ensure cryptodisk is listed === wildlander is now known as wildlander_ === wildlander_ is now known as wildlander [18:38] have an Acer Aspire 4750G with 18.04.1 - shutdown process stops and freezes. can anyone help? [18:38] The output of that seems to extended beyond the limit of the screen. Is there a way to move up? [18:39] surprise_inquisi: "set pager=1" [18:39] I do see "crypto 27" [18:39] cYnIxX3, that location it's no more in use [18:40] The "set pager=1" didnt appear to have an effect [18:40] surprise_inquisi: to your earlier question, no, the cryptomount device,partition does not need ( ) [18:40] ioria, where can I check gdm3 failures? [18:40] cYnIxX3, gdm logs to systemd journal [18:40] I am new to linux, I want to grant access to a folder to 2 different group ? [18:40] surprise_inquisi: Oh silly me! doh, we need to insmod other modules! [18:41] cYnIxX3, man journalctl [18:41] surprise_inquisi: "insmod luks" ... I'll look up the others [18:41] Oh, of course :P [18:41] hi. I just checked gparted. Does this mean I have everything under one partition, ie both system and ~/ ? http://i.imgur.com/1ieWQFd.png [18:41] I'm on debian and wanna switch to ubuntu, for multiple reasons. [18:42] hence asking about my ~/ [18:42] surprise_inquisi: "insmod gcry_rijndael" and "insmod gcry_sha1" then try "cryptomount hd1,gpt3" again [18:42] I see how pager works now. Cryptodisk and luks are showing now [18:43] surprise_inquisi: I'm so used to those modules built into GRUB core, I forgot them :) [18:44] bhanz: you'd likely need to use the Access Control List functions, see "man 5 acl" [18:44] okay [18:45] TJ-: Woot woot! I got "Slot 0 opened" [18:45] surprise_inquisi: we're rolling! Now "ls" and you should see a (crypto/xxxxx) device [18:46] surprise_inquisi: we know that device is a Linux root-fs, so we know it'll have a /boot/ containing the kernel/initrd files, so if you do "ls (crypto/XXXXX)/boot/" you should see some vmlinuz-* initrd.img* files amongst others [18:47] New devices; crypto0, proc, lvm/ubuntu--vg-swap_1, lvm/ubuntu--vg-root [18:47] surprise_inquisi: so, "ls (crpyto0)/boot/" [18:47] surprise_inquisi: but don't copy my typo :) [18:47] "Error: unknown filesystem" [18:48] surprise_inquisi: ahhh ... maybe "insmod ext2" and try again [18:49] Same error after that command [18:49] surprise_inquisi: Let me go check the bootrepair report [18:50] surprise_inquisi: no clue there; do you recall if there was LVM inside the crypto container? If so "insmod lvm" [18:51] Its still the same error after insmod lvm [18:51] RandomSerb: sda1 contains system and user data in one partition [18:53] surprise_inquisi: hmmm, I'm playing in a virtual machine here, let me see what else could be missing. Can you recall what file-systems you used. Could it be ZFS for example? [18:54] Isnt the default Ext3? [18:55] surprise_inquisi: default is an ext (ext4) [18:55] Or, could I just try adding all possible filesystems? [18:55] surprise_inquisi: but when using the FDE install mode it does LUKS > LVM > FS [18:56] Im fairly certainly when I installed Ubuntu, I just let it do its thing. I dont believe I specified any filesystems [18:56] surprise_inquisi: what does "ls (crypto0)" report ? [18:57] Device crypto0: No known filesystem detected [18:57] surprise_inquisi: OK, so we've not just got a bad path. check the ext2 module is loaded with "lsmod" [18:57] surprise_inquisi: and the lvm module too [18:57] Sector size 512B - Total size 498830336KiB [18:58] ioria, I see a line AVC apparmor="status" line with lightdm-guest-session, even though the display-manager said it was using gdm3, is this unusual? [18:58] lvm and ext2 both are loaded [18:58] cYnIxX3, timestamp ? [18:59] timestamp should be relivant to this issue. Is that what you mean? [19:00] surprise_inquisi: I'm wondering if the problem is not a missing module, but that the module required wasn't loaded before we unlocked the crypto device (when it would be scanned) - I'm pretty sure that isn't the case but...! [19:00] Does loading the module automatically load its dependencies? [19:00] surprise_inquisi: I'm checking the GRUB lvm module's source-code in case there's a command we could use [19:00] ioria, it was a few reboots ago, but i believe was on a boot that was experiencing the issue [19:00] surprise_inquisi: That's the general idea, yes [19:00] pavlos, any arguably safe way to repartition said disk so that I have my home dir in separate partition, so that I can overwrite debian with ubuntu? [19:01] surprise_inquisi: lvm auto scans, and should be triggered with a simple "ls" [19:02] Hmm... [19:03] surprise_inquisi: shall we start again just to be sure? [19:03] Sure, Ill restart [19:04] So load the modules, then unencrypt? [19:11] sparr, nice find, thanks for feedback [19:11] surprise_inquisi: indeed... sorry, trying to cook here as well as type [19:13] No worries man, dont let me hold you up :) [19:13] surprise_inquisi: got the notebook with me now; how's the 2nd attempt going? [19:15] See for yourself: https://ibb.co/m9ML2Kk [19:18] surprise_inquisi: If you did a standard install then it looks like you have al the required modules loaded, but at this point your earlier suggestion of loading other file-system modules is worth a shot [19:19] Dang. Do you happen to have a list of the file systems names? [19:21] surprise_inquisi: not easy if you want to read it, let me link you to a list on the web [19:21] Thank you [19:24] surprise_inquisi: it's to the source-code so you have to do some intelligent conversion of .c filenames :) http://git.savannah.gnu.org/cgit/grub.git/tree/grub-core/fs [19:25] By intelligent conversion, you mean simply dropping".c"? :p [19:26] surprise_inquisi: hopefully! Or if you're on a similar system now, you could do "ls /boot/grub/x86_64/*fs*.mod" but that won't get them all (e.g. ext2.mod) [19:31] surprise_inquisi: I've just had an idea! I wonder if there's another partition table