[00:57] <oerheks> = rescan?
[01:23] <goddard> i have a gpdwin 3 and i installed ubuntu on it.
[01:24] <goddard> but screen orientation won't change even though i could change it just fine when "Trying Ubuntu"
[04:31] <Samian> Which term do you prefer? "Terminal", "Command line interface",
[04:31] <matsaman> Samian: to refer to what
[04:32] <matsaman> to who?
[04:32] <matsaman> I think 'terminal' and 'command prompt' are more general and useful to most people
[05:35] <Guest59> I want to disable copy past shortcut (using command) ? without installing any additional application.
[07:44] <kieto> Hi! Has anyone encountered this libpython error when upgrading an ubuntu 20.04.2 LTS? https://imgur.com/a/MyHEsIc
[07:48] <kieto> nvm, fixed it with `apt install -f`
[09:00] <rob> I've lost the handle of the people who helped fix my DNS problems yesterday, but thought you might like to know that I reported this to Proton, who have acknowledged a bug in their client and are working on a fix.
[09:05] <Guest22> hellp
[09:05] <Guest22> when i connect ethernet wifi will disconnect
[09:05] <Guest22> any idea ?
[09:19] <frib> My wifi gets disabled randomly in Ubuntu 20.04. I have tried all manner of power management settings configuration and nothing seems to work.
[09:20] <frib> What can I do to monitor for events that are causing wifi disabling?
[09:25] <rob> frib: I had a similar problem recently with a Dell XPS with Killer wifi, which is notoriously unreliable. In the end it turned out an unfixable hardware problem and I had to get a new machine on warranty.
[09:25] <rob> In my case the problem was that the wifi card would just entirely vanish from the system for a while, even after reboots, then just as mysteriously it would return again.
[09:32] <frib> rob, mine doesn't appear to be a hardware problem as it visibly gets disabled in Network Manager and I can simply manually turn it back on again
[10:27] <theMajor> goodafternoon
[10:28] <lotuspsychje> welcome theMajor
[10:30] <madhan> Hi
[10:31] <madhan> i want to config gui and vnc on ubuntu server
[10:31] <madhan> server is Oracle cloud instance
[10:33] <p3rL> hello
[10:34] <lotuspsychje> welcome p3rL
[10:34] <p3rL> how do i fix or avoid this error
[10:34] <p3rL> -bash: warning: setlocale: LC_ALL: cannot change locale (cs_CZ.UTF-8)
[10:34] <lotuspsychje> madhan: please no crosspost on several ubuntu channels
[10:34] <p3rL> when ever i login my server it get this error
[10:34] <p3rL> i am getting this on my ubuntu server
[10:36] <madhan> lotuspsychje yes i tried
[10:36] <toddc> p3rL: https://askubuntu.com/questions/89976/how-do-i-change-the-default-locale-in-ubuntu-server
[10:38] <theMajor> ok the big question and don't get angry for asking but ubuntu vs popos, why should one choose ubuntu over popos
[10:38] <p3rL> toddc thanks
[10:38] <p3rL> letme read this
[10:39] <cbreak> theMajor: support by third parties for example
[10:40] <cbreak> native zfs for example
[10:40] <lotuspsychje> start the topic in #ubuntu-discuss if you like theMajor
[10:41] <cbreak> various client / server variants, and so on.
[10:42] <iomari891> is there a sleep alternative that will break with any key stroke and not just ctrl-c?
[11:02] <sixwheeledbeast> I assume you mean in bash? a loop looking for a keystroke would be an alternative.
[11:05] <sixwheeledbeast> you can use $SECONDS in the loop for the sleep time element.
[11:07] <sixwheeledbeast> you could also use `date` if a specific time is important not elapsed time.
[12:25] <goddard> i got an nvme drive that is dying
[12:26] <goddard> what is the safest way to try and copy the disk to a file?
[12:26] <goddard> i drive gnome disks but it just craps out on me
[12:27] <lotuspsychje> goddard: try a photorec recover on it
[12:28] <lotuspsychje> save as much data as you can
[12:36] <goddard> ill give it a shot
[12:36] <goddard> not super critical as its just a game save i want off it
[12:36] <goddard> but it would be nice to not have to play that far
[12:39] <stevenm__> hey at some point I told a ubuntu 18.04 box to stop loading the graphical UI on startup ... and now I can't figure out which service needs re-enabling and starting
[12:39] <stevenm__> I'd have thought lightdm - but it doesn't seem to be there
[12:39] <goddard> gdm ?
[12:39] <oerheks> something with graphical target ...
[12:40] <stevenm__> root@prometheus:/lib/systemd/system# grep -RIi graphical .
[12:40] <ioria> sudo systemctl set-default graphical.target
[12:40] <stevenm__> yeah I looked at that :P
[12:40] <stevenm__> aha!
[12:40] <oerheks> ioria, +1
[12:40]  * stevenm__ tries rebooting
[12:42] <stevenm__> nah that doesn't seem to have done it
[12:42] <stevenm__> graphical.target                                                                             loaded active active    Graphical Interface
[12:42] <stevenm__> active  - but not running
[12:43] <ioria> stevenm__, so you did something else; can you restart gdm with  sudo systemctl restart gdm ?
[12:43] <stevenm__> ioria, according to 'dpkg -l' this system is lightdm, not gdm
[12:43] <ioria> ubuntu 18.04 has gdm
[12:44] <stevenm__> and 'systemctl enable lightdm' just gives me a warning
[12:44] <oerheks> systemctl enable gdm
[12:44] <stevenm__> ah sorry - good point, this is actually ubuntu mate 18.04
[12:44] <oerheks> oh
[12:44] <ioria> stevenm__, sudo dpkg-reconfigure lightdm
[12:44] <stevenm__> k - it did... something? returned me to the prompt anyway
[12:45] <oerheks> no message = ok
[12:45] <stevenm__> I know :)
[12:45] <stevenm__> doesn't mean ok means it did anything useful :P
[12:46] <ioria> stevenm__, how did you 'tell' ubuntu to not load the graphical UI  ?
[12:46] <stevenm__> no idea - was probably over a year ago - or more
[12:46] <stevenm__> aha! graphics on boot!
[12:46] <ioria> stevenm__,  cat /proc/cmdline
[12:46] <theMajor> stevenm__: did you manually adjusted the runlevel of the system
[12:46] <stevenm__> that did it :)
[12:47] <theMajor> stevenm__: what was the solution?
[12:47] <stevenm__> possibly that dpkg-reconfigure of ioria's ?
[12:47] <stevenm__> it's booted straight into the desktop though - no login screen
[12:47] <stevenm__> which is ... odd
[12:48] <oerheks> oh, auto,ogin
[12:48] <theMajor> stevenm__: that makes total sense. Hmm maybe check if autologin is enabled
[12:48] <ioria> not if you use autologin
[12:48] <stevenm__> yup it's on in /etc/lightdm/lightdm.conf
[12:48] <ioria> or edited /etc/lightdm/lightdm.conf
[12:49] <stevenm__> :)
[12:49] <ioria> well, iirc, that file is not default (in that location) you need to create it
[12:49] <theMajor> stevenm__: glad it is solved for you
[12:50] <stevenm__> this "desktop" is actually just a shallow-depth 1U rackmount in a datacentre
[12:50] <ioria> i see
[12:50] <stevenm__> originally I wanted to remote into it using NoMachine... but then found it wasn't very quick
[12:50] <stevenm__> so then disabled the graphics.. swapped to X2Go
[12:50] <stevenm__> but now I'm sick of X2Go and I'm swapping back :)
[12:52] <theMajor> stevenm__: my experience with nomachine was actually pretty good never used X2GO so can't compare
[12:53] <stevenm__> theMajor, I think since NoMachine 4 it very much depends on how good the GPU is at helping with the encoding
[12:53] <stevenm__> which this machine isn't great at
[12:53] <theMajor> stevenm__: the only thing with nomachine is that it doesn't like wayland
[12:53] <stevenm__> but there are so many quirks and issues with X2Go (even the latest firefox has issues) and I've found with one false move it'll kill your sessions :(
[12:54] <stevenm__> theMajor, actually I've got a Fedora 34 box running gnome 40 & wayland... and I can NoMachine into it just great
[12:54] <stevenm__> and it works far better than the built in VNC & RDP of gnome-remote-desktop which is now in gnome 40
[12:56] <oerheks> gnome40 should be a lot better with remote access, but ubuntu is not that far yet.
