[00:22] <Nick01> hello!
[00:28] <bobdobbs> Hi all. I'm using ubuntu 22.04. I tried to update my drivers for my RTX 3070, and now my display is completely borked. I've been using linux/ubuntu for many years, but I've got no idea how the video system works or how to fix it when it collapses.
[00:28] <bobdobbs> At the moment I can run gnome, but only one of my two displays is running. The second display is blank
[00:29] <bobdobbs> I can run 'nvidia-settings', but it doesn't show information on my monitors
[00:29] <bobdobbs> Also, I have no gnome toolbars available, so I can't access settings.
[00:29] <Bashing-om> bobdobbs: How did you do that driver update ? -- Is there now a need to Un-do ?
[00:30] <bobdobbs> Bashing-om: I'm not even exactly sure what I was doing. Initially I was trying to install drivers from nvidia. But I was following instructions from online forums without really understanding them.
[00:32] <bobdobbs> So yeah - there may well be things that I have to undo
[00:32] <Bashing-om> bobdobbs: Oh Gosh darn - Nvidia advises not to do OEM install on linux systems - lemme pull up that warning,
[00:33] <bobdobbs> what does 'oem install' mean?
[00:34] <Bashing-om> bobdobbs: "Note that many Linux distributions provide their own packages of the NVIDIA Linux Graphics Driver in the distribution's native package management format. This may interact better with the rest of your distribution's framework, and you may want to use this rather than NVIDIA's official package.".
[00:34] <bobdobbs> hmmm
[00:34] <oerheks> use the drivers ubuntu proveides, they are tested
[00:36] <bobdobbs> well, at the moment I don't know if I've got the ubuntu driver or the nvidia driver installed. How do I check? I really don't know a thing about ubuntu video stuff.
[00:37] <bobdobbs> the command 'nvidia-driver' returns nvidia-driver-520. But I'm not sure what that means
[00:38] <Bashing-om> bobdobbs: Let's see about cleanup: what shows ' sudo find / -name "NVIDIA-Linux-* ' ?
[00:38] <oerheks> 470 and up should work https://www.nvidia.com/Download/driverResults.aspx/177145/en-us/
[00:38] <oerheks> but do not use their drivers..
[00:40] <bobdobbs> Bashing-om: that find command has been running for a bit. Hasn't returned anything yet
[00:40] <bobdobbs> oerheks: why should I avoid their drivers?
[00:42] <Bashing-om> bobdobbs: Patience, "Find" has to do a lot of looking :P
[00:43] <bobdobbs> Bashing-om: command completed. It returned a few "permission denied" line, and a few "no such file or directory". But no hits, it seems.
[00:45] <Guest67> hello. i am on my flash drive
[00:45] <Guest67> ubuntu * thumb drive, live ISO
[00:45] <Guest67> I don't see an IRC client, and the App Store doesn't work
[00:46] <oerheks> just use webchat, guest
[00:46] <Guest67> for now...
[00:46] <Guest67> I am downloading Konversation
[00:47] <Guest67> is there a list you can get to of installed Applications?
[00:47] <Guest67> instead of those large obnoxious icons?
[00:47] <Guest67> Maybe a desktop menu that shows a hierachy of apps?
[00:49] <Bashing-om> bobdobbs: Now then I am uncertain of a best way to proceed - however - pastebin the result of ' dpkg -l | grep -i nvidia '.
[00:49] <Nick02> hello.
[00:50] <oerheks> guest the is gnome-menus, but what is the use in a live system that does not save any.. and you will run out of memory soon with many apps installed
[00:50] <oerheks> !info gnome-menus
[00:51] <Nick02> how do i get the app store to work?
[00:51] <Nick02> it wont load anything
[00:52] <oerheks> run sudo apt update before entering app store.. on  a live session
[00:52] <oerheks> but why?
[00:52] <Nick02> hmm
[00:58] <bobdobbs> Hi. I'm back. I just rebooted. I now have a toolbar and panel. Still no activity on my second monitor
[00:59] <Bashing-om> bobdobbs: And my last " Now then I am uncertain of a best way to proceed - however - pastebin the result of ' dpkg -l | grep -i nvidia '.
[01:00] <Bashing-om> bobdobbs: And my last " Now then I am uncertain of a best way to proceed - however - pastebin the result of ' dpkg -l | grep -i nvidia '.
[01:00] <bobdobbs> Bashing-om: sure thing...
[01:01] <bobdobbs> https://hastebin.com/usabejarod.apache
[01:03] <Diagon> I'm starting to get problems with packages being kept back on dist-upgrade with ubuntu 22.04.  First it was some dependencies of jami, now it's some dependencies of jsonlint.  I've never had this problem before.  What gives?
[01:03] <Bashing-om> Diagon: !phasedupdates
[01:04] <bobdobbs> Bashing-om: well, I certainly seem to have lots of nvidia-related stuff installed
[01:05] <Bashing-om> bobdobbs: Yeah ,, and looks like a driver conflict between 2 veesions. How about we clean things up and have the kernel install the driver it thinks best ?
[01:05] <bobdobbs> sounds good :)
[01:05] <Bashing-om> !updates |  Diagon
[01:06] <Diagon> Thanks, Bashing-om.  Finding something here: https://askubuntu.com/questions/1431940/what-are-phased-updates-and-why-does-ubuntu-use-them
[01:10] <bobdobbs> Bashing-om: in terms of diagnostics, is this helpful?  https://hastebin.com/qinecivole.yaml
[01:13] <bobdobbs> Bashing-om: so, how do we go about cleaning things up?
[01:19] <Bashing-om> bobdobbs: the grep output shows a driver comflict -as the find result does not find a Nvidia (O)rininal (E)quipment (M)anufacture driver - All I know to do now is purge the present Nvidia driver(s) installs - update; and have the kernel install a driver.