in the crypto device (I'd be surprised but!) You've got part_gpt loaded but try "insmod part_msdos" then do "ls" just in case [19:33] I just added all of the filesystems in that page (some didnt exist). And Im still getting "unknown filesystem" [19:33] Ill try that suggestion [19:34] ls spits out the same partitions as before :( [19:34] surprise_inquisi: we're missing something obvious, I suspect it is a crypto module! [19:34] surprise_inquisi: as in the cryptomount unlocked the LUKS header but the master key uses some other crypto algorithm [19:37] Im at a complete loss myself. Do you have any ideas? [19:38] At this point, Im considering a full reinsrallation. Id juat hate to do that everytime I encounter this issue [19:39] Having ntp issues, googled a lot, no go [19:39] cognition: define 'ntp' issues [19:39] https://pastebin.com/zucW2se7 [19:39] cognition: dead link is dead, please use paste.ubuntu.com [19:40] surprise_inquisi: The only thing I can think is to boot into the hd0 OS and use cryptsetup to determine exactly what crypto algo/hash is required. that would at least confirm we have the correct modules loaded (or not) [19:42] https://paste.ubuntu.com/p/jCtwkvZrhF/ [19:42] How would I boot into hd0? [19:43] surprise_inquisi: as in "sudo cryptsetup luksDump /dev/nvme0n1p3" and check the "Cipher name" and "Hash spec" [19:43] surprise_inquisi: if you're in the GRUB command shell now just press Esc to get back to the menu, then choose the default entry I would assume [19:45] Hey all, I cannot seem to get inbound SSH packets to show up in the iptables dropped log. Here is a pastebin of my rules, they're fairly simple: https://paste.ubuntu.com/p/SDX3ywQ5FQ/ -- But when I SSH in from a different server on port 22, I do not see a dropped message at all. [19:45] Can anyone see anything I'm missing ? [19:47] TJ-: this is what I get with that; https://pasteboard.co/HWKT8Mh.jpg [19:48] surprise_inquisi: HASH! it's sha256 not sha1 [19:49] jwitko_, you are being helped in ##linux :-D [19:49] surprise_inquisi: so what we need is "insmod gcry_sha256" before loading cryptodisk [19:49] Alright, diving back in :) [19:49] OerHeks, i've posted here because no one has an answer there :( [19:49] they told me to double check my iptables rules... but I've checked them quite a bit here and have them available in the pastebin i linked above ( https://paste.ubuntu.com/p/SDX3ywQ5FQ/ ) [19:50] for the life of me i cannot understand why, for example, if I ssh into the server on port 22 I'm not getting a blocked message. [19:50] surprise_inquisi: such an annoying little but vital difference [19:50] Anyone have experience with RX 550 (laptop) and 18.04 ? [19:50] It seems that it's always underclocked while on battery [19:51] Even under usage it's underclocked [19:55] TJ-: same issue ; https://pasteboard.co/HWKWHia.jpg === capella|away is now known as capella [19:59] surprise_inquisi: it's possible we need other modules again too, but it's hard to know for sure because we have no diagnostic tools available in GRUB. I'm going to suggest an alternative. Boot into the default Linux and we'll set up a chroot to the broken install and use that to try to reset it [19:59] cYnIxX3, guest session on 18.04 with lightdm sounds weird, yes [20:00] Got it, im in default linux now [20:01] surprise_inquisi: unlock it with something like "sudo cryptsetup open /dev/nvme0n1p3 crypt_target" [20:02] surprise_inquisi: then show me "pastebinit <( lsblk )" [20:06] fresh install of server 18.04, wget a file locks the host. anyone ever experience this? [20:07] what directory are you saving to [20:07] compdoc: /root/ [20:07] http://paste.ubuntu.com/p/457hrzb7sj [20:08] is there enough space ? [20:08] yeah [20:09] compdoc: for reference, I'm trying to get the linux version of "prometheus": https://github.com/prometheus/prometheus/releases/download/v2.6.1/prometheus-2.6.1.linux-amd64.tar.gz [20:10] it legit LOCKS the host. [20:10] surprise_inquisi: that wasn't what I expected. When the LUKS device is unlocked it shold be scanned and file-systems/LVM discovered. It is starting to look like it may be corrupt [20:10] strixdio: what does "sudo file -s /dev/mapper/crypt_target" tell you? [20:11] oh flip! [20:11] sorry strixdio! surprise_inquisi: what does "sudo file -s /dev/mapper/crypt_target" tell you? [20:11] np [20:14] "/dev/mapper/crypt_target: symbolic link to ../dm-0 [20:15] surprise_inquisi: oh, of course :D "file -s /dev/dm-0" [20:18] surprise_inquisi: did you do something to that crypt device before all this started, it's proving to be very weird so far. [20:19] http://paste.ubuntu.com/p/jhx3DF5CBX [20:19] anyone know if there is a place to request applications be published as snaps? [20:20] I dony think so. I was using it fine for weeks, then I booted the other Ubuntu, updated/upgraded it, and noticed the GRUB boot loader got updated as well [20:20] surprise_inquisi: well that's progress. LVM, so it should have been auto-detected and the device-mapper nodes created if there is a volume group and logical volumes inside it [20:21] surprise_inquisi: try this: "sudo pvs /dev/mapper/crypt_target" -- should give you a Volume Group (VG) name [20:21] RandomSerb: you'd have to resize sda1, create a new partition with mountpoint /home, copy your data there and do all this from a liveCD using gparted. I would probably take a backup of my files in /home, then wipe the disk and install ubuntu, then bring back your data. [20:22] "pvs command not found" [20:22] I tried to install, and it wasnt found there either [20:23] surprise_inquisi: aha! possibly no lvm2 package installed "sudo apt install lvm2" [20:23] can anyone recommend a more simplistic image editor than gimp? [20:23] surprise_inquisi: that explains why right now it didn't auto-scan and create the device-mapper nodes [20:24] Alright! The previois command worked [20:24] surprise_inquisi: doesn't explain yet why GRUB's lvm module didn't