[12:56] <stevenm__> yeah I'm hoping to use Ubuntu (rather than Ubuntu MATE) on my work lappy when it hits 22.04 - with hopefully both RDP and AD out of the box
[12:57] <theMajor> stevenm__: have to be honest that i have not tried it in gnome40 yet. i was running on xorg than for the time being. but hey maybe i switch and give it a try again as i have noticed that wayland performs better than xorg. I don't know if it is much better but they developed it for a reason
[12:57] <oerheks> Nope, it will not be introduced with an LTS.
[12:57] <stevenm__> 40 won't be?
[12:57] <goddard> any way i can rotate my screen from an ssh session?
[12:57] <goddard> i wanna see what is going wrong and why the screen just goes black
[12:58] <stevenm__> goddard, maybe an xrandr command?
[12:58] <stevenm__> with DISPLAY=:0.0 at the start
[12:58] <oerheks> rotating screen, does that help?
[12:59] <goddard> i tried xorg and wayland
[12:59] <goddard> its the gpd win3
[12:59] <oerheks> oh wait, impish 21.10 will have gnome41 https://discourse.ubuntu.com/t/impish-indri-release-schedule/18540
[12:59] <goddard> its screen is a sideways screen
[12:59] <goddard> makes no sense whatsoever but that is just the way it is
[12:59] <stevenm__> goddard, maybe?  DISPLAY=:0.0 xrandr -o right
[12:59] <goddard> thanks ill try
[13:00] <stevenm__> (or 'left', or 'normal' or 'inverted')
[13:00] <goddard> plasma was able to rotate no problem
[13:00] <goddard> only gnome has some kind of issue
[13:00] <stevenm__> oerheks, are you trying to say there will definitely be no GNOME 40 in 22.04?  and if you are, how can you know for sure?
[13:01] <stevenm__> i mean in 9 months GNOME 40 might be considered stable enough to support for 5 years
[13:01] <oerheks> i found 21.10 will have 41, see last post
[13:01] <stevenm__> yeah I saw that too
[13:01] <stevenm__> which is why it made no sense you saying 22.04 wouldn't have 40 :P
[13:02] <stevenm__> oh yeah - i missed your link
[13:02] <oerheks> stevenm__, if it was introduced, not with an LTS
[13:03] <goddard> yeah the command works
[13:03] <goddard> but still black screen
[13:03] <goddard> doing DISPLAY=:0.0 xrandr -o normal doesn't bring it out of the blackness of the void
[13:03] <stevenm__> hurrah - never done that before, yay for 2 seconds of googling (although it helps to know what to google :P)
[13:11] <goddard> i guess ill just use plasma
[13:12] <theMajor> i never tried rdp in gnome 40 is that possible out of the box?
[13:12] <goddard> vnc is
[13:12] <locsmif> Hi all. I install lsyncd and it works when I run it with options on the foreground but won't start as a service. I've tried to look at logs, but don't find any relevant errors (yet). It seems to be old-fashioned in that it start via a script in /etc/init.d/lsyncd ..anybody seen this before?
[13:13] <goddard> they call it screen sharing in ubuntu
[13:17] <_moep_> hi there! Is it possible to restrict the rights so that a user is not allowed to install new snaps (via snapd) without sudo?
[13:18] <theMajor> goddard: so basicly it is vnc and not actual rdp
[13:18] <goddard> it has a rdp client
[13:18] <goddard> but not rdp server
[13:19] <theMajor> goddard: and that last one would really be helpfull
[13:19] <goddard> everyone has their own use case i suppose
[13:19] <goddard> personally never needed it
[13:20] <theMajor> goddard: if that option was there we would not have to fiddle with things like nomachine. no i am a more terminal kind of guy so i can always ssh in and stuff but sometimes you just need your desktop for whatever reason
[13:20] <theMajor> goddard: even now i am ssh-ed into my machine from work
[13:21] <theMajor> goddard: meaning this is my irc from home :)
[13:21] <oerheks> moep any new user has no sudo rights
[13:21] <oerheks> only the 1st created during install *
[13:24] <goddard> theMajor so why not vnc?
[13:24] <goddard> personally i dont want to use any microsoft technologies
[13:24] <goddard> no samba, rdp, edge, vscode, etc..etc..
[13:25] <goddard> even if its just a translation technology created by some one else like samba
[13:25] <goddard> but i understand some sysadmins may need to work in an environment with other OS
[13:25] <goddard> i dont
[13:25] <goddard> usually you get a better experience doing it the purely Linux way
[13:26] <goddard> proton is the only thing i tolerate at the moment
[13:26] <rbasak> _moep_: isn't that how it is already?
[13:27] <bigpod> well i use edge all the time and vscode and samba and am primarily c# programmer honestly i dont care if microsoft created it to me at the end of the day that isnt a reason to use good piece of software
[13:27] <gordonjcp> goddard: interesting idea
[13:27] <gordonjcp> goddard: what's the linux equivalent of vscode?
[13:27] <bigpod> i dont am just a hobby programmer ;)
[13:27] <stevenm__> goddard, i'm not particularly happy that the wayland-effort (in it's many many ways) doesn't collectively seem to have considered an alternative for VNC or X11 forwarding
[13:28] <goddard> neovim is better than vscode
[13:28] <stevenm__> instead we're seeing GNOME building in freerdp's RDP server implementation into their desktop / wayland
[13:28] <goddard> and much lighter
[13:28] <gordonjcp> goddard: "better" in what way?
[13:28] <stevenm__> which is - err... nicer than VNC I guess?  but really? just imitating MS?
[13:28] <goddard> but still as a programmer myself a full IDE will always be better
[13:28] <goddard> and quicker to get going
[13:28] <goddard> if you dont need quick then learning NeoVim will be tremendously useful
[13:29] <bigpod> gordonjcp: u havent yet found an equivalent for vscode
[13:29] <goddard> stevenm__its a different protocol
[13:29] <stevenm__> what is?
[13:29] <gordonjcp> bigpod: not really, although I've only relatively recently started using vscode and it's quite a game-changer
[13:29] <goddard> vscode is a web browser that consumes massive amounts of ram to show a text file
[13:29] <goddard> it is pretty silly to use it
[13:29] <lotuspsychje> !discuss
[13:29] <goddard> sorry
[13:30] <lotuspsychje> please stay ontopic guys, we got a nice #ubuntu-discuss channel
[13:30] <goddard> ill be in offtopic
[13:30] <bigpod> stevenm__: honestly anything is better then VNC
[13:33] <_moep_> rbasak: Im not really sure, when I install the daemon for non-root users, its possible
[13:33] <_moep_> but tbo: Im new to snap
[13:42] <punkgeek> I have encrypted my total hard drive in ubuntu VM. clearly, the sysadmin can have access like physical access in the virtual environment. But there is a problem, every time my VM has been restarted, It requires entering the password to boot up. Is there any way to solve it? to boot up without entering the password?
[13:42] <gordonjcp> punkgeek: yes, don't use encryption
[13:44] <rbasak> punkgeek: how do you expect the system to get the decryption key?
[13:45] <rbasak> There are other options. You can provide a script that obtains the key however you want.
[13:45] <dob1> hi, where can I find information about motherboard compatibility ?
[13:45] <dob1> google the motherboard model + linux  gives me not much infos
[13:48] <ducasse> dob1: ubuntu has a compatibility list, but it's mostly for assembled systems
[13:49] <ducasse> there's no database for stuff like that i know of
[13:50] <bigpod> most if not all montherbards should work with ubuntu
[13:50] <bigpod> motherboard generaly doesnt have anything to do with linux compatibility
[13:55] <punkgeek> Actually, I want to protect my data in the webserver directory. But it should be mounted after boot up automatically. Do you have any suggestions?
[13:55] <lotuspsychje> !hardware | dob1
[13:55] <gordonjcp> punkgeek: yes, don't use disk encryption
[13:56] <punkgeek> For example, can I use a dynamic password and receive it from an API server to mount it?