[01:20] <Bashing-om> conflict*
[01:20] <bobdobbs> ok
[01:22] <bobdobbs> how would I do that/
[01:23] <Bashing-om> bobdobbs: In small steps - 1) run ' sudo apt remove --purge nvidia-* ' -- rboot; should come back up in the nouveau driver and 2) next run ' sudo apt update ; sudo apt upgrade '. Advise when at this point to continue.
[01:23] <Bashing-om> reboot*
[01:24] <bobdobbs> ok. I might as well just nuke it all.
[01:24] <bobdobbs>  
[01:25] <bobdobbs> k, I've purged everything. Will reboot. Crossing my fingers that I'll get video signal after reboot
[01:25] <bobdobbs> brb (I hope)
[01:30] <bobdobbs> k, I'm back. I do indeed have video signal
[01:30] <bobdobbs> back to boring-old single-monitor
[01:31] <Bashing-om> bobdobbs: Progress made :D .. ok now let's get rid of the cruft: dpkg -l | awk '/^rc/{print $2}' | xargs sudo dpkg -P
[01:32] <bobdobbs> dang, that's one gnarly command
[01:33] <bobdobbs> Bashing-om: k, that purged a lot of stuff. Mostly nvidia-stuff, it seems
[01:33] <Bashing-om> bobdobbs: While there is no built in way to remove all of your configuration information from your removed packages ( rc) you can remove all configuration data from every removed package.
[01:34] <bobdobbs> thats what 'purge' does, right?
[01:34] <tomreyn> Diagon: jsonlint is in universe, though, i would not expect that to be phased, so maybe you have another issue.
[01:34] <bobdobbs> removes configs along with everything else?
[01:34] <Bashing-om> bobdobbs: confirm that you have updated/upgraded the system at this point.
[01:35] <bobdobbs> yup. after I logged back in I did 'apt update' and then 'apt upgrade'
[01:35] <tomreyn> Diagon: apt policy jsonlint     should either say that it's "phased" or not
[01:35] <Bashing-om> bobdobbs: The flag ^ -- is "rc' where that means config files remain.
[01:37] <bobdobbs> ah, I see
[01:37] <Bashing-om> bobdobbs: Let's have the kernel install the proprietary driver ' sudo ubuntu-drivers autoinstall' . Reboot again  to see the effect.
[01:38] <bobdobbs> kk
[01:38] <bobdobbs> oh dear
[01:39] <bobdobbs> https://hastebin.com/ubiqoqiwil.sql
[01:40] <bobdobbs> Bashing-om: that looks like it might be an actual ubuntu bug
[01:44] <bobdobbs> Bashing-om: yeah, it does look like an ubuntu bug. But I think I've found a workaround. An edit to a file in a system library file
[01:44] <Bashing-om> bobdobbs: You on the Wayland desk top ? As it is known that Wayland and Nvidia do not play nice together.
[01:46] <bobdobbs> afaik I'm not on wayland. I experimented with wayland when I got this sytsem. It was a complete disaster, so I nuked it
[01:46] <bobdobbs> I nearly had to completely reinstall my system
[01:46] <bobdobbs> k, I got ubuntu-drivers working. Will paste the output...
[01:47] <bobdobbs> https://hastebin.com/azecegizaq.apache
[01:48] <bobdobbs> Bashing-om: k, what I read from that is that the system wants to install a driver version that is not compatible with the kernal
[01:49] <bobdobbs> oh wait. I'm wrong
[01:49] <bobdobbs> The problem is "held broken packages". But I'm not sure what that means. Googling it.
[01:52] <Bashing-om> bobdobbs: Ouch ! Let's try ' sudo apt autoclean ; sudo apt -f install ; sudo dpkg --configure -a ' . Then run this again: sudo apt -f install - see if the package manager will heal it's self.
[01:54] <bobdobbs> Bashing-om: alas, no. I ran those commands in the order you gave them, and then ran 'ubuntu-drivers autoinstall' again. I get the same result
[01:54] <bobdobbs> one monitor makes me sad
[01:55] <tomreyn> i haven't followed the conversation, but https://hastebin.com/azecegizaq.apache looks like packages are installed in a version that's not available for ubuntu (i.e. third party, without active apt source).
[01:55] <tomreyn> if that's intentional, please just ignore
[01:57] <bobdobbs> tomreyn: I may have installed something without an apt source. I don't know.
[01:57] <Bashing-om> !info nvidia-driver-470 | jammy
[01:57] <bobdobbs> Actually, I'm pretty sure I've installed things on this system after directly downloading. So I'm likely to have things installed that don't have an apt source.
[01:58] <tomreyn> apt list --installed | grep ',local\]$'      should tell you which packages / package versions are installed without active apt source providing it
[01:58] <bobdobbs> https://hastebin.com/alezudezat.http
[01:59] <bobdobbs> I don't think I know what surge-xt is
[01:59] <bobdobbs> That list has 'cuda-repo...'. I'm guessing that's a package that touches apt sources files?
[01:59] <tomreyn> cuda-repo-ubuntu2204-11-8-local may have added file:/ URIs to apt sources
[02:01] <bobdobbs> tomreyn: yeah. I think that that adds a source for an nvidia driver. Or perhaps for nvidia's cuda toolkit software.
[02:01] <tomreyn> apt policy libnvidia-gl-520     should show which sources this is available from
[02:01] <bobdobbs> https://hastebin.com/owasemixuw.apache
[02:02] <tomreyn> see, the newer version that's not available in ubuntu is from the local apt repository.
[02:02] <bobdobbs> huh
[02:04] <tomreyn> what's your goal? removing third party drivers, going back to ubuntu's?
[02:04] <bobdobbs> tomreyn: yeah, now I'm wondering which direction I should Gorenoise_
[02:04] <bobdobbs> lol. damn autocorrect
[02:05] <bobdobbs> tomreyn: I'm wondering which direction I should go.
[02:05] <bobdobbs> In the short term I want to just get my displays working. Ubuntu's driver should work for that.