do that though, which is still worrying [20:24] One sec, gonna switch to hexchat [20:24] surprise_inquisi: show me the single line, here, that it reports [20:26] is there a command line option to switch display managers for a user? [20:26] cYnIxX3: not sure you can do that for just 1 [20:26] http://paste.ubuntu.com/p/hhf4kbNr9d/ [20:27] okay, I'll go for systemwide. [20:27] cYnIxX3: you can try sudo apt-get install lightdm that would install it to the pc for all users [20:27] Thank you [20:27] np [20:27] humm, already newest version. [20:27] set it as the default dm [20:30] cYnIxX3, sudo dpkg-reconfigure lightdm (but why , if i may ?) [20:31] How do I execute *.sh files? [20:31] ioria, I am stuck in a login loop. I can sign in, see my desktop for a second, then get logged out. Adding a new user is able to login and see the desktop fine. [20:31] surprise_inquisi: sorry, battery exhausted! [20:32] mchelen: check Pinta [20:32] cYnIxX3, using what ? gdm3 or lightdm ? [20:32] both actually. [20:32] :( [20:32] cYnIxX3, nope... you can't :þ [20:33] TJ-, No worries :) [20:33] ioria, I cant what? [20:33] surprise_inquisi: at this point, with lvm2 installed, I'd expect "lsblk" to show the VG and any LVs [20:33] cYnIxX3, use both (at the same time) [20:33] surprise_inquisi: can you do "pastebinit <( lsblk; sudo blkid; )" [20:33] http://paste.ubuntu.com/p/KmfFbNPZM3/ === daniel is now known as Guest90532 [20:34] loltest: make it executable ... chmod u+x something.sh [20:34] http://paste.ubuntu.com/p/zvJt4c9XhJ/ [20:34] ioria, I am not attempting to use both at the same time. I switched from gdm3 to lightdm and the problem continued, [20:34] ah I see thanks! [20:34] That second link is the one you want. I got a little trigger happy [20:34] cYnIxX3, i see ... already checked xauthority and root owned files ? [20:35] ioria, I moved .Xauthority and that was not successful, All directories and files in /home/user are owned by user [20:35] cYnIxX3, ls -l /home/$USER | pastebinit [20:36] surprise_inquisi: right, lets see if we can wake the VG up! "sudo vgchange -ay ubuntu-vg" then do "lsblk" again to check you find some new nodes reported. If they are then pastebin the output [20:36] pavlos: ok thanks! [20:37] cYnIxX3, i mean the toubled user , not your current [20:37] *troubled [20:37] ioria, http://termbin.com/rpxh [20:38] cYnIxX3, user user ? [20:38] sed [20:38] cYnIxX3, user is your main $USER ? [20:38] "2 logical volume(s) in volume group "ubuntu-vg" now active" [20:39] ioria, no, i stream edited the output of the command to obscure the user. [20:39] http://paste.ubuntu.com/p/HX6zZKBdzG/ [20:40] cYnIxX3, i see [20:40] ioria, sorry for the confusion but I used: ls -l ~ | sed 's/$USER/user/g' | nc termbin.com 9999 [20:40] surprise_inquisi: Yay, finally! [20:41] I'm enthused if you're enthused! [20:41] cYnIxX3, did you edit backup .profile or .bashrc ? [20:41] cYnIxX3, did you edit .profile or .bashrc ? [20:41] I have no idea what any of this means though [20:42] surprise_inquisi: now we start motoring! "sudo mkdir -p /target; sudo mount /dev/mapper/ubuntu--vg-root /target; for n in proc sys dev dev/pts run etc/resolv.conf; do sudo mount --bind /$n /target/$n; done" [20:42] i have made some customizations to .bashrc I forget if I did profile, will check ioria [20:42] cYnIxX3, backup them, and re-login [20:42] ioria, negative, no .profile edits. [20:42] wilco [20:42] I ran that command, but there was no output [20:43] surprise_inquisi: assuming no errors (no output) then show me "pastebinit /target/etc/fstab" [20:43] ioria, negative, still looping. [20:43] http://paste.ubuntu.com/p/vfWQxt8f5s/ [20:43] cYnIxX3, something in /etc/lightdm/lightdm.conf.d/ ? [20:44] empty directory owned by root. [20:44] surprise_inquisi: :D going well. "sudo chroot /target mount -a" (again, expect no output) [20:44] cYnIxX3, what is .bash.conf ? [20:45] "mount: -: can't find in /etc/fstab." [20:45] ioria, one of my own scripts configurations. [20:45] ioria, junk. [20:45] Whoops, a typing error on my part. It worked (no output) [20:45] surprise_inquisi: darn, that's not supposed to happen! Did you type the command exactly as I showed? [20:45] I want to connect to my schools FTP server to access my files on it. I was using WinSCP on Win 7 before. Any recommendation for a similar program for Ubuntu? [20:46] surprise_inquisi: phew! OK, "sudo chroot /target" and now you're *inside* and using the broken install [20:46] loltest, filezilla? [20:46] loltest, terminal scp user@IP:directory/file host [20:46] great! it worked [20:46] loltest: is it an ftp server or an sftp server? [20:46] surprise_inquisi: first let's ensure the packages are all up to date: "apt update && apt upgrade" === msalvatore_ is now known as msalvatore [20:47] well that's a question I cannot answer right away [20:47] cYnIxX3, using what desktop ? the default gnome-shell ? [20:47] ioria, correct. [20:47] I will have a look at filezilla [20:47] cYnIxX3, so what is that .gnome folder ? [20:47] By the way, are you just doing this from memory, or are you using a resource? [20:48] surprise_inquisi: from memory; I do this all the time [20:48] How did you learn about this? [20:48] cYnIxX3, you should just have .gnome2 ... [20:48] surprise_inquisi: you may need to "apt install pastebinit" so we can gather some more data. Once you've done that show me "pastebinit <( ls -latr /boot/ )" [20:48] surprise_inquisi: breaking things :p [20:48] loltest: you can also use nautilus, the derfault file brwser, and once you'Re connected you can add a bookmark so you can quickly connect next time. or use the terminal. [20:48] ioria, I am not sure but it has a bunch of .gnome/apps/chome-*...