[13:57] <gordonjcp> punkgeek: maybe, but why?
[13:57] <gordonjcp> disk encryption doesn't actually solve anything
[13:59] <punkgeek> So how can I protect the files from physical access?without encryption
[14:00] <gordonjcp> punkgeek: you can't
[14:00] <gordonjcp> punkgeek: unless you enter the key yourself on boot-up
[14:00] <gordonjcp> if you automatically decrypt the drive, how do you stop someone else doing that?
[14:00] <gordonjcp> punkgeek: think about your threat model
[14:00] <gordonjcp> punkgeek: what are you actually trying to protect against?
[14:02] <ogra`> punkgeek, you could get a USB key with physical number pad that handles the encryption on HW level internally ...
[14:04] <leftyfb> ogra`: that would still require manual intervention
[14:04] <ogra`> yes
[14:04] <leftyfb> punkgeek: if anything automatically decrypts the drive, then there's no point in encrypting the drive
[14:04] <ogra`> well ...
[14:05] <ogra`> if you want to protect yourself from drive theft this is actually a proper protection ...
[14:05] <punkgeek> I've thought about USB key and it is good but I can't give it to sysadmin
[14:05] <ogra`> but thats about the only use-case i guess
[14:06] <leftyfb> punkgeek: in your use case, there is no point in encrypting the drive if you're just going to automate decrypting it
[14:06] <gordonjcp> exactly
[14:06] <gordonjcp> punkgeek: where physically is the data?
[14:06] <leftyfb> punkgeek: either encrypt it and someone decrypts it at every boot or you don't encrypt it
[14:06] <gordonjcp> punkgeek: how likely is it that someone will physically attempt to steal the disk that the data is on?
[14:07] <bittin> hey
[14:07] <gordonjcp> hello
[14:08] <punkgeek> it can happen from our communist government. and I can't move the data outside the country.
[14:08] <leftyfb> punkgeek: then come up with a policy of manually decrypting the drive
[14:08] <gordonjcp> yup
[14:09] <gordonjcp> you will need to enter some sort of key that only you know
[14:09] <gordonjcp> automatically decrypting the drive won't help you here
[14:09] <leftyfb> punkgeek: either way, this is beyond the scope of this channel. Feel free to ask for more input in #ubuntu-security
[14:09] <gordonjcp> punkgeek: also, consider that this is the least likely way that anyone will exfiltrate your data
[14:10] <gordonjcp> punkgeek: you basically have two distinct infosec threats
[14:10] <gordonjcp> punkgeek: your two threats are Mossad, and Not-Mossad
[14:11] <gordonjcp> punkgeek: if Mossad are after your data, you can't do anything about it.  They'll abduct you, beat the password out of you, shoot you on national television, and then say they didn't have anything to do with your mysterious self-inflicted decapitation by rocket launcher
[14:11] <leftyfb> !ot | gordonjcp
[14:11] <gordonjcp> punkgeek: if Not-Mossad is after your data, then generally speaking ensuring your code has no obvious remote exploits and you don't reuse passwords on hornypornygirls.ru is going to be sufficient
[14:12] <leftyfb> gordonjcp: please stop
[14:12] <gordonjcp> punkgeek: but in general, disk encryption is going to be a tiny piece of your puzzle
[14:12] <leftyfb> gordonjcp: very innapropriate
[14:19] <punkgeek> What about this theory. I can write a script to find the hard disk hash. then I send it to an API server. If the has is be correct, the API server will send the password to the VM to decrypt the partition. So in this case, If someone goes through the rescue mode, the total disk hash wouldn't be correct.
[14:21] <gordonjcp> punkgeek: I don't know if that would work
[14:21] <bittin> Hello, Any SRU/MRU meeting today ?
[14:21] <gordonjcp> punkgeek: the hash would change all the time surely?  Like, as you write logs/edit pages/log in and out storing stuff in wtmp?
[14:23] <leftyfb> punkgeek: it's still automated. No point. Please bring further ideas to #ubuntu-security.
[14:23] <leftyfb> bittin: #ubuntu-devel maybe?
[14:23] <punkgeek> Okay, Thank you sir
[14:23] <bittin> leftyfb: #ubuntu-meeting but thx anyways
[14:33] <BluesKaj> Hi all
[14:35] <FreeBDSM> Hi. Even after reading the docs https://www.freedesktop.org/software/systemd/man/systemd.mount.html#x-initrd.mount - I don't really understand what the mount option 'x-initrd.mount' in /etc/fstab does. Could someone, please, explain?
[14:36] <rbasak> https://www.recompile.se/mandos is interesting
[14:37] <rbasak> But separately, punkgeek could you just have a separate partition and filesystem for the one directory you want encrypted?
[14:37] <rbasak> Then boot without that, and unlock and mount it manually afterwards
[14:44] <leftyfb> rbasak: encrypted home gives the same functionality and has already been suggested in #ubuntu-security
[15:05] <Maxii> Hi, I have limited space on an SSD and would like to have an encrypted boot and root partition on the ssd, and then have an encrypted /home partition on a separate hard drive. Is this possible?
[15:05] <Maxii> I'm assuming I could just follow this https://help.ubuntu.com/community/Full_Disk_Encryption_Howto_2019 and create another luks that has /home in, just like for / and /boot ?
[15:09]  * cbreak has an encrypted root pool via zfs native encryption
[15:10] <cbreak> and zfs encryption is by dataset, so my homedir could have its own key
[15:20] <asd_> hey, my j1800(asrock d1800m) can't play any media unless pulseaudio is killed in a loop; it doesn't detect the analog output from realtek sound card and the hdmi out audio is broken/no audio plays
[15:20] <asd_> can someone help me out; even if I don't kill pulseaudio in a loop
[15:20] <asd_> no audio plays
[15:21] <asd_> if I just reboot rn and launch it up it won't really work at all
[15:21] <asd_> other than letting me superficially set loudness
[15:23] <asd_> pulseaudio daemon doesn't write any errors to stderr
[15:24] <asd_> mpv/application doesn't respond with any failures
[15:24] <asd_> but it also doesn't play the audio it decodes
[15:24] <asd_> but at the same time can partially(at times with dropped frames) decode 1080p@60 vp9 video
[15:57] <shimbles> i upgraded focal->groovy->hirsute, and now vncserver with startplasma-x11 & in ~/.vnc/xstartup doesn't work, complaining about not being able to find the DISPLAY; DISPLAY=:1 and VNCDESKTOP is also set correctly
[15:58] <Mekaneck> shimbles: haven't you been through all that yesterday with the same question?
[16:00] <shimbles> i'm not sure i understand the question; is there additional information i can provide that will help you suggest a way to figure out how to fix it?
[16:01] <Mekaneck> i don't have that knowledge. Just saying that you been here yesterday too with the same issue and was helped by others?
[16:01] <shimbles> i think you need to stop talking to me, now.
[16:01] <leftyfb> shimbles: echo $XDG_SESSION_TYPE
[16:02] <Mekaneck> if nothing works either reinstall 21.04 and see if it works then if not, stick with LTS
[16:02] <leftyfb> shimbles: please be respectful to those that are offering assistance
[16:02] <Mekaneck> thanks leftyfb
[16:02] <shimbles> he's not offering assistance, he's telling me to stop talking
[16:02] <peetaur> Howdy...I have/had a headless Ubuntu 18.04 machine where I did "apt install ubuntu-desktop" and have 2 problems... (1) is when I `systemctl start gdm` the thing that pops up has "Welcome!" instead of login, and says my user already exists so can't create one. How do I get the login? (2) `systemctl enable gdm` complains "The unit files have no installation config..." so can't enable.
[16:02] <Mekaneck> i do not shimbles
[16:02] <shimbles> if he continues to stalk me around different channels, I will file harassment charges and subpoena his identity
[16:02] <Mekaneck> sigh
[16:03] <Mekaneck> just do as you please, IF you actually know what harrasment is
[16:03] <shimbles>  XDG_SESSION_TYPE was set to tty
[16:03] <leftyfb> shimbles: ok, No sense in helping someone who is hostile and threatening legal proceedings. good luck.