[02:05] <tomreyn> i'm not so much into nvidia, especially not cuda. all i could possibly guide on is how to restore what ubuntu provides.
[02:06] <bobdobbs> But I also want to get stable diffusion working locally. And I don't know if ubuntu's drivers get in the way of that or not. I suspect not, but I don't know.
[02:06] <bobdobbs> afaict, stable diffusion relies on python's torch library to use cuda.
[02:07] <bobdobbs> actually, I should be able to find out if ubuntu's native drivers should work. I'll need to do a bit of googling.
[02:10] <Nick02> hello!
[02:11] <bobdobbs> If I tried to run the nvidia drivers, would I still be eligable to ask for help on this chanel?
[02:12] <Bashing-om> bobdobbs: Again Nvidia suggest *NOT* to do that !
[02:13] <Nick02> can someone tell me how to get the app store working on ubuntu?
[02:13] <Nick02> fresh live cd
[02:14] <bobdobbs> Bashing-om: yeah, I've encountered that suggestion. But still... I'm not 100% that stable diffusion can make use of an RTX with the native drivers
[02:14] <bobdobbs> Bashing-om: I think I should be able to. But I'm still figuring it out
[02:15] <Diagon> tomreyn - so what do you make of this? https://termbin.com/rwh6
[02:16] <Bashing-om> bobdobbs: Ubuntu and Nvidia do have a close working relationship - if something does not work - then one files a good bug report to get the issue resolved.
[02:17] <Diagon> I know smartctl is hit or miss with USB, depending on pass-through, but how about with USB-C?  Does anyone know?
[02:17] <tomreyn> Diagon: not phased
[02:17] <Diagon> tomreyn / oh.  So what might be going on there?  What might I look into?
[02:18] <bobdobbs> well, I haven't found anything about issues related to the ubuntu native drivers and pytorch. So... might as well go with the native drivers
[02:18] <bobdobbs> how do I get started?
[02:19] <Diagon> tomreyn / I mean, I can install them individually and then set them back to auto, but I'd like to know what's going on...
[02:19] <tomreyn> Diagon: apt list --installed | grep ',local\]$'      should tell you which packages / package versions are installed without active apt source providing them. i suspect you have packages / package versions installed from an apt source that is not ubuntu's, which is no longer configured.
[02:20] <Diagon> tomreyn / Oh, ok.  That's possible as this is an upgrade from 20.04.  I'll check, hold on...
[02:22] <Diagon> tomreyn / I'll have to look through these ... https://termbin.com/spe7
[02:23] <tomreyn> bobdobbs: remove the local nvidia repository, probably by apt purge'ing this package providing it
[02:24] <bobdobbs> tomreyn: sorry, I'm a bit of a dimwit when it comes to apt. How do I go about doing this?
[02:24] <tomreyn> Diagon: this doesn't seem immediately related, but you should remove those packages you don't know / know you need to be installed like this, without an upgrade path and without getting security patches.
[02:25] <Diagon> tomreyn / right, so if I'm understanding correctly, if the package is in a proper sources.list and the repo is working, then the package should be ok.  Otherwise, I should be checking what's up.
[02:25] <tomreyn> bobdobbs: https://docs.nvidia.com/cuda/cuda-installation-guide-linux/index.html#removing-cuda-tk-and-driver
[02:25] <bobdobbs> tomreyn: thanks
[02:26] <Diagon> tomreyn / so I might do a reverse depends on those php packages and see which of that list are depending on them, yes?
[02:27] <tomreyn> Diagon: sorry, i'm loosing track, can't seem to focus on supporting two apt issues at a time
[02:27] <Diagon> :)
[02:28] <bobdobbs> tomreyn: Bashing-om: k, that nvidia documentation for purging their drivers had some... purge comands. I've run them. No errors. I'm just gonna reboot. Back in a few minutes
[02:29] <tomreyn> Diagon: had you run   sudo apt update     before you ran     sudo apt-get dist-upgrade    ?
[02:30] <Diagon> Yes.
[02:30] <tomreyn> Diagon: can you show   sudo apt policy php8.1-common
[02:31] <tomreyn> actually no sudo needed there
[02:31] <Diagon> bobdobbs / I will need to purge an nvidia driver in the future too.  To save my asking for help, can you point me to the doc?
[02:31] <tomreyn> the same documentation you followed when installing it.
[02:32] <Diagon> tomreyn / https://termbin.com/dole4
[02:32] <tomreyn> Diagon: okay so *this* is phased, as you can see
[02:32] <tomreyn> so nothing to do here
[02:32] <tomreyn> Diagon: don't use aptitude though, it has problems
[02:33] <tomreyn> at least not for package installations and apt cache update
[02:33] <Diagon> tomreyn / what's the alternative to `aptitude why` and ... some reverse depends syntax...
[02:34] <tomreyn> apt-rdepends
[02:34] <tomreyn> and 'apt policy' for the 'why'
[02:35] <tomreyn> and 'apt depends'
[02:36] <Diagon> tomreyn / Sorry, one more: `aptitude show` is the one way I know to get long output about a pkg.
[02:37] <tomreyn> i don't remember what that looks like, but suspect it'll be similar to   apt show <package>    and    apt-cache show <package>
[02:37] <Diagon> tomreyn / apt depends is not in the man page!  (Bug report necessary...)
[02:37] <tomreyn> got for it
[02:38] <tomreyn> *go for it
[02:38] <tomreyn> !bug | Diagon
[02:38] <tomreyn> it would be good to first check whether that already exists
[02:39] <Diagon> apt-rdepends is a separate package ...
[02:40] <Diagon> tomreny / ok, so is it possible that the reason php was held back, is that there is a package that needs the older version?
[02:40] <bobdobbs> hi ho
[02:41] <bobdobbs> Bashing-om, tomreyn. I'm back. I have video. And I think I've installed a driver.
[02:41] <tomreyn> Diagon: the reason the php update was not yet installed is phased updates. i was just assuming otherise initially since i had you check whether an unrelated package was phased.