*-Default.desktop files [20:49] surprise_inquisi: I've worked on kernel/GRUB development for about 15 years [20:49] by the way, i got an interesting waring when upgrading [20:49] update-initramfs: Generating /boot/initrd.img-4.15.0-43-generic [20:49] cryptsetup: WARNING: invalid line in /etc/crypttab for crypt_target - [20:49] cryptsetup: WARNING: invalid line in /etc/crypttab for crypt_target - [20:49] update-initramfs: Generating /boot/initrd.img-4.15.0-42-generic [20:49] cryptsetup: WARNING: invalid line in /etc/crypttab for crypt_target - [20:49] cryptsetup: WARNING: invalid line in /etc/crypttab for crypt_target - [20:49] !paste | surprise_inquisi [20:49] surprise_inquisi: 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. [20:49] cYnIxX3, backup it [20:49] ioria, I installed ubuntu on this several years ago and have upgraded a lot since, Its possible an upgrade didnt fully apply I suppose? [20:49] ioria, roger [20:49] surprise_inquisi: that's fine, we have to sort that out yet, because we chose our own device-mapper name that doesn't match what /etc/crypttab has [20:49] surprise_inquisi: you are currently muted, will be able to talk again shortly [20:50] tomreyn: thanks :) [20:50] ioria, negative login. [20:50] ;-) [20:50] surprise_inquisi: whilst you're muted, "pastebinit <( cat /etc/crypttab; ls -latr /boot/ )" [20:50] cYnIxX3, paste .xsession-errors [20:52] ioria, http://termbin.com/xmfa [20:52] http://paste.ubuntu.com/p/N7XpYdpXhT/ [20:53] How are yall direct messaging me? [20:54] surprise_inquisi, type a few letters and then tab. [20:54] and its not direct message, your just tagged in the message. [20:55] cYnIxX3, Did this work? [20:55] yep [20:55] surprise_inquisi: so now we need first to create a device-mapper entry for the LUKS device using the name in crypttab - once you've pastebinned the /etc/crypttab we can get on [20:55] cYnIxX3, Thanks! [20:55] surprise_inquisi: only about 4 more steps until we should be done [20:56] TJ-, Did I not already send it? [20:56] TJ-, http://paste.ubuntu.com/p/QWw4NZnJnZ/ [20:56] TJ-, Awesome to know we're on the cusp of the solution :) [20:56] cYnIxX3, i don't see a .dbus folderin your home [20:56] surprise_inquisi: OK so "ln -s ../dm-0 /dev/mapper/nvme1n1p3_crypt" [20:57] TJ-, Done [20:57] ioria, I was previously advised to move it. I will recreate it. [20:57] cYnIxX3, no need i guess [20:57] surprise_inquisi: then "update-initramfs -u -k 4.15.0-43-generic" and this time there should be no "cryptsetup: WARNING: invalid line in /etc/crypttab for ..." warnings (I hope) [20:58] created an empty .dbus to test anyways, problem presists. [20:58] Anyone? googling did not help: https://paste.ubuntu.com/p/jCtwkvZrhF/ [20:58] cYnIxX3, sy, heading out, maybe keep asking or come back tomorrow. good luck [20:58] "cryptsetup: WARNING: invalid line in /etc/crypttab for crypt_target -" [20:59] surprise_inquisi: drat. OK. let me think if we can hack this rather than having to exit the chroot and recreate it. [21:00] TJ-, I defer to your judgment, grand wizard [21:00] surprise_inquisi: let's try renaming the device-mapper name. First we have to delete the symlink I had you create: "rm /dev/mapper/nvme1n1p3_crypt" - then try "dmsetup rename crypt_target nvme1n1p3_crypt" [21:01] surprise_inquisi: if no errors reported then retry "update-initramfs -u -k 4.15.0-43-generic" [21:01] No output from last command [21:01] No warnings! [21:01] surprise_inquisi: eeek, did you notice a big potential bug there!?!? [21:02] TJ-, Hmm... something to do with symlinks, maybe? [21:02] When I run ssh-keygen or apt-get in a terminal, as soon as a prompt comes up (asking for save location or continue Y/n or similar) the process aborts and I am back at my prompt, as if I had pressed ctrl+C. What could cause that? It only happens in Terminal, not in xterm. [21:02] surprise_inquisi: we're actually working with /dev/nvme0n1p3 but in the install that device must have been detected 2nd, not first, so it got named nvme>>1< surprise_inquisi: won't hurt but could catch the unaware out :) [21:03] TJ-, So, should it be renamed? Or can we force it to be detected first? [21:04] surprise_inquisi: It's fine for now at least :). Anyhow, let's get GRUB sorted then we're done: "grub-install /dev/nvme0n1" [21:04] surprise_inquisi: the name is just text, it isn't actually used to find the device, but it could be confusing if someone forgot that (as I almost did!) [21:05] TJ-, Alright, installation finished! [21:05] TJ-, now that youve made it more apparent, it should make it easier to remember [21:05] surprise_inquisi: darn! and I got it wrong! I just pointed you at the other drive, arghh! which is likely how all your issues started [21:06] surprise_inquisi: let's just get *this* disk sorted for now! do "grub-install /dev/nvme1n1" (we are currently working with nvme1n1p3 !!) [21:07] TJ-, Good golly, this is intense [21:07] surprise_inquisi: better than a thriller movie :) [21:07] surprise_inquisi: once that is done for nvme1n1 do "update-grub" [21:07] TJ-, Its certainly raising my heart rate! [21:07] surprise_inquisi: that writes the /boot/grub/grub.cfg menu file [21:08] TJ-, It looks like that probably did it. It saw Drive A Ubuntu [21:08] test [21:08] surprise_inquisi: at this point, in theory, the install is now correct. But we need to leave the chroot and fix-up that mistake I just made. So, "exit" (now you're on the booted host again) then do "sudo grub-install /dev/nvme0n1" and "sudo update-grub" [21:09] surprise_inquisi: we also need to check if your having 2 separate installations has caused one to over-write the other in the PC's UEFI firmware menu. If so, we need to do additional steps to make them unique [21:10] surprise_inquisi: so, "pastebinit <( sudo efibootmgr -v )" [21:10] "Could not parse device path: No such file or directory" [21:10] surprise_inquisi: hmmph! was that for 'grub-install' ? [21:11] TJ-, No, that was for pastebin [21:13] surprise_inquisi: oh I bet the efivars or EFI-SP aren't mounted. Check "mount | grep efi" you should see 2 lines, one for efivars, the other for /boot/efi [21:14] ssh-keygen and apt are both behaving as if my shell is