[16:04] <shimbles> [10:01:15 AM] <Mekaneck> i don't have that knowledge. Just saying that you been here yesterday too with the same issue and was helped by others?
[16:04] <shimbles> [10:01:25 AM] <shimbles> i think you need to stop talking to me, now.
[16:04] <shimbles> he literally showed up here and in #kde to chastise me for asking the same question yesterday, and not managing to fix it
[16:04] <jailbreak> shimbles, just ignore him. he's the "grumpy" dude of the channel. don't worry about it! :-)
[16:05] <Mekaneck> troll elswhere shimbles
[16:05] <Mekaneck> ignore it is
[16:05] <shimbles> you are the one trolling. you are not providing assistance. you are pointing out to others that i asked the same question yesterday, and that they should not help me.
[16:05] <Aison> hello, pulseaudio is not started here after reboot
[16:05] <Aison> I always have to start it manually (and then it works...)
[16:06] <Aison> I tried: systemctl --user reenable pulseaudio
[16:06] <shimbles> leftyfb hopped on to add to your toxicity without reading what you actually said; and what you actually said "look, he was in here yesterday!!! dont help him!!!" made no sense.
[16:06] <Aison> but then I get the error: Failed to connect to bus: No such file or directory
[16:06] <Aison> any idea what's wrong? I guess it's related to dbus
[16:06] <Mekaneck> shimbles: it tried to help but if you see it that way, have a nice day
[16:06] <shimbles> you did not try to help. where did you provide technical assistance?
[16:07] <Mekaneck> sigh
[16:07] <oerheks> Mekaneck, now stop please, thanks.
[16:07] <leftyfb> shimbles: in reading your posts from yesterday, it looks like the only purpose you have for a GUI and VNC is to access a web URL listening on localhost. You know you can still access this by way of your remote machine and ssh right?
[16:07] <Mekaneck> oerheks: i'm doing nothing wrong... but ok, if you guys see it that way...
[16:08] <shimbles> leftyfb i am specifically trying to get Apache Guacamole to continue to work on Ubuntu Hirsute. Guacamole itself is working, but something broke from groovy to hirsute such that I can't get the vnc server to serve up startplasma-x11
[16:08] <oerheks> shimbles, are you in Wayland or Xorg session?
[16:09] <oerheks> wayland is standard now ..
[16:09] <leftyfb> oerheks: I asked the same, but was told "tty" which means they are running the command over ssh and not locally
[16:09] <shimbles> oerheks I am on AWS, the instance was upgraded from focal->groovy->hirsute; i have not figured out how i might disable any wayland config changes as of yet; i read that for KDE Plasma on Hirsute, wayland is not the default yet.
[16:11] <peetaur> I guess I solved #1 ...   just let it create a useless user, then `killall -u uselessuser ; userdel -r uselessuser`
[16:11] <shimbles> on focal and groovy, i had not yet managed to get vncserver to run at startup and was running it manually, and it was working; so my ~/.vnc/xstartup file, which contains xrdb $HOME/.Xresources; startplasma-x11 &     at least used to work
[16:11] <oerheks> shimbles, get a fresh image from aws? .. i would not do such upgrade path.
[16:11] <oerheks> and make sure you run xorg ..
[16:12] <shimbles> oerheks since I am going to patch QtWebEngine, i want to be fully upgraded; and i need QtWebEngine running in Guacamole to test my patch
[16:13] <shimbles> i have focal / groovy images which work, but the software is somewhat out of date; i could compile the full stack (chromium, qt, pyside) although that's rather large;
[16:13] <shimbles> on that topic, "make sure you run xorg" how would you recommend I do that? via startx? I was not doing that previously
[16:14] <shimbles> furthermore, since DISPLAY and VNCDESKTOP are set correctly when ~/.vnc/xstartup runs, isn't X being started correctly?
[16:16] <shimbles> since i can't fix it, i really have no choice but to compile everything on groovy; still, it would be sweet to get Plasma working on Guacamole on Hirsute on AWS; i feel like there must be some minor thing i can fix :/
[16:17] <shimbles> i wonder about disabling wayland? what do you think oerheks, perhaps something with systemd?
[16:17] <oerheks> echo $XDG_SESSION_TYPE , then use https://ubuntu.com/tutorials/ubuntu-desktop-aws#1-overview
[16:31] <BinarySavior> hello, how can i create a "dummy connection" for debugging purposes to any ip (ex 127.0.0.1) on a specific port (ex, port 4000)
[16:31] <BinarySavior> tcp connection
[16:33] <rbasak> BinarySavior: to listen, or to connect?
[16:33] <rbasak> netcat (nc) can do both
[16:33] <rbasak> Or you might be interested in socat
[16:34] <BinarySavior> i have a python script that monitors for any connection on a specific port, i just want to create a connection to activate parts of my python app
[16:35] <BinarySavior> it doesn't listen, just verifies that a connection exists
[16:46] <BinarySavior> rbasak, nc worked perfectly, i just set up a listener on localhost then connected to it
[16:46] <BinarySavior> ty
[17:15] <punkgeek> Before I've decrypt a directory in the ubuntu that people can have physical access, what cases should I checked before decrypting the data? because it needs to be in mounted position until restart. For example, I should check  the root password not to be changed, there is not another account has been added, there is no running bad script or a script won't be run after a while to change the password.
[17:17] <shimbles> oerheks i found switching to gnome on hirsute (which allowed me to disable wayland) didn't work either :/
[17:17] <shimbles> looks like i've got some make -j in my future
[17:27] <punkgeek> Let me explain in another way. I have a debian VM, I'am suspect the sysadmin to stoling my encrypted data. So when ever I login to my system and want to decryp my encrypted partition, What cases should I check before mounting? For example, the sysadmin may go to in the rescue mode and add another user with sudo access, So when I mount my encrypted partition, he can access to my data.
[17:27] <punkgeek> or ubuntu VM, doesn't matter*
[17:29] <leftyfb> punkgeek: your situation is no to be trusted, period. You have no way of knowing what is done to your VM by any means. It cannot be trusted no matter what you do. That said, your question is beyond the scope of this channel. Please stop cross-posting. Wait for an answer in #ubuntu-security or try #ubuntu-offtopic. You are going to get the same answer no matter where you ask though.
[17:37] <dreamon> good evening. having issue, to suspend my laptop. "sudo pm-suspend" works. but it dont locks screen, after awaking. systemctl suspent dont work .. it only locks screen
[17:38] <dreamon> https://pastebin.ubuntu.com/p/sWjwdCDgwy/
[18:16] <goddard> is it possible to black a screen but not shut it off?
[18:17] <sarnold> goddard: xset dpms has several different tools for that kind of thing, though what they do to your device is dependant upon your device
[18:17] <sarnold> goddard: you can also just put up a big black window, some screen lockers do that kind of thing
[18:18] <goddard> yeah that is what i was thinking
[18:18] <goddard> I have an OLED so i want to save the screen from any chance of burn in, but then this OLED has some bug with Nvidia which prevents it from the signal going back to the screen once its shut off
[18:18] <goddard> really strange
[18:19] <goddard> if i wait long enough i can get it to come back
[18:22] <sarnold> ewww
[18:22] <goddard> nvidia is such a pain sometimes
[18:22] <cart_man> Hey everyone. If I added this Repo like -> sudo add-apt-repository ppa:beineri/opt-qt-5.12.2-bionic    ...and did the update. How can I remove it just for safety
[18:23] <sarnold> cart_man: depending upon what you want to do, you can either comment the lines out of your sources, in case you want to put it back again at some point.. or you can use ppa-purge from the ppa-purge purge package to remove the ppa and everything that was installed from it
[18:24] <goddard> In gnome settings "Performance" option is missing.  See - for example of what it should look like https://askubuntu.com/a/1343879/37577
[18:25] <goddard> What i see https://ibb.co/fn5vKj9
[18:30] <goddard> nvidia wont support HDMI-CEC
[18:30] <goddard> https://forums.developer.nvidia.com/t/hdmi-cec-support/31445
[18:30] <goddard> if they did it would probably make this bug go away
[18:33] <cart_man> sarnold: I always get shit with adding these repos... "Held broken packages" is basically a guaranteed for me so I just want to clean them afters
[18:33] <cart_man> afterwards
[18:34] <cart_man> sarnold: Where is that file located that I can comment it out in??