[02:41] <bobdobbs> I'm not sure if it's loaded though. And I suspect that it's an nvidia one?
[02:42] <tomreyn> bobdobbs: if nvidia-xsettings works, then you're running the proprietary nvidia driver.
[02:42] <Bashing-om> bobdobbs: The file /var/log/gpu-manager.log shows all happy happy ?
[02:42] <bobdobbs> I ran 'ubuntu-drivers autoinstall'. It installed a driver. If I run it now, it says "All the available drivers are already installed". So that's a start
[02:43] <bobdobbs> tomreyn: do you mean 'nvidia-settings'? That runs but doesn't bring up information on the displays.
[02:43] <Diagon> Ok, many thanks tomreyn!  Saved the day again!
[02:44] <bobdobbs> Bashing-om: the output of that log file: https://hastebin.com/didelaqafu.apache
[02:44] <Diagon> Now, back to a previous question.  I know smartctl will only work with some USB->SATA bridges, if they have pass-through.  Is that also true with USB-C?
[02:44] <tomreyn> Diagon: you're welcome. the 'fail to full-upgrade issue' should sort itself out in a few days.
[02:45] <tomreyn> bobdobbs: i guess i mean nvidia-settings, yes if it doesn't work properly maybe that's because you're on wayland, or because you're running a different driver, or both.
[02:46] <bobdobbs> Pretty sure I'm not on wayland. I'll try and figure that out...
[02:46] <tomreyn> Diagon: i have no first hand experience, but would expect so, yes
[02:47] <bobdobbs> tomreyn: I'm using x11
[02:47] <tomreyn> so     echo $DESKTOP_SESSION $XDG_SESSION_DESKTOP $XDG_CURRENT_DESKTOP    does not say wayland anywhere?
[02:48] <bobdobbs> Yes. None of those variables show anything that looks like 'wayland'
[02:49] <tomreyn> and    lspci -knnd ::0300 | grep 'in use'    says "nvidia"?
[02:49] <bobdobbs> yes
[02:50] <tomreyn> then, unless it failed to load, you're probably running ubuntu's nvidia driver
[02:51] <tomreyn> you'd need to look for the current xorg log if you wanted to know whether the driver failed to load, i'm not aware of another way
[02:53] <tomreyn> glxinfo -B | grep 'renderer string'    would also hint on it if you have glxinfo installed.
[02:53] <bobdobbs> That returns: OpenGL renderer string: llvmpipe (LLVM 13.0.1, 256 bits)
[02:54] <tomreyn> hmm, that sounds like software rendering. could be normal for the -open driver, though, if you're using that
[02:55] <tomreyn> you may want to look at the log file Bashing-om had pointed out 14 minutes ago
[02:55] <bobdobbs> Does  line 12 of this output show us that an nvidia driver is being used? https://hastebin.com/xuqicerala.yaml
[02:56] <bobdobbs> oh. I thought I pastebinned that
[02:56] <bobdobbs> lets see...
 Bashing-om: the output of that log file: https://hastebin.com/didelaqafu.apache
[02:56] <tomreyn> sorry, i missed that
[02:57] <bobdobbs> https://hastebin.com/gobiwihuro.apache
[02:57] <Bashing-om> tomreyn: bobdobbs: "Error : Failed to open /dev/dri" indicates to me that the driver did not load :(
[02:57] <bobdobbs> drat
[03:00] <bobdobbs> "operating system says no"
[03:00] <tomreyn> the log states the nvidia kernel module was loaded, but i agree the 2d and 3d drivers seems to have failed to initialize.
[03:00] <bobdobbs> I wish it would tell us *why* it's not loading the driver
[03:00] <Bashing-om> bobdobbs: I do not presently have a clue what in the world these "/kernel/nvidia-515srv" could be :(
[03:01] <tomreyn> this should be the headless drivers, no?
[03:01] <bobdobbs> headless?
[03:01] <tomreyn> the -server packages
[03:02] <tomreyn> https://packages.ubuntu.com/search?keywords=nvidia-driver-515
[03:04] <tomreyn> so https://packages.ubuntu.com/jammy-updates/nvidia-dkms-515-server
[03:05] <tomreyn> bobdobbs: the xorg initialization logs should be in ~/.local/share/xorg/Xorg.*.log
[03:07] <tomreyn> if not there, then somewhere "Xorg"-ish in /var/log/
[03:07] <bobdobbs> tomreyn: I do have a log there. But it's too big to copy and paste to hastebin
[03:07] <bobdobbs> oh, I think the system log is more copyable
[03:07] <tomreyn> log for lines with (EE)
[03:08] <tomreyn> *look for
[03:08] <bobdobbs> https://hastebin.com/opazecavuw.less
[03:09] <bobdobbs> That's just the output of 'cat Xorg.0.log | grep  EE'
[03:10] <tomreyn> hmm, i'm afraid it doesn't tell us what exactly failed.
[03:11] <tomreyn> maybe it says near those "[drm] Failed to open DRM device for pci:0000:2d:00.0: -19" and "open /dev/dri/card0: No such file or directory" lines
[03:11] <bobdobbs> from memory there's a commandline tool that can shunt a file to a pastebin. I'll see if I can find it
[03:12] <tomreyn> pastebinit is one
[03:12] <tomreyn> but i'm getting really tired now, need to sleep. good luck on your quest!
[03:12] <bobdobbs> thanks tomreyn. sleep well.
[03:12] <tomreyn> thanks
[03:15] <bobdobbs> dammit. pastebinit requires an Ubuntu One account
[03:16] <Bashing-om> bobdobbs: termbin is the one I use :)
[03:18] <Bashing-om> bobdobbs: However, be aware I have been to this point of the "Failed to open /dev/dri" a couple of times before - have yet to have the skill to get beyond :(
[03:24] <bobdobbs> Bashing-om: I can't get it to work, and I'm having trouble finding an alternative.