non-interactive [21:15] TJ-, http://paste.ubuntu.com/p/dR9Z8yFXBz/ [21:15] sparr, does `which ssh-keygen` reveal more than one location? [21:15] cYnIxX3: sorry, I fixed the problem by closing that shell and starting a new one, so I'm not sure what caused it or how to reproduce it. [21:15] congrats then. [21:16] surprise_inquisi: ouch, we've something wrong there. Let me check back in the other pastebins - there should not be 2 different /boot/efi's from the same disk [21:17] TJ-, Sure thing, I'll wait here :) [21:18] surprise_inquisi: Aha! now all you've experienced is becoming clear. You've got a total of THREE EFI System partitions (EFI-SP), when there should be at most 1 per disk, and preferrably one per PC [21:19] TJ-, Hmm... I'm not sure of the implications of that, or how to fix/prevent it [21:19] surprise_inquisi: in summary, the current nvme0n1 has 8 partitions including, apparently, 2 EFI-SPs (p2 and p6) and the target OS we've been fixing also has one on nvme1n1 in p1 [21:20] TJ-, My ideal is to have three OSes between 2 drives. DRIVE A: Windows 10 and Ubuntu 18.04. DRIVE B: Ubuntu 18.04 (Encrypted) [21:20] surprise_inquisi: so what is happening currently is the encrypted OS on nvme1 is using the nvmen0 EFI-SP, which means both Ubuntu installs over-write each other's boot files - explains your issue [21:21] surprise_inquisi: this would have happened when you install the encrypted OS I think and it chose an existing EFI-SP from the other SSD [21:21] TJ-, Well thats a huge pain in the ass, especially when everythign was working fine until I upgraded DRIVE A Ubuntu [21:21] surprise_inquisi: Anyhow, now we know, we can fix up. Several more commands I'm afraid but we can do it [21:22] TJ-, Awesome. I cant express how much much I appreciate your help [21:24] TJ-, So, whats our course of action? [21:26] surprise_inquisi: " sudo mkdir -p /mnt/p1 && sudo mount /dev/nvme1n1p1 /mnt/p1 && pastebinit <( echo "nvmen0n1p6"; find /boot/efi -ls; echo "nvme0n1p2"; find /target/boot/efi -ls; echo "nvme1n1p1"; find /mnt/p1 -ls ) " [21:28] TJ-, http://paste.ubuntu.com/p/XRcRr2gtfV/ [21:28] TJ-, Also, "find: ‘/target/boot/efi’: Permission denied" [21:28] surprise_inquisi: uhoh, might have to redo with sudo, let me check [21:29] surprise_inquisi: " sudo mkdir -p /mnt/p1 && sudo mount /dev/nvme1n1p1 /mnt/p1 && pastebinit <( echo "nvmen0n1p6"; find /boot/efi -ls; echo "nvme0n1p2"; sudo find /target/boot/efi -ls; echo "nvme1n1p1"; sudo find /mnt/p1 -ls ) " [21:29] surprise_inquisi: actually, no, stop [21:29] surprise_inquisi: " pastebinit <( echo "nvmen0n1p6"; find /boot/efi -ls; echo "nvme0n1p2"; sudo find /target/boot/efi -ls; echo "nvme1n1p1"; sudo find /mnt/p1 -ls ) " [21:29] http://paste.ubuntu.com/p/XXC3QDx59P/ [21:31] surprise_inquisi: OK, that shows the SSD with the encrypted install doesn't have the EFI bootloader files on it; it puts them on the other (Windows) SSD [21:32] What lines shows that? [21:33] surprise_inquisi: lines 269-270. No files. So I think we have two tasks. The first is relatively easy. Make the encrypted install use the EFI-SP on it's own SSD (currently nvme1n1). [21:34] surprise_inquisi: the other is to *change* the name of the install directory and EFI menu entry so the 2 Ubuntu installs do not both use the name "ubuntu" [21:34] surprise_inquisi: I need to check on this latter issue to be sure it is possible to configure, I've not come across it before [21:35] TJ-, Wouldnt that just be a custom GRUB menu entry? [21:36] TJ-, https://help.ubuntu.com/community/Grub2/CustomMenus [21:41] surprise_inquisi: no, this is slightly different. I think GRUB_DISTRIBUTOR= sets it but I'm checking the source. It is how GRUB constructs the string it passes to efibootmgr when creating the firmware menu entry I'm concerned about too. [21:42] surprise_inquisi: No, GRUB_DISTRIBUTOR only controls the entries in grub.cfg, not the ones that matter! [21:43] TJ-, Well of course, that'd be too easy :P [21:43] surprise_inquisi: give me 10 minutes to read the source code some more [21:43] TJ-, Take your time! [21:44] surprise_inquisi: It's controll by grub-install --bootloader-id ... but I need to find out how we can configure the system to always use the one we choose === capella is now known as capella|away [21:45] there was some bug a while ago where people showed up with /boot/efi/efi/linux and it was missing the grub.cfg there as well causing a grub shell [21:49] surprise_inquisi: OK, It seems like 'grub-install' is never called in an automated fashion (only update-grub is) so we can manually do it and it'll 'stick' - but if you ever need to redo grub-install you need to remember to use this option [21:49] TJ-, Ok, I'll certainly try to keep that in mind [21:49] surprise_inquisi: I need to make a coffee to lubricate my brain-cells then well sort out the EFI-SP file-systems so we can do this [21:50] surprise_inquisi: we can stick a wrapper script in the PATH that always uses this alternate bootloader-id on the encrypted install so you don't need to remember [21:50] TJ-, Thats fair, you've been working on this all day! [21:53] surprise_inquisi: whilst the drink brews let's check the nvmen1p1 is an EFI-SP. "pastebinit <( sudo sgdisk --info 1 /dev/nvme1n1 )" [21:53] surprise_inquisi: we're expecting the Globally Unique IDentitifer to be "C12A7328-F81F-11D2-BA4B-00A0C93EC93B (EFI System)" [21:53] http://paste.ubuntu.com/p/dpNvwY4Jc5/ [21:54] TJ-, Looks right to me :) [21:54] surprise_inquisi: good! it is as I thought. When Ubuntu installed the partitioner created this partition, but later, the installer mistakenly chose an EFI-SP on the other SSD [21:55] TJ-, Dang installer doin me dirty [21:56] I have come seeking aid regarding Bluetooth in Ubuntu 18.04. I've been struggling to correct the issue on my own for three weeks [21:56] surprise_inquisi: so I don't mess this up, show me again "pastebinit /target/etc/fstab" [21:57] http://paste.ubuntu.com/p/kBcPvvjNv8/ [21:59] Mythikos, what's happening [22:00] funtapaz, I cannot seem to get the system to recognize the adapter (a Gigabyte WiFi-BT combo device via PCIe) [22:00] surprise_inquisi: we're going to replace the UUID F21D-9FBF with the nvme1n1p1 UUID 0FBD-AE3C: "sudo umount /target/boot/efi && sudo sed -i 's/UUID=F21D-9FBF/UUID=0FBD-AE3C/' /target/etc/fstab " [22:00] model? [22:00] specific [22:01] It worked on Linux Mint 18.3, though not incredibly well, and was fine on Windows 7. Certainly. One moment [22:01] TJ-, Done [22:01] surprise_inquisi: check it changed yourself with "cat /target/etc/fstab" [22:02] surprise_inquisi: if you see "UUID=0FBD-AE3C /boot/efi ..." it changed [22:02] TJ-, Yep, lookin good! [22:03] surprise_inquisi: now we'll go back into that OS and add a grub-install wrapper, so "sudo chroot /target" [22:03] TJ-, Done === Tobbi__ is now known as Tobbi [22:04] surprise_inquisi: first we'll make the bootloader ID set from the grub config with " echo "GRUB_BOOTLOADER_ID="Ubuntu LUKS" >> /etc/default/grub " [22:05] surprise_inquisi: now we'll create a wrapper script, I'll put it in a pastebin so you can copy it over. [22:05] TJ-, Wait, is this going to be the name in the boot menu? [22:05] TJ-, "Ubuntu LUKS"? [22:07] surprise_inquisi: it'll get lower-cased and used as a directory name in /boot/efi/EFI/ and in the firmware boot menu itself as the choice [22:08] TJ-, In that case, it should be able to be any string, right? [22:09] surprise_inquisi: indeed, I've writing it that way. See the pastebin https://paste.ubuntu.com/p/q4xXnmFcyq/ [22:09] surprise_inquisi: create that file as /usr/local/sbin/grub-install and then do "chmod +x /usr/local/sbin/grub-install" [22:10] surprise_inquisi: I set the default - if no entry in /etc/default/grub - to be "ubuntu 2" [22:11] TJ-, I made naming changes, but everything else is the same [22:11] surprise_inquisi: that's fine :) [22:12] if you've added the file and made it executable try using it: "grub-install /dev/nvme1n1" then do "efibootmgr -v" and you /should/ see a new entry in the firmware's boot menu with your chosen name [22:12] TJ-, Running that echo command created a new terminal line, as though it expects further input [22:12] surprise_inquisi: you have unmatched quote marks then [22:12] surprise_inquisi: my fault too [22:13] surprise_inquisi: " echo 'GRUB_BOOTLOADER_ID="Ubuntu 2' >> /etc/default/grub " [22:13] surprise_inquisi: darn, another typo, scratch that! [22:13] surprise_inquisi: " echo 'GRUB_BOOTLOADER_ID="Ubuntu 2"' >> /etc/default/grub " [22:13] TJ-, I was gonna say, that doesnt look quite right :P [22:13] surprise_inquisi: I need that drink! brb [22:16] surprise_inquisi: OK, if you've got those added try the command as I suggested 9 lines above [22:17] funtapaz, I do not understand the question [22:18] TJ-, Looks like everything is good so far [22:19] funtapaz, http://susepaste.org/view//90914649 [22:19] surprise_inquisi: I think I forgot a step earlier... let me check [22:20] Mythikos, I meant the specific model of the card you were using. [22:20] There are still some hardware oddities out there that just flat out do not have support in the kernel. [22:20] Especially among older and cheaper oddpal PCI/PCI-e peripherals [22:20] I see. One moment [22:21] oddball* [22:21] Broadcom used to be notorious for poor support and some of their older stuff is still kinda garbage [22:22] funtapaz, https://www.gigabyte.com/Motherboard/GC-WB867D-I-rev-42#ov [22:22] surprise_inquisi: darn, yes I did, forgot to mount the new EFI-SP so we're written the boot files into the root-fs instead. Grrr. OK, lets fix that. "umount /boot/efi" (that should throw an error about no mount, which is what we want). Then "rm -r /boot/efi/*" to clean out the files, then "mount -a" and "mount | grep /boot/efi" to ensure /dev/nvme1n1p1 is mounted at /boot/efi [22:22] A recent purchase. My limitedresearch indicated it was Linux-friendly [22:23] TJ-, "umount: /boot/efi: not mounted." [22:23] surprise_inquisi: good [22:23] Mythikos, and you said ubuntu flat out isn't seeing the card at all? [22:23] TJ-, /dev/nvme1n1p1 on /boot/efi type vfat (rw,relatime,fmask=0022,dmask=0022,codepage=437,iocharset=iso8859-1,shortname=mixed,errors=remount-ro) [22:23] surprise_inquisi: after all those commands, if you've got /dev/nvme1n1p1 /boot/efi then redo "grub-install /dev/nvme1n1" [22:24] surprise_inquisi: then do "update-grub" [22:25] surprise_inquisi: and to check you now have a unique EFI menu entry: "efibootmgr -v" [22:25] funtapaz, I don't believe so [22:25] TJ-, So you want me to run those last 3 commands? [22:25] surprise_inquisi: yes, from "grub-install ..." [22:26] TJ-, I beleive so [22:29] surprise_inquisi: 'efibootmgr -v' should show the original 'ubuntu', the new alternative 'ubuntu 2' and presumably 'Windows xxxx' as well [22:30] TJ-, Yes, thats what its showing [22:30] TJ-, Although its using that actual given name instead of the "fallback" name [22:31] funtapaz, https://imgur.com/a/Scy7j0Y [22:32] https://www.maketecheasier.com/setup-bluetooth-in-linux/ Trying this out [22:33] On an 18.04 system... it's hard to tell but I think networkd isn't attempting prefix discovery at all... can this configured or is it straight-up not supported on the networkd version included in 18.04? [22:34] surprise_inquisi: good. So right the encrypted OS should be fixed, and have an entry on the firmware boot menu. As it was also added last it should show as first in the BootOrder as reported by efibootmgr [22:35] TJ-, Correct, its Encrypted Ubuntu, Ubuntu, Windows [22:35] surprise_inquisi: what you now need to remember is, the last OS to update the EFI boot menu will set itself as the default - so to boot another you have to interrupt the PC at start-up with whatever key brings up the manual boot menu (something like F10 or F8 usually) - this is entirely separate, and *before* tapping Esc to get the GRUB