[18:35] <goddard> cart_man look at ppa purge
[18:35] <goddard> its a gui to manage your ppas
[18:36] <goddard> super simple and usually works without incident
[18:36] <cart_man> do I install it first?
[18:36] <cart_man> How do I run it?
[18:36] <cart_man> ppa purge seems ...well
[18:36] <cart_man> I can not run it like that
[18:37] <goddard> y-ppa-manager
[18:37] <goddard> if you want a gui
[18:38] <sarnold> cart_man: it'll be in /etc/apt/sources.list or a file in /etc/apt/sources.list.d/
[18:48] <Slartiba1t> Guys? I used to have 2 external monitors connected to my laptop, one displayport and one HDMI. Now I've added another displayport monitor, so I have one HDMI cable and a displayport hub connected to the laptop. It works in the dualboot Windows 10, but when I try to boot Ubuntu 20.10 it freezes during boot. If I restore the old setup with 2 monitors Ubuntu boots as well. Anyone had similar problems?
[18:48] <Slartiba1t> Should I file a bug report somewhere? (Where? Never had this kind of problem before..)
[18:52] <cart_man> goddard: Thanks thats a nice too
[18:52] <cart_man> tool
[18:53] <cart_man> ALSO.... I am trying to build a program but my QT libs inside /usr/local/lib/  overshadow the libs that I am trying to use for the build and HENCH my build fails. How can I temporarily remove the libs from being used by my build? I still need /usr/local/lib in the path since other things need the libs in that path
[18:54] <cart_man> runtime library [libQt5Network.so.5] in /usr/lib/x86_64-linux-gnu may be hidden by files in: /home/rt/Qt/5.12.11/gcc_64/lib
[18:55] <punkgeek>  I'm going to use remote ssh for enter LUKS password. Is the interface name same as the network configuration file? I've tried this but it doesn't worked: I use this configuration but it doesn't works: GRUB_CMDLINE_LINUX="IP=51.68.177.130::51.68.177.254:255.255.255.0:linux:ens160:none"
[18:55] <TJ-> Slartiba1t: I'd do some isolation/combination tests first. Will it work with only the displayport hub and 1 DP monitor? the hub and 2 DP monitors? the hub, 1 DP monitor, 1 HDMI?
[18:56] <TJ-> punkgeek: no - at boot time it's the kernel name - you'd be better using the interface MAC address
[18:57] <Slartiba1t> TJ-: Fair enough. I'll try to cover all the cases.
[18:58] <punkgeek> TJ-: How can I find the interface name from the kernel?
[18:59] <TJ-> punkgeek: Ethernet interfaces start at eth0 and the number increments for each discovered interface
[19:01] <TJ-> punkgeek: I'm not sure if NFS's ip= can do MAC addresses now I'm reading up on it. I'm pretty sure I've used it that way many years ago but kernel admin guide doesn't mention it. However, the ip= command-line option is intercepted and interpreted by the initramfs-tools /init shell script and there are hooks to handle ip= in those
[19:06] <TJ-> punkgeek: aha, yes! see "/usr/share/initramfs-tools/scripts/functions" and the function configure_networking() where it translates a BOOTIF=01-aa-bb-cc-dd-ee-ff that is set by PXE boots to a corrected MAC address e.g. aa:bb:cc:dd:ee:ff then looks up which device name has that MAC address and sets the IP= DEVICE accordingly
[19:07] <TJ-> punkgeek: so you can 'pretend' it booted using PXE and add a BOOTIF=01-yo-ur-ma-ca-dd-re ss on the kernel comand line and moit a DEVICE from the IP= line and it should fill it correctly for you
[19:07] <TJ-> s/moit/ommit/
[19:13] <diskin> Hi all, I was in a meet video call, when the sound started repeating and the system become unresponsive to mouse or keyboard. REISUB did not do anything. Powercycled and found nothing in logs for the moment of the crash. How to debug such problems?
[19:15] <sarnold> diskin: maybe you could make progress using a crash kernel to collect information when it hangs, but that's not the easiest thing to set up, and then you'd have to figure out how to do the kernel debugging. it wouldn't be fun, unless that's your idea of fun..
[19:16] <oerheks> this 'meet'? https://meet.google.com/
[19:16] <sarnold> diskin: it's perhaps more approachable to remove devices from your system one at a time and try to recreate it
[19:16] <sarnold> diskin: or perhaps if you've got 'funny' kernel modules loaded like vbox, remove those, and try to recreate it
[19:16] <sarnold> it's also not fun, but far more approachable if you're not good at debugging kernels
[19:18] <sarnold> another possibility is that your power supply just isn't up to the task; I've seen power supplies that were pushed too far, or failing, do things like that
[19:28] <diskin> oerheks, yes, this meet.
[19:29] <diskin> sarnold, re: power supply, it's a laptop with a working battery, plugged in to AC power source, does it matter?
[19:29] <sarnold> diskin: ah that's probably enough unless you've got crazy usb accessories attached
[19:29] <oerheks> if that runs in your chrome browser, look there for logs? .. or wait,  logs aren't generated automatically. You need to turn on logging first. https://support.google.com/chrome/a/answer/6271282?hl=en#zippy=%2Clinux
[19:30] <diskin> and I do have virtualbox installed, some modules loaded... but it happened 1st time in a year or so. normally the system is very stable, upwtime several weeks
[19:31] <diskin> oerheks, are you saying that Chrome could have logged the system crash in it's own logs?
[19:31] <oerheks> only when enabled, yes.
[19:31] <sarnold> when the sound card starts playing the same clip over and over again it means the kernel wasn't able to put new audio data in its buffer in time, so it's just playing what it's got.. probably the machine is hosed enough at that point that google meet logging won't be able to help
[19:33] <diskin> yes, the sound card was repeating some part of a word which was just said by a call participant, initially I thought it was something with his machine, but very quickly understood it was mine :)
[19:35] <oerheks> when using meet with 5+ users, you will need 4 gb ram...
[19:36] <diskin> oerheks, I have 16Gb.
[19:36] <oerheks> ohhh, you could do HD, that needs 8 gb
[19:36] <oerheks> no, without those logs no way telling what went wrong.
[19:36] <diskin> But I also have a lot of browser tabs open (bad habit!) and memory consumption is high. But I try to avoid swapping (also have 2Gb swap partition)
[19:39] <Guest27> Hi y'all. I'm freaking out. Today I turn on my laptop and it wouldn't boot so I end up turning off and back on again and I started getting something along the lines: Error dev sda blk_update_request, RUN fsck MANUALLY etc
[19:40] <Guest27> So I googled a bit about it and I landed on a website that suggested running an Ubuntu live cd and running 'fdisk -l', getting what the device was and running fsck etc
[19:41] <Guest27> I did that and got something along the lines 'magic number yada yada superblock'
[19:41] <Guest27> so more googling took me to 'you need to run e2fsck etc' which I did but when trying to restart from one of the superblocks, I'd either get an dos partition table error or I/O error
[19:42] <Guest27> I turned it off and on again since I needed to go afk and decided to come back later (now) to get help from IRC
[19:43] <Guest27> I just turned it on and it says "Error reading block XXXX (Input/output error) /dev/sad1: UNEXPECTED INCONSISTENCY; RUN fsck MANUALLY
[19:43] <Guest27> fsck exited with status code 4
[19:43] <Guest27> The root filesystem on /dev/sda1 requires a manual fsck
[19:43] <leftyfb> Guest27: it's more than likely a bad drive. I would get a new one ASAP, leave your machine off till you can restore it and then once your new one is installed and Ubuntu re-installed, reinstall your applications and restore from backup
[19:44] <Guest27> leftyfb is there a way to remount the current drive once I have a clean install on a new hdd?