[03:29] <Bashing-om> bobdobbs: To relate a file to termbin - something like ' nc termbin.com 9999 < <(cat /etc/hosts;) '.
[03:32] <bobdobbs> oh, I see the problem. I downloaded the wrong file. termbin.com links to the repo for a dependancy. I'll have to hunt around for the actual termbin script
[03:36] <bobdobbs> huh. Where do I get the actual termbin script from? I can't find it.
[03:39] <bobdobbs> I know that seems like a dumb question. I should be able to just Google for it and find it. But I can't find it online. It' own website doesn't even link to it.
[03:40] <matsaman> bobdobbs: it's a web app
[03:40] <Bashing-om> bobdobbs: struggling here to to know what nc entails.
[03:41] <matsaman> nc being netcat
[03:42] <bobdobbs> Bashing-om: nc is netcat. It's a network debugging tool. I think it's basically meant to be pluggable into scripts and applications as a way to send data directly to TCP ports.
[03:43] <bobdobbs> So termbin uses netcat to send text files to pastebin-like services that offer a port for public service
[03:43] <bobdobbs> but a couple of years ago most of those services shut down en-masse, because they were being abused
[03:43] <matsaman> I don't know about that
[03:43] <Bashing-om> bobdobbs: Many uses - trying to determine how I *DID* install netcat for my use cases :)
[03:43] <matsaman> http://ix.io/
[03:44] <bobdobbs> I mean, those services shut down the capacity to accept pastes directly from things like termbin, because termbin and related tools can be part of automated toolsets for abusers
[03:45] <matsaman> that's how it has always been with web apps
[03:45] <matsaman> and will always be
[03:45] <matsaman> you have to safeguard against it on the receiving end
[03:45] <littlequestion> hello everyone
[03:45] <matsaman> hi little
[03:45] <littlequestion> is ubuntu servers on maintaince ?
[03:46] <littlequestion> i mean packages.ubuntu.com/search?keywords= giving me error 500
[03:46] <bobdobbs> In the meantime - I'm still kinda screwed. I need to get wiser eyes than mine to look at my log files. But I can't paste them to a public service, cos they are too big.
[03:47] <littlequestion> now start to work
[03:47] <matsaman> bobdobbs: what's the broader issue?
[03:47] <littlequestion> looks like an update or ddos happend . but look like it fixed
[03:47] <littlequestion> thx
[03:48] <littlequestion> nope still same
[03:48] <bobdobbs> matsaman: I'm trying to get an nvidia driver working on my RTX card on ubuntu 22.04
[03:48] <matsaman> yeah packages.ubuntu.com failing for me just now
[03:48] <matsaman> bobdobbs: which model?
[03:48] <bobdobbs> 3070ti
[03:48] <matsaman> and which driver/version are you trying?
[03:49] <bobdobbs> I've installed one from the ubuntu repos. It's installed but not loading. I'll try and discover which version
[03:50] <matsaman> dpkg -l | grep -i nvidia
[03:51] <bobdobbs> matsaman:  https://hastebin.com/qimodeloja.apache
[03:56] <bobdobbs> Are there specific forums for getting help with ubuntu video issues?
[03:57] <matsaman> here is probably best
[03:58] <Guest1> IHi guys, I'm seeking advice on good full-stack web-dev frameworks, considering Django vs Ruby on Rails, and WHY. =)
[03:58] <Guest1> I'm seeking advice on how to install Hardened Linux Kernel for my webserver, and if the hardened kernel would break/change any software, or is the hardening 100% passive/doesnt change settings?
[04:00] <matsaman> bobdobbs: what makes you think it's not loading?
[04:00] <matsaman> Guest1: rails is _horrible_
[04:00] <matsaman> just try installing it compared to django and you will see why
[04:00] <matsaman> honestly I find ruby to be a pretty nice language, but rails is unbearably mismanaged and beastly
[04:01] <matsaman> django is pretty good, though you should probably also look at flask & fastapi if you look at django
[04:01] <matsaman> Guest1: hardening with enough effort doesn't limit functionality, but it can be much effort
[04:01] <matsaman> Guest1: for a _server_ it's less problematic, because servers are less complicated than desktops
[04:01] <Guest1> Is there a way to harden easily without breaking software
[04:02] <Guest1> Like, how about Hardened Kernel patch
[04:02] <Guest1> for starting
[04:02] <matsaman> there is a certain amount of drop-in stuff like
[04:02] <matsaman> apparmor and things
[04:02] <matsaman> I forget what Ubuntu supports the most
[04:03] <Guest1> How about the hardened linux kernel
[04:03] <matsaman> if you really care about security I would move over to Debian stable, though
[04:03] <Guest1> Debian more than Ubuntu
[04:03] <Guest1> why
[04:04] <matsaman> Debian has a stable branch, it's a bigger organization
[04:04] <matsaman> Ubuntu is sourced largely from Debian's unstable branch
[04:04] <matsaman> it is fundamentally less secure
[04:09] <Guest1> hmmm
[04:09] <Guest1> wow
[04:09] <Guest1> are there are debian light distros
[04:09] <Guest1> like lubuntu or xubuntu
[04:09] <Guest1> for debian though
[04:10] <matsaman> you can install Debian without a DE and then choose Xubuntu, LXDE, GNOME, etc
[04:10] <matsaman> but for an actual server install you of course would not
[04:10] <matsaman> (install a DE)
[04:10] <matsaman> and then choose Xfce*, LXDE... whoops
[04:10] <Guest1> What do u think is more lightweight
[04:10] <Guest1> Lubuntu, or Debian with LXDE
[04:11] <matsaman> historically it would've been LXDE but I feel like they've changed their configuration some
[04:11] <matsaman> Xfce is my favorite
[04:11] <matsaman> but I use Xfce on a distro that allows it to be built without optional GNOME support
[04:12] <matsaman> even using LXDE on a Debian or Debian-based distro (such as Ubuntu) might draw in a lot of GNOME stuff, though
[04:12] <Guest1> xfce then
[04:12] <Guest1> debian xfce
[04:12] <Guest1> nice
[04:12] <Guest1> I like how easy it is to install software on Ubuntu
[04:13] <Unit193> If by "GNOME stuff" you mean GTK, then yes. :>
[04:29] <matsaman> Guest1: how do you personally install stuff in Ubuntu?