boot menu! [22:36] TJ-, Hmm... So, I wont be automatically shown the GRUB boot menu, where I can easily be shown the OS I want to boot? [22:36] surprise_inquisi: with that out of the way lets leave the chroot "exit". Now we'll clean up the now unused EFI-SP in nvme0p2 ! [22:37] TJ-, Sounds good! [22:37] surprise_inquisi: you've got to understand you've now got TWO levels of boot menu. The PC's own UEFI boot manager which selects amongst the entries you've added with 'efibootmgr' - Those lead to two separate installations of GRUB - 1 on each SSD [22:39] surprise_inquisi: Thinking about it, we should probably configure GRUB not include the other operating systems in its menus, so as not to get really confused! [22:39] No luck. Now reading https://github.com/blueman-project/blueman/wiki/Troubleshooting#no-adapter-or-not-able-to-turn-on-bluetooth [22:39] Okay, if I'm reading things right, this theoretically should be configured in /etc/netplan, but netplan doesn't actually support prefix discovery [22:39] surprise_inquisi: If you want to do this, we need to add "GRUB_DISABLE_OS_PROBER=true" to each OS's /etc/default/grub and re-run 'update-grub' for each ! [22:40] Or am I wrong on that [22:40] TJ-, Aww dang, I thought that was really convenient [22:40] surprise_inquisi: well you can stick with it, but just be VERY aware of it and if it leads to issues, come bac here another time and we'll isolate them [22:41] Well, amI likely to run nto this issue again when I upgrade either Ubuntu? [22:41] surprise_inquisi: let's carry on cleaning up the currently booted OS and nvme0n1 then you can *finally* test the darned thing [22:41] TJ-, Because if so, I'd rather disable as you suggest [22:41] surprise_inquisi: no, you shouldn't suffer this issue we've been fixing again [22:41] Ok, then lets clean and test :) [22:44] surprise_inquisi: so to summarise. On nvme0n1 there are several EFI-SPs! I need to be sure which are in use and I've got lost in all the pastebins so need to collect some fresh data. "pastebinit <( lsblk; cat /etc/fstab )" [22:45] TJ-, http://paste.ubuntu.com/p/MWBCYBXRPK/ [22:46] surprise_inquisi: so we're still inside the encrypted chroot. Let's leave it: "exit" then do the same command again :) [22:47] TJ-, "Could not parse device path: No such file or directory" [22:48] surprise_inquisi: erp! did the pastebin not create a link either? [22:49] TJ-, http://paste.ubuntu.com/p/MwPHhdKS2m/ [22:50] surprise_inquisi: that wasn't the command I expected ! "pastebinit <( lsblk; cat /etc/fstab )" [22:50] http://paste.ubuntu.com/p/qDrNKsjvZK/ [22:52] surprise_inquisi: notice how in the fstab the EFI-SP changed from the one used on install (different UUIDs and went from p2 to p6). I wonder if that was something bootrepair did [22:54] TJ-, I think it might have been my install order [22:55] TJ-, I installed as Windows > Ubuntu > Ubuntu LUKS [22:55] surprise_inquisi: but the efibootmgr shows Windows is set to find its files in p2 whereas ubuntu in p6 - so right now, if I recall the output of those 'find' commands, Windows cannot boot! [22:56] TJ-, I'm not sure where you're seeing that. But I accept that I am far removed from my element [22:57] surprise_inquisi: lets check what's in p2 and p6. "sudo mkdir -p /mnt/p2 && sudo mount /dev/nvme0n1p2 /mnt/p2" then do "pastebinit <( sudo find /mnt/p2 -ls )" [22:58] surprise_inquisi: don't worry this stuff is mostly intuitive to me. I'm trying to explain as we go so you can understand at least a bit, and more importantly, my reasoning for actions [22:58] TJ-, http://paste.ubuntu.com/p/96sGKzKwGY/ [22:59] surprise_inquisi: so, that shows Windows is in there (I must have mis-remembered about that from earlier). It also has all the bootrepair logs, plus an 'ubuntu' install which we should remove [23:01] surprise_inquisi: we'll leave the bootrepair/log/ for now just in case it is needed later. But lets remove the Ubuntu stuff. "sudo rm -r /mnt/p2/EFI/ubuntu" [23:01] Gotcha, I appreciate th ecommentary :) [23:01] TJ-, Done [23:02] surprise_inquisi: "umount /mnt/p2" now we're done with it so we don't get confused, then lets address p6 with "pastebinit <( sudo find /boot/efi -ls )" [23:02] surprise_inquisi: you'll need "sudo umount /mnt/p2" - forgot we're not in the chroot now [23:05] TJ-, http://paste.ubuntu.com/p/Z7qP8rjcDB/ [23:05] HI there [23:06] no-name-yet, was it a domain name for ubuntu project at the beginning? [23:07] ubuntu dot com [23:08] surprise_inquisi: nice and clean just as we prefer it. So one more to check, the one belonging to the encrypted install (which may have both 'ubuntu' and your alternative u-hc in it). "pastebinit <( sudo find /target/boot/efi -ls )" [23:09] http://paste.ubuntu.com/p/WWXS35Q7V2/ [23:10] surprise_inquisi: nice and clean, fantastic. let's do a few unmounts and then you can do a reboot test, then we're finished [23:10] a space in uefi name? [23:10] OerHeks: yes :) [23:11] did not know that is valid [23:11] * foldername [23:11] TJ-, Woot woot! I can see the light at the end of the tunne; [23:12] surprise_inquisi: "sudo umount /target/{boot/efi,boot,etc/resolv.conf,dev/pts,dev,run,sys,proc,}" [23:12] surprise_inquisi: that ^^^ should unmount everything currently attached to /target ! [23:12] surprise_inquisi: if no errors reported, so "mount | grep target" to confirm no remaining mounts there [23:13] Three weeks in and Bluetooth adapter still not detected by Ubuntu 18.04, despite being detected by Linux Mint 18.3 & Windows 7. I do not understand the issue [23:13] TJ-, Confirmed :) [23:13] surprise_inquisi: OK... acid test. Reboot, use the firmware's boot manager and choose your U-HA encrypted option and let us hope! [23:14] surprise_inquisi: it should boot by default but best to do a manual selection first :) [23:14] TJ-, Here goes nothin! I'll be back in a minute or two [23:14] I thought Linux Mint was based on Ubuntu, and both were better than Windows, but this is