[19:44] <Guest27> there were some things i hadn't backed up yet
[19:45] <Guest27> i know this is on me. i just haden't gotten around to backing up some files in the last week
[19:45] <lounge-user> Help! My DisplayPort is not detected ever since updating my Lenovo ThinkPad T510 to 20.04 from 18.04
[19:45] <leftyfb> Guest27: worry about that once you're back online with a new drive. The more you power that on and mess with it, the less likely you are to retrieve anything from it
[19:45] <jdaviescoates> Help! My DisplayPort is not detected ever since updating my Lenovo ThinkPad T510 to 20.04 from 18.04
[19:45] <Guest27> ugh... i did mess with it
[19:45] <leftyfb> !repeat | jdaviescoates
[19:45] <Guest27> hopefully i'll be able to restore something
[19:46] <jdaviescoates> leftyfb thanks, I just figured if someone replied to lounge-user I'd miss it after changing nick
[19:46] <Guest27> how should i go about it btw? leftyfb
[19:46] <Guest27> i have this laptop with ubuntu as well
[19:47] <Guest27> how could i restore the old drive?
[19:48] <leftyfb> Guest27: Using an external, usb drive reader via USB, mounted as read-only. Or ideally, using something like ddrescue to dd an entire image to a new drive and location for forensics on the created image
[19:49] <jdaviescoates>  /msg NickServ VERIFY REGISTER jdaviescoates 85lbGDcTwy9cvwFy
[19:49] <leftyfb> jdaviescoates: you might want to pick a different password now
[19:50] <jdaviescoates> haha
[19:50] <jdaviescoates>  /NickServ VERIFY REGISTER jdaviescoates 85lbGDcTwy9cvwFy
[19:50] <leftyfb> jdaviescoates: please go to #libera for help on how to use IRC properly
[19:51] <JonathanD> It's just a verify key, so it's not leaking the pw :)
[19:51] <jdaviescoates> it's been too long, totally forgotten how to use this! (but not the first time I've done that - although that's just... yeah what JonathanD said
[19:51] <jdaviescoates> I have previously leaked pw in a similar fashion before too though! :P
[19:51] <Guest27> leftyfb thank you. i actually have a usb drive reader, so i guess i should plug that and then mount it as read only and then i should be able to access whaterver it's in it? since i'm not booting from it
[19:52] <leftyfb> Guest27: I would go with the ddrescue option first, if you've got the space
[19:52] <Guest27> leftyfb what do you mean?
[19:53] <leftyfb> Guest27: https://help.ubuntu.com/community/DataRecovery#Imaging_a_damaged_device.2C_filesystem_or_drive
[20:07] <Rotofeiul> hi! Will work my nvidia GeForce 1660 GTX on Ubuntu 21.04?
[20:09] <oerheks> Rotofeiul, no reason why not; sudo ubuntu-drivers autoinstall # it will install recommended ones
[20:09] <sarnold> nvidia has a big table somewhere saying which cards they still support and with which drivers
[20:10] <sarnold> it might be faster to just try it out as oerheks says ..
[20:13] <Rotofeiul> thanks!
[20:23] <bw2> hello
[20:33] <nomike_> Hi
[20:35] <nomike_> Is it possible to install the ubuntu installer in a running/installed ubuntu? I want to install ubuntu on a portable USB thumb drive, so I have my personal full installation which I can boot on my work laptop which is running Windows 10 without influencing any of the data on internal hard drive.
[20:36] <TJ-> nomike_: there are a couple of ways you can do that. Probably the easiest is to create a virtual machine and set the USB thunb drive, attach the Ubuntu installation ISO file as a virtual CDROM and boot it and go through the install as normal.
[20:36] <nomike_> I've booted from live USB drive A and am installing to USB drive B and it works well, though it is painfully slow, the setup process is running for a couple of hours already) and I was jyst curious if I would be able to just install the installer on my other laptop which has a running installtio of ubuntu.
[20:44] <TJ-> nomike: it's a 30 second task to create a virtual machine and attach a USB device directly to it as its storage device, and attach the installer ISO file as CDROM and boot the VM from that. Your slowness is likely due to reading/writing USB2 devices
[20:44] <oerheks> nomike, yes, it will be horribly slow, installing, and running.
[20:45] <oerheks> TJ-, +1
[20:45] <nomike> W
[20:45] <nomike> oops
[20:46] <nomike> Well, both USB drives are USB 3.0 and are connected to USB 3.0 ports. But whatever, it works and it's almost done. I was just curious. After all it's an odd thing to want to start an installer in an already installed system.
[20:47] <nomike> And it's running somewhere in the background. I don't sit next to it and twist my thumbs waiting for it to finally finish.
[20:47] <TJ-> nomike: I do it using debootstrap :)
[20:49] <ubbuntiti> What code does ubuntu is developed?
[20:49] <ubbuntiti> I meant language
[20:49] <ubbuntiti> C?
[20:50] <nomike> ubbuntiti, which part of ubuntu? Ubuntu is not one software.
[20:51] <oerheks> C++, python3, and more
[20:51] <TJ-> ubbuntiti: Ubuntu is approximately 50,000 different packages... pick one!
[20:51] <ubbuntiti> all right
[20:51] <ubbuntiti> that answer
[20:51] <nomike> The linux kernel is written in C. A lot of tools are written in C (e.g. bash, grub, etc.). Some tools might be written in python or perl. There are shell scripts. Maybe some go.
[20:52] <ubbuntiti> Does perl make front end programms?
[20:52] <nomike> SOme tools are even written in mono (a variant of the .net family)
[20:53] <nomike> (some people are going to slap me for that variant thingy, but I think that's a quite accurate description).
[20:54] <gordonjcp> TJ-: I wonder how much straight assembler there is in a modern Linux distro
[20:55] <TJ-> gordonjcp: that'd be an interesting challenge to figure out!
[20:55] <gordonjcp> TJ-: probably some in the kernel, deep down in the driver layers
[20:56] <oerheks> now there is rust support .. https://github.com/Rust-for-Linux
[20:56] <TJ->  " apt-file search -x '(\.S|\.asm)$' | wc -l " => 230
[20:56] <TJ-> oops, 340
[20:57] <julio> HELLO WORLD!
[21:00] <nomike> According to github on the kernel source code: C 98.4%, Assembly 0.9%, Shell 0.3%, Makefile 0.2%, Perl 0.1%, Python 0.1%
[21:00] <nomike> .2 % Makefiles? That's a lot.
[21:05] <TJ-> nomike: every directory in kernel has a Makefile, or more than 1
[21:05] <linsux> how do i remove application icon on desktop
[21:05] <nomike> TJ-, ah yes. I already thought so.
[21:06] <punkgeek> How can I install md5sum on busybox from ubuntu?
[21:10] <TJ-> punkgeek: busybox has the md5sum applet built in
[21:14] <punkgeek> Thank you
[21:15] <punkgeek> Is it possible to remoev the grub password on the full encrypted hard drive?
[21:16] <TJ-> punkgeek: grub password?
[21:17] <punkgeek> yes, Grub Boot Loader Password
[21:17] <punkgeek> the actual grub password, the one protecting the grub menu itself
[21:18] <TJ-> punkgeek: the usual way I presume, by resetting it
[21:19] <leftyfb> punkgeek: please stop cross-posting
[21:22] <noarb-> in the default Settings > Date & Time > Automatic Date & Time, is there a way to set to a local ntp server?
[21:39] <jhutchins> Does Ubuntu track which repository a given package came from?
[21:42] <ice9> how can i run xorg app in wayland session?
[21:46] <sarnold> jhutchins: sortof, I think; try apt-cache policy on a package from a ppa and on a package from the archive, and compare the output
[21:46] <sarnold> jhutchins: this is the mechanism used by ppa-purge
[21:55] <Guest63> i'm trying to follow this guide (https://help.ubuntu.com/community/DataRecovery#Imaging_a_damaged_device.2C_filesystem_or_drive) suggested by leftyfb
[21:56] <Guest63> however, I'm utterly confused with how to read fdisk -l
[21:57] <Guest63> i think it is: Disk /dev/sdb: 465.8 GiB, XXX bytes, XXX sectors. What is a Device? I see /dev/sdb1 * Start End Sectors, yada yada yada
[21:58] <Guest63> How do I know if my disk / device is mounted?