[04:29] <matsaman> Unit193: I don't mean that, but it is true GTK and GNOME are increasibly inseparable
[04:29] <matsaman> which is unfortunate
[04:29] <Guest1> sudo apt install
[04:29] <Guest1> many software is built for ubuntu
[04:30] <matsaman> Guest1: well, it's built for Debian really
[04:30] <Guest1> i thought ubutnu had a dedicated decently good security team too
[04:30] <matsaman> apt is a Debian thing
[04:30] <matsaman> Ubuntu is almost entirely actually Debian
[04:30] <matsaman> Ubuntu has very, very few employees
[04:30] <matsaman> I suppose it has many more volunteers, but not more than Debian, not remotely close
[04:32] <bobdobbs> Bashing-om: hiya. I tried something else:  I discovered the ubuntu GUI for installing proprietary drivers.
[04:33] <wolfdale_> aa
[04:33] <bobdobbs> Bashing-om: I used it to install the nvidia driver. It installed the driver. But after a reboot, I'm still not able to load the card
[04:33] <bobdobbs> I just realized something - is there a chance that their is a hardware issue with the card?
[04:38] <Bashing-om> bobdobbs: As I have said - this is not the 1st time I have ran into this - a stone wall for me- I have yet to be able to see why the module - that builds fine - does not load :(
[04:39] <mozpedia> where does Ubuntu rank in popularity among distros?
[04:39] <bobdobbs> Yeah, it's a mystery to me too.
[04:40] <mybalzitch> bobdobbs: that would usually result in crashes or other weird behaviour
[04:40] <bobdobbs> Bashing-om:  I think I'll remove everything, then try and do a clean install of the proprietary drivers, then gather information and post on the nvidia forums.
[04:42] <bobdobbs> Bashing-om: I think I'd prefer to run the open drivers, but the nvidia forums offer support from paid staff. I've always found it hard getting help for video issues on ubuntu. It's just mysterious shiz that only a very few people seem to understand deeply enough to be able to help others.
[04:43] <Bashing-om> bobdobbs: If you are to do that - I would ne interested in seeing the present results from ' sudo modprobe nvidia ' .
[04:44] <bobdobbs> Bashing-om: at the moment that command produces no output
[04:45] <Bashing-om> bobdobbs: Hummm - and again I do not know what to make of "no output" :(
[04:45] <bobdobbs> same
[04:46] <Bashing-om> bobdobbs: No output many times means success --- what shows dor a driver from ' sudo lshw -C display ' ?
[04:48] <bobdobbs>  lshw -C display
[04:48] <bobdobbs> https://hastebin.com/ozesukibih.yaml
[04:49] <bobdobbs> I just tried something else. Gonna reboot and see what happens
[04:57] <Diagon> tomreyn told me to avoid aptitude.  I've been using it only for aptitude search/ show.  What other tool might show me the state of a package on a search?  (p,c,i,v,A) ie. installed, automatically installed, configuration files remain, virtual, etc.
[04:57] <Bashing-om> Diagon: ' apt policy <package> ' .
[04:59] <Bashing-om> Diagon: then too maybe ' apt show <package> ' .
[04:59] <Diagon> Bashing-om not quite doing it ...
[05:03] <Bashing-om> dia
[05:04] <Bashing-om> Diagon: As I have but sparringly ever used aptitude - I can not make the translations to apt :(
[05:07] <hiya> After installing Ubuntu with secureboot I get a screen that says perform mok management
[05:07] <hiya> What do I do there?
[05:08] <hiya> Enroll MoK ? Enroll key from disk? Or hash from disk?
[05:08] <hiya> I entered a password for secure boot
[05:08] <hiya> It's first boot post live iso
[09:17] <elias_a> I am using 22.04.1 on Thinkpad T450s. The fingerprint does not wake up sometimes when trying to log in after screen lock. Any hints how to get rid of this nuisance?
[09:27] <lotuspsychje> elias_a: can you keep a journalctl -f open and paste the output after lock/resume so the volunteers can have a look?
[10:33] <elias_a> lotuspsychje: Good idea! I will.
[11:38] <m> hey
[11:38] <lotuspsychje> welcome m
[11:38] <m> are you real?
[11:38] <lotuspsychje> yes, what can we do you for m ?
[11:40] <m> i just came here to tell you that i appreciate your work
[11:41] <lotuspsychje> did someone help you fix an issue m ?
[11:41] <m> didnt have an issue, just wanted to thank you guys for your work
[11:42] <lotuspsychje> !cookie  | everyone
[12:07] <meandrain> Hi, I have an ubuntu installation on regular SSD but I want to move it on nvme ssd. So I've installed plain ubuntu 22.04 on nvme and rsynced regular ssd installation to the nvme install. Now what do I need to do in grub or in other places so I would boot into nvme installation ?
[12:07] <meandrain> my motherboard does not know nvme boot (nvme drive is not seen in bios), so I would still have to use the ssd for grub boot loader
[12:33] <tomreyn> meandrain: if you have an efi system partition (ESP) on the SSD (but, preferrably, not the NVME) and have this configured on the new installations' /etc/fstab and run     sudo update-grub && sudo grub-install   there then this should install grub on the SSD's ESP
[12:34] <tomreyn> since rsync works on a file system level (not block device level), hopefully you had also created the same / suitable partitions on the nvme?
[12:35] <meandrain> I did because first I've installed a new ubuntu on the nvme, then I rsynced over that minimal installation
[12:35] <meandrain> is there a way to chroot to nvme and do grub-install ?