becoming ridiculous [23:15] Mythikos: how is the BT device connected internally, is it on a USB interface (I know it is part of a PCI wifi device, but mostly the BT side is exposed on USB) [23:15] PCIe device for WiFi with a short cable to the F_USB [23:16] Mythikos: so does the USB BT chipset show up with "lsusb" ? [23:16] Your question inspiries me to test a theory [23:17] Mythikos: even without drivers, lsusb will show all devices that will enumerate [23:17] TJ-, No dice :( [23:17] The option was there, but it dumped me into GRUB CLI [23:18] TJ-, http://susepaste.org/view//23925367 [23:18] surprise_inquisi: oh, that's not too bad! :D was it a "grub rescue >" or regular "grub>" shell prompt? [23:19] Mythikos: hm. the entries from Intel and the empty one seem the most worth investigating. [23:19] TJ-, Im not sure how to tell the difference [23:19] Maybe even the microsoft one. [23:20] surprise_inquisi: literally, the prompt you see on screen looks like the two strings I mentioned above [23:20] ryuo, Indeed. I think it is time to test that port, try an alternative F_USB, reboot, cha-cha-cha [23:21] TJ-, https://ibb.co/9VB8wFD [23:21] Mythikos: I cannot find the USB vendor for 28da in /usr/share/misc/usb.ids [23:22] TJ-, What are the consequences? [23:22] surprise_inquisi: well that is GOOD news. That shows that the crypto device unlocked (is that correct - did you get prompted for the passphrase) ? [23:22] Device 002: ID 28da:1102 seems to be a BCM20702A http://forums.debian.net/viewtopic.php?f=7&t=134276 and same fix https://dev-pages.info/ubuntu-bluetooth/ [23:22] ryuo, Mythikos ^^ [23:23] TJ-, Unfortunately not. I wasnt prompted for a passphrase [23:23] Mythikos: well for a driver to load, the vendor:product ID (28da:1102) needs to match a modalias [23:23] why does that vendorID/productID look familiar... [23:23] https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1065400 [23:23] Launchpad bug 1065400 in linux (Ubuntu Trusty) "Support for loading Broadcom bluetooth firmware" [Undecided,Confirmed] [23:24] old bug, still in use [23:24] surprise_inquisi: So the GRUB that loaded could be the wrong one. Tell you what, to make following easier, lets move to a separate channel. Do /join #luksfix [23:25] Broadcom's usually 0a5c - at least it is on my Dells [23:26] I've just done update-usbids and it still doesn't have 28da [23:26] I guess it is just confusing that it worked in LM 18.3 [23:27] I just switched, actually [23:27] Mythikos: it is strange. Can you show us "pastebinit <( lsusb -v -d 28da:1102 )" [23:29] http://susepaste.org/view//23701073 [23:29] Is that helpful? [23:30] how can i open the settings pane with the default ubuntu windows manager? [23:30] Mythikos: aha! well, still don't know who the 28da is! But, the device contains USB device class HIDs so it is one or more input devices (likely keyboard/trackpad) [23:31] left down icon on the panel, opens the menu, and you can search/start systemsettings [23:31] you might want to install gnome-tweak tool too [23:32] OerHeks if thats aimed at me, then i hear you, but i'd like to open the default settings pane via the command line and force it to use gtk or whatever [23:32] or gnome, rather. [23:33] TJ-, My keyboard uses one port for power, the other for data [23:33] Perhaps that is it. Works fine though too [23:33] duoi, interesting [23:34] i kind of confused myself [23:34] i was basically after this [23:34] env XDG_CURRENT_DESKTOP=GNOME gnome-control-center [23:34] Mythikos: show us "pastebinit <( lspci -nnk )" so we can see the manufacturer of the associated wifi device [23:37] TJ-, http://susepaste.org/view//66862113 [23:39] Mythikos: oh, so it'd be Intel! "Intel Corporation Wireless 8265 / 8275 [8086:24fd] " [23:41] duoi, not sure why your solution would not work, Exec=env XDG_CURRENT_DESKTOP=GNOME gnome-control-center --overview [23:42] Mythikos: so the USB device is 8087:0a2b which looks to possible use the USB Bluetooth class driver, at least accoridng to https://certification.ubuntu.com/catalog/component/usb/4103/8087%3A0a2b/ [23:42] OerHeks it did work, i had just forgotten it [23:42] Mythikos: so now the question is whether the device has been set up: "pastebinit <( ls -l /sys/class/bluetooth/* )" [23:45] TJ-, https://www.gigabyte.com/Motherboard/GC-WB867D-I-rev-42#ov [23:45] http://susepaste.org/view//49761483 [23:46] Sorry for the delay. Had to do some digging [23:46] Mythikos: good news, the Bluetooth device is there. Now check its radio isn't disabled with "pastebinit <( rfkill list )" [23:47] Did that earlier. http://susepaste.org/view//99483842 But worth checking === [1]MrMobius is now known as MrMobius [23:51] Mythikos: looking good so far. Now do "bluetoothctl" then do "help" understand what commands available, then do "show" to report the controller info, then "discoverable on" then "scan on" and see if it can detect other devices [23:52] TJ-, http://susepaste.org/view//31219138 [23:53] TJ-, Same thing with "scan on" [23:53] Mythikos: OK, do "quit" [23:53] Mythikos: I think the BT firmware is missing is the issue [23:54] jups [23:54] That was my thinking, but I'm not sure how to remedy that on here [23:54] Mythikos: "pastebinit <( journalctl -b 0 -p warning )" [23:54] or 'firmware' [23:56] TJ-, http://paste.ubuntu.com/p/cf89DGZJxJ/ [23:57] Mythikos: might be worse than missing firmware: line 26 "kernel: Bluetooth: hci0: Reading Intel version information failed (-110)" [23:57] :/ [23:59] line #8 powernow_k8 .. https://askubuntu.com/questions/1064450/powernow-k8-on-lubuntu-18-04-is-flooding-my-kernel-log-with-cpu-frequency-errors [23:59] maybe this cpu is the culprit [23:59] Mythikos: you've also got a serious-looking kernel panic there from the Wifi device itself, as as they are associated I'd suspect both issues are related [23:59] TJ-, From my amateur perspective, this output of that command implies the whole system is broken, and ready for scrap. lol