[21:59] <Guest63> got it. the device part is the partition tables. i reckon that means i have only one partition on this disk
[21:59] <sarnold> the /proc/mounts file will probably show you if the device is mounted, and where
[22:01] <Guest63> sarnold /proc/mounts | grep sdb shows nothing
[22:01] <Guest63> i reckon it ain't mounted, right?
[22:01] <sarnold> yeah :)
[22:03] <goddard> any way to make ubuntu dock show the windows on hover?
[22:04] <goddard> if you have two windows of the same app you gotta right click and then click all windows and then click the app you want
[22:07] <Guest63> i guess i have two questions: when running "sudo mount /dev/xxx /mnt/media" should xxx be sdb or sdb1? should /mnt/media be an existent directory?
[22:09] <sarnold> Guest63: probably /dev/sdb1, unless you created the filesystem directly on the drive without making partitions first .. that's insanely rare, but possible
[22:09] <sarnold> Guest63: yes, the directory must exist first
[22:09] <Guest63> i guess i do have a partition since fdisk shows sdb1
[22:11] <lordley> do command df and you see the name of the drive
[22:11] <Guest63> ugh. this is bad news i think
[22:11] <Guest63> sudo mount -r /dev/sdb1 /mnt/media returns
[22:11] <Guest63> mount: /mnt/media: cannot mount; probably corrupted filesystem on /dev/sdb1
[22:12] <Guest63> i thought i needed to mount to follow this https://help.ubuntu.com/community/DataRecovery#Imaging_a_damaged_device.2C_filesystem_or_drive
[22:13] <Guest63> i see "ddrescue --no-split /dev/hda1 imagefile logfile" can imagefile be any directory?
[22:14] <sarnold> Guest63: ddrescue doesn't mount to a directory, it saves to a file, and probably the file shouldn't already exist
[22:18] <Guest63> thanks. i went with sudo ddrescue -n /dev/sdb1 /mnt/media/image /mnt/media/logfile
[22:18] <Guest63> i re read the guide and it didn't mention mounting the partition at any point
[22:18] <Guest63> i guess i just got confused
[22:18] <Guest63> so i'm just running ddrescue now to save as much as i can... hopefully i can get my files back :'(
[22:20] <Guest63> wtf. remaining time 4h lol
[22:20] <Guest63> didnt know it was going to take this long
[22:22] <sarnold> reading every sector of the hard drive can take a while
[22:22] <TJ-> Guest63: if the source is in any way damaged, it can take days. ddrescue marks blocks it cannot read initially and goes back and forth around them trying to extract the sectors either side. Every read error fault can takes a few seconds to recover from
[22:23] <sarnold> a healthy hard drive can do roughly 100MB/s when doing 'large' reads, and ddrescue may not be doing large reads if it expects some / many of the blocks to be broken
[22:23] <Guest63> well from the ubuntu wiki
[22:23] <Guest63> i'm following the forensics recommendation
[22:23] <Guest63> running --no-split first and then with retry=3
[22:25] <TJ-> Guest63: ddrescue can use a larger --cluster-size= (default is 128) to speed up reads... the more sectors read in each operationg the faster the transfer will be
[22:25] <Guest63> oh, i guess i should've done that
[22:25] <Guest63> wait. i'm just freaking out
[22:26] <TJ-> it's like using 'dd bs=150M' instead of its default of 512 bytes
[22:26] <Guest63> my corrupted disk is 500gib but i had only used about 30-40 gib
[22:26] <Guest63> will ddrescue go only over those 30-40gib or the entire thing?
[22:26] <Guest63> if so, i don't even have that much room for 500 gib backup or whatever
[22:26] <Guest63> did i mess up?
[22:27] <TJ-> it's doing the entire block device
[22:27] <Guest63> oh, shoot, that's not good
[22:27] <Guest63> i don't have that much free space on this computer
[22:28] <TJ-> if you need to limit the range of sectors it should read, you'll first need to use file-system tools to identify what blocks the file-system is using
[22:28] <Guest63> can i do that when the partition is corrupted though?
[22:29] <TJ-> 1. identify file-system type 2. identify tools that can tell you about used/free blocks in that file-system 3. figure out the range, or multiple ranges, of sectors needing recovery 4. use that info to craft a ddrescue command-line
[22:29] <TJ-> Guest63: partitions don't get corrupted, file systems do
[22:30] <sarnold> well, a partition table can also get corrupted, but that's also pretty rare..
[22:30] <Guest63> thanks for the correction
[22:30] <sarnold> (it's almost always a user-error :)
[22:30] <TJ-> A partition is nothing more than a starting sector number, a quantity, and some optional type info
[22:30] <Guest63> any recommended tools to tackle 1 and 2? i'm pretty sure it was ext4 but i'm not entirely sure
[22:30] <TJ-> if it's ext4 then dumpe2fs can give you info
[22:31] <Guest63> i'm still a linux newbie and i was even more so when i install this one
[22:31] <TJ-> start of with e.g. "dumpe2fs -h /dev/sdb1"
[22:31] <TJ-> that'll display a concise summary from any superblock it can find
[22:32] <Guest63> lol i got ddrescue: Error writing mapfil 'mnt/media/logfile': No space left on device
[22:32] <Guest63> how did i even run out of space this quick?
[22:32] <Guest63> i'm supposed to have ~445gib on this laptop
[22:32] <TJ-> Guest63: are you sure you are trying to recover from a physical disk hardware corruption, or is this /just/ file-system corruption
[22:33] <Guest63> TJ- yes, so here's what happened after my computer failed to boot and people here, namely, leftyfb pointed me to that ubuntu guide
[22:33] <Guest63> i took apart my laptop and got the hdd out
[22:33] <Guest63> so now i'm trying to recover it from another laptop
[22:33] <Guest63> does that answer your question? i can expand more if not
[22:34] <TJ-> Guest63: so there are errors reported in the SMART data, using smartctl, or I/O errors are/were seen in the kernel logs?
[22:35] <TJ-> Guest63: because if this is simply some structures in the file-system being over-written or corrupted, but the underlying physical storage is OK, ddrescue isn't the solution
[22:36] <Guest63> TJ- you're speaking greek to me. my computer wouldn't boot and i came here earlier with this: https://pastebin.com/raw/zqELrNk4
[22:36] <Guest63> TJ- that's a good insight. however, i'm not sure whether it is either or
[22:36] <Guest63> what i shared above (pastebin) is what i was getting
[22:37] <Guest63> is there a way to test that now that the disk isn't on the laptop anymore?
[22:37] <TJ-> Guest63: OK, physical fault ""Error reading block XXXX (Input/output error)"
[22:37] <Guest63> i decided to go this route as to not keep trying the fsck thing and that would override everything making my things def not recoverable
[22:38] <TJ-> Guest63: what you could do is re-test by doing a fast sequential read of the entire device and monitoring the kernel log for any more I/O errors.
[22:38] <Guest63> wouldn't that lower my chances of getting my data back?
[22:38] <Guest63> sorry, meant to say would*
[22:39] <Guest63> i'm def in no position of saying wouldnt. my knowledge isn't too advance yet lol
[22:39] <Guest63> TJ- is there a guide on how to go about that?
[22:40] <TJ-> Guest63: I'd do it with ddrescue but with different options :)
[22:41] <TJ-> Guest63: what was the ddrescue command-line you used before it ran out of space?
[22:41] <Guest63> sudo ddrescue -n /dev/sdb1 /mnt/media/image /mnt/media/logfile
[22:42] <TJ-> Guest63: how large is /mnt/media/logfile ? use "df -h" to check
[22:43] <Guest63> df -h /mnt/media/logfile returns /dev/sda1 28GB 28GB 0 100$ /
[22:43] <Guest63> Filesystem Size Used Avail Use% Mounted on
[22:43] <Guest63> however, ls -lah /mnt/media shows image is 24G and logfile 0
[22:44] <Guest63> it might've been because when prompted with the ddrescue error i just canceled the thing
[22:44] <TJ-> so it's out of space due to the previous run
[22:44] <Guest63> i truly don't get the answer given by df -h
[22:45] <Guest63> seems like it is talking about /dev/sda1 rather than logfile itself
[22:45] <TJ-> the block device is 28GB, and 28GB is used
[22:46] <Guest63> oh okay
[22:46] <TJ-> and I think you said that you think the amount of space used on the damaged device was 30-40GB? if so, there's not enough space to recover even the used sectors
[22:47] <TJ-> 'df' reports block device space, not files. it reports /dev/sda1 since that is the block device containing the file-system that has /mnt/media/logfile in it
[22:47] <Guest63> correct. though, could i use /home/Guest63?