[12:35] <tomreyn> oh right, you said so, sorry
[12:35] <meandrain> (from usb stick)
[12:36] <tomreyn> you'll need to chroot into the new installation to run the    sudo update-grub && sudo grub-install     commands (after editing fstab)
[12:40] <tomreyn> https://help.ubuntu.com/community/Grub2/Installing#via_ChRoot
[12:40] <meandrain> tomreyn: tunning grub-update from chroot returns:    /usr/sbin/grub-probbe: error: cannot find a device for / (is it /dev mounted?)
[12:41] <tomreyn> see the link i just posted, you probably did not (bind) mount some of the virtual file systems
[12:42] <tomreyn> this means you'll need to unmount file systems you mounted from within the chroot, exit the chroot, mount the missing file systems, enter the chroot again
[12:43] <tomreyn> step 10 is where /dev gets bind mounted into the chroot
[12:44] <tomreyn> you will also need to mount the ESP in the chroot
[12:48] <meandrain> tomreyn: yes, mount bind was missing, I had to mount /dev/, /dev/pts, /proc, /run, /sys
[12:50] <tomreyn> and /boot/efi
[12:51] <tomreyn> meandrain: ^ because otherwise grub would have installed to the wrong place.
[13:04] <gneeriiloeepdeer> im using vlc to listen to old music on headphones with very bad stereo, meaning soundis only balanced if I use mono, but I have to change this parameter each time I replay the same song or any new song. How do I change it so vlc always plays mono?
[13:18] <PeGaSuS> gneeriiloeepdeer: try this: Tools > Preferences > Audio > Bottom left corner "Show settings" enable all > Audio > expnad `Output modules` > ALSA (or whatever you use) > Audio output channels > Mono - video example: https://filehost.0bin.xyz/files/HYguHxGNEik1bzAy.webm
[13:33] <BluesKaj> Hi all
[13:48] <xatrix> Hi, can you advice ? I'm using Ubuntu 22.04 with XFCE (latest). Also, i use Chrome as by default web browser. A few days ago something happened with it, and i can't ignite Open/Save dialog. So i can't upload images for example on a websites. Maybe i have some problems with QT/GTK or whatever it uses? This problem only affects Google Chrome. Any other software works fine. I also have a second laptop with the same system(cloned),
[13:48] <xatrix> and it has the latest updates as my primary system as well. And this laptop has no such issue. What should i inspect on my prim. system to find out what's wrong with it ?
[13:51] <tomreyn> xatrix: you could run chrome from a terminal and see whether it prints relevant error messages there. as well as watch   journalctl -f   for anything related logged there.
[13:52] <xatrix> tomreyn: thanks for the answer! Yea, i tried it. But i received no warning or debug messages when hit a hotkey CTRL+O or S
[13:52] <tomreyn> xatrix: i'd suggest also comparing to chromium-browser
[13:52] <xatrix> If it's possible, i'd like to stay in a stable branch of google-chrome
[13:53] <tomreyn> you'll need to get support from google then
[13:53] <xatrix> Yea, i wrote them. But.. No answer.
[13:53] <xatrix> Maybe it's not exactly a google problem
[13:53] <xatrix> Maybe something is wrong with my GTK\QT libs, or dbus or whatever else?
[13:54] <tomreyn> unlikely if there are only issues with chrome
[13:55] <gneeriiloeepdeer> PeGaSuS, thanks!
[13:55] <xatrix> Yes, but i have the same system on a second laptop. And all the same versions. And it works fine. Ofcource it was cloned a few months ago, and it has some minor differences as for now.But i'm not sure it affects
[13:58] <Guest2> Not sure if this was the best place to ask but I was looking at a buying couple of newer ultrabooks & notice some had DDR5 ram while others had LPDDR5. I was just wondering if LPDDR5 generates less heat than DDR5 even though, from what I've read, the only difference an average user would notice between the two is that LPDDR5 may extend battery
[13:58] <Guest2> life. Since my current laptop is LPDDR3, which runs really cool quiet, I'm just wondering if that translates to LPDDR5 or if plain DDR5 would still run the same as long as the CPU is power efficient.
[13:58] <_4four5ive> hello?
[13:58] <tomreyn> hello!
[13:59] <gneeriiloeepdeer> PeGaSuS, thanks but it doesnt work as intended
[13:59] <lotuspsychje> Guest2: sounds more like a question for  #hardware
[14:02] <_4four5ive> hey, i'm just getting started with Ubuntu. i'm trying to set up Swizzin on my old laptop, but the dashboard only shows up on ethernet (the laptop has ethernet and wireless). how do i make it so i can access the dashboard on both wireless and ethernet? i'm guessing i have to set up a static ip? i tried doing that for both wireless and ethernet, but it didn't work for me.
[14:05] <tomreyn> _4four5ive: is this "swizzin"? https://swizzin.ltd/
[14:06] <_4four5ive> yes, that's swizzin
[14:07] <tomreyn> okay, i had not heard of it until today. it's good to point out what third party software you're using and also, unless it is very common, where you retrieved it from, and how you installed it.
[14:07] <_4four5ive> yeah, i'm new to it too. it's for making your own seedbox. good to know, sorry.
[14:07] <lotuspsychje> _4four5ive: we got a bit of similar app in the ubuntu repos, stacer if you like
[14:08] <lotuspsychje> just not the seedbox part
[14:08] <tomreyn> i assume in this case you just downloaded the installation script over an insecure http download and gave it remote command ececution on your computer, as the website recommend
[14:08] <_4four5ive> lotuspsychje, oh, this is not just monitoring
[14:09] <_4four5ive> monitoring is only part of it
 just not the seedbox part, oh...
[14:09] <_4four5ive> but good to know
[14:09] <lotuspsychje> !info stacer  | _4four5ive
[14:09] <_4four5ive> tomreyn, lol what?