[22:48] <TJ-> Guest63: use simply "df -h" and see which of your mounted file-systems will have sufficient space
[22:48] <Guest63> df -h /home/Guest63 shows /dev/sda6 441G 127M 418G 1% /home
[22:48] <TJ-> OK, so that's where the saved data needs to go
[22:49] <TJ-> and you should use some additional command line options to make this much more efficient, in terms of speed and of actual space used
[22:49] <Guest63> okay. that sounds smart. i'm all ears
[22:50] <TJ-> So, we'll put the data under /home/$USER, OK
[22:51] <Guest63> okay
[22:52] <TJ-> We'll write a sparse file (any empty sectors aren't actually copied but the image is marked as having an empty sector. Means you could copy a 500GB block device and it'll only need space for the non-empty sectors )
[22:52] <TJ-> and we'll read many more sectors in each read operation
[22:52] <Guest63> understood
[22:53] <TJ-> I'd suggest: "sudo ddrescue --sparse --cluster-size=4096 --no-scrape /dev/sdb1 /home/$USER/sdb1.img /home/$USER/sdb1.mapfile"
[22:57] <Guest63> TJ- running
[22:57] <Guest63> it went for 11s and then paused for about 1m and now seems like it's reading again
[22:57] <Guest63> remaining time: 1h 2m
[22:58] <Guest63> yeah, for some reason it "hangs / freezes" from time to time
[23:00] <TJ-> That's it dealing with I/O errors
[23:01] <TJ-> If you open another terminal and do "journalctl -k -f" that'll '(f)ollow the kernel log as it is written and you'll likely see more of those I/O errors reading ....
[23:03] <TJ-> after a few tries, if a sector cannot be read, it'll makr it as bad in the mapfile and move on some way and try again. Later it'll return to the bad areas and try to read smaller numbers of sectors around the damaged area(s)
[23:03] <Guest63> TJ- oh yeah. i see them now. 'print_req_error: critical target error, dev sdb, sector xxx' and 'buffer i/o error on dev sdb1, logical block xxx, async page read'
[23:03] <TJ-> Guest63: if there are lots you can expect the recovery run to take days!
[23:04] <TJ-> each error can cause a pause whilst the device recovers of several seconds
[23:04] <Guest63> there are two so far
[23:04] <TJ-> that is promising
[23:04] <Guest63> ~remaining time is 4h according to ddrescue
[23:04] <TJ-> I have to go now, gone midnight here
[23:05] <Guest63> oh. should i ping you tomorrow on what the next steps will be?
[23:07] <TJ-> I may or may not be around but there should be others around who can advise. Basically, if ddrescue manages to get most of the data then you can try using the fsck tool (initially in the do-not-change-anything mode!) on the sdb1.img file (which is a file-system image). If that suggests there aren't too many faults and it can fix them, then its worth allowing it too.
[23:08] <TJ-> If there are too many faults then you might need to use testdisk and/or photorec - likely only photorec is going to be of use in that case and it isn't certain how good a job it can do to recover the data in an understandable form
[23:09] <Bashing-om> Guest63: ^ I might suggst that you work on a copy of the sdb1.img file for safety reasons/
[23:10] <Guest63> TJ- thank you
[23:10] <Guest63> Bashing-om thank you. will def do
[23:10] <TJ-> Guest63: as Bashing-om  says, if you've got space, make a copy of sdb1.img and work on the copy :)
[23:11] <Guest63> i do and i will
[23:11] <TJ-> and ensure you use the 'cp --sparse=always ...' option so the file size doesn't balloon
[23:12] <Guest63> will read what that means. i wasn't aware of that flag
[23:12] <Guest63> btw, 16m in and 2 errors so far
[23:12] <TJ-> it'll ensure those unused sectors that don't take up space in the sdb1.img remain that way in the copy
[23:12] <Guest63> oh, okay
[23:13] <TJ-> got to go, eyes are dying
[23:13] <Guest63> tyvm. g'night
[23:16] <dreamon> hello. how can I remove a service from systemctl?
[23:17] <dreamon> systemctl list-dependencies systemd-suspend.service → shows nvidia-resume.service .. I want to remove this service
[23:18] <dreamon> Or reset this service to its standard defaults
[23:18] <sarnold> dreamon: there's a systemctl preset  command, check the manpage to see if it does what you want it to do
[23:18] <SysGhost> you can use "disable" with systemctl to disable it directly. But if something still needs it it will still be used.
[23:20] <dreamon> thanks. systemctl disable nvidia-suspend.service Unit /etc/systemd/system/nvidia-suspend.service is masked, ignoring.
[23:20] <Bashing-om> dreamon: I.E, ^^ a "disable" stops the service but it can be restarted by you or another process. A "mask" makes the service un-startable until you "unmask" it.
[23:21] <dreamon> Bashing-om, I had a nvidia gpu instide and removed it. now nvidia-suspend stops suspend from working normally, I beleave.
[23:21] <dreamon> How kann I unmask?
[23:21] <SysGhost> masking dependencies might result in other services to fail that needs it.
[23:21] <sarnold> there's a systemctl unmask command
[23:22] <Bashing-om> dreamon: ^ According to the man pages "systemctl unmask NAME" will undo the effect of "systemctl mask NAME".
[23:22] <dreamon> maybe I uninstalled nvidia in wrong way. so some parts are still alive.
[23:23] <dreamon> Is there a uninstall routine for nvidia that fixes all that stuff?
[23:25] <dreamon> I can suspend by calling "pm-suspend" but "systemctl suspend" dont work
[23:25] <Bashing-om> dreamon: Most depends on how you initially installed the driver,. what shows now ' dpkg -l | grep -i nvidia ' ?
[23:26] <dreamon> there is nothin more install only (rc) no (ii)
[23:29] <Bashing-om> dreamon: rc == (R)emoved but (C)onfig files remain. One can run ' dpkg -l | awk '/^rc/{print $2}' | xargs sudo dpkg -P ' to also remove these. But in this case do observe what the system will remove.
[23:30] <dreamon> It works.. after unmasking and "systemctl disable nvidia-resume.service" and "systemctl disable nvidia-suspend.service" I can suspend!!
[23:31] <Bashing-om> !yay | dreamon
[23:31] <sarnold> wootr
[23:31] <dreamon> Bashing-om, yes this command remove lot of config stuff..
[23:32] <dreamon> that cost me lot of days.. thank you so much guys. I have it back.. so lucky
[23:33] <Bashing-om> retep: Uh Huh - if you have an older install there will be lots of cruft around, Just note what will be removed and "do it".
[23:33] <dreamon> Bashing-om, (C) config remain files I didnt knew. Thanks!
[23:35] <Bashing-om> dreamon: Thing here is that the package manager will not remove config files without that you tell it too, in the event that you want a package re-install to have those old config files to use.
[23:38] <dreamon> I cannot test. but think it would be interessting to know, if removing this config would bring back suspend again.
[23:39] <Bashing-om> dreamon: I would not think so at all - config files after all control application action.
[23:41] <dreamon> Bashing-om, Is it possible to list, all those config files that a pakage is using?
[23:43] <Bashing-om> dreamon: ' dpkg -L <packagename> '  is good to see the full list of files a package installs.
[23:44] <sarnold> I think the maintainer scripts for a package can also maintain files outside of that list
[23:45] <sarnold> because dpkg's handling of config files can be pretty brutal
[23:46] <dreamon> Cool. this makes it much easier to unterstand for me.
[23:47] <dreamon> but it dont seperates between config and programm stuff.
[23:51] <Bashing-om> dreamon: /var/lib/dpkg/info/<package_name>.conffiles :list of configuration files. (user modifiable).
[23:54] <dreamon> Bashing-om, I think I have to come here more often. I learned a lot today!!
[23:55] <Bashing-om> dreamon: The collective knowledge here is indeed amazing :D