[14:10] <tomreyn> _4four5ive: i just described the "installation process" provided on the "swizzin" website
[14:10] <_4four5ive> i think i used wget
[14:10] <_4four5ive> to install swizzin
[14:11] <tomreyn> the rest of what i said still applies there
[14:11] <_4four5ive> im confused
[14:11] <_4four5ive> i think im missing something
[14:20] <meandrain> tomreyn: I did as you told me and it did work, thank you !
[14:21] <meandrain> I had to chroot grub update boot menu for nvme and also for the ssd where starts the boot loader
[14:21] <lotuspsychje> !cookie  | tomreyn
[14:21] <meandrain> I think a small part of boot loader (grub) starts from ssd and other part from nvme (I do not fully understand the process yet)
[14:22] <meandrain> (I have to boot from ssd because the motherboard has no nvme boot support)
[14:26] <meandrain> also grub-install did not worked in chroot, only update-grub
[14:26] <meandrain> ubottu: indeed, he is a great helper
[14:27] <tomreyn> meandrain: if you're EFI booting then the (grub) bootloader is fully stored on the ESP, but the grub configuration file is stored on /boot, as well as the initrd and kernel
[14:27] <meandrain> (noted !)
[14:28] <tomreyn> if grub-install did not work then you're still using the old grub binary from the other installation on the ESP
[14:29] <tomreyn> but you should ensure that grub.install can work now that you booted this new installation, otherwise installing updated grub package versions (and potential security updates) would also fail to install
[14:29] <tomreyn> * security + compatibility updates
[14:31] <tomreyn> _4four5ive: "swizzin" seem to provide support over discord, according to their website. their issue tracker is at https://github.com/swizzin/swizzin/issues
[14:43] <_4four5ive> tomreyn, thanks for letting me know. i dont think its a swizzin issue, though. btw, i dont have discord.
[14:50] <tomreyn> _4four5ive: i did not mean to recommend using discord (or swizzin). i assume the issue is related to it, but then i don't know what it does or how. if you run 'ss' or 'lsof' and look for its listening port, you will see which IP address the service making the dashboard available binds to, and this will tell whether it will be available over one or the other network link.
[14:51] <_4four5ive> tomreyn, brilliant idea. thank you!
[14:53] <tomreyn> you're welcome. and to answer your initial question: a fixed ip address assignment should not be needed (but may make it more usable)
[14:53] <_4four5ive> hmm, i wonder what the issue would be then
[14:57] <www2> Hi i have an problem with imputing the euro symbol when i press AltGr+5 i switch to an diffent tab in state of typing the euro symbol (ubuntu version 22.10)
[14:58] <tomreyn> and your keyboard layout is?
[14:58] <tomreyn> try   AltGr + e
[15:00] <EriC^^> exit
[15:01] <www2> @tomreyn US with euro symbole on the 5
[15:01] <EriC^^> you heard me, everyone leave, there's a fire!
[15:02] <www2> @tomreyn just found out that the seting that turn to do this was turn off in my settings
[15:03] <tomreyn> i see
[15:03] <tomreyn> btw, there is no need for the '@' (in '@tomreyn') on IRC
[15:04] <_4four5ive> ^
[15:04] <www2> oke it is more my style ;)
[15:05] <www2> and i have done this before twiter
[15:05] <EriC^^> www2: it actually might not highlight the person if you put something before the nickname
[15:05] <_4four5ive> i think im done with linux for today... spent about 6 hours configuring this....
[15:09] <www2> EriC^^ i have test between my accound and the @ before name symbol works
[15:10] <EriC^^> www2: it's client-dependent
[15:10] <www2> thanks i use hexchat as my client
[15:12] <_4four5ive> same
[15:12] <xatrix> tomreyn: i've tried chromium, and it seems like it doesn't open me Open/Save as well
[15:12] <EriC^^> no i mean it depends on the person you're highlighting's client, if it's written in a way that takes "nickname" and you put "@nickname" it might not let him know you messaged him
[15:12] <xatrix> tomreyn: but it opens me a print dialog
[15:13] <EriC^^> www2: ^ it's just something to consider
[15:36] <skr4to> hi all , skraito with Lord Jesus Christ Here ... . Open https://edx.org v72 Complete and always follow news from https://facebook.com/skraito.0day ... . ask your governor to open facebook and edx.org and those that We post in https://facebook,com/skraito.0day .... . for nothing more can We teach you ... .  https://worldhacker.org will be up soon for my  phd thesis .... .
[15:36] <Drone> Error: You don't have the #ubuntu,op capability. If you think that you should have this capability, be sure that you are identified before trying again. The 'whoami' command can tell you if you're identified.
[15:36] <jhutchins> www2: It also depends on what chat system you're on, irc clients tend to ignor the name if it has extra characters (leading @).
[15:38] <hiya> How to force discharge on power source with /sys/class/power_supply/BAT0 file? so it reaches 0 and charges then
[15:52] <Enissay> I need some guidance please: I have 3 servers each geographically separate from each other and all connected to internet. I would like to configure them so all inter-communication is passed through VPN while remain accessible directly via internet. I have already setup wireguard server on each and is working fine. But I am not sure how to proceed from here...
[16:25] <n0st0m0> join #haskell
[16:30] <alkisg1> Enissay: well, if they can see each other over wireguard in e.g. the 10.x subnet, then use the 10.x IPs for intercommunication; there's nothing more to it
[16:31] <Enissay> alkisg1: Not yet.. That's the point I am missing... So one of them is the server and the others are clients ? Also, I dont want all traffic to go through tunnel, only intra-traffic must be
[16:32] <alkisg1> Enissay: you can create a mesh in which everyone directly sees everyone, to avoid having to route all the traffic via a specific server
[16:34] <alkisg1> E.g. if you assign 10.1.1.1 to first, 10.1.1.2 to the second, and 10.1.1.3 to the third, and you assign the appropriate routes, then traffic to 10.1.1.x goes over wireguard, while the rest of the traffic goes over the normal routes
[16:34] <alkisg1> Not sure what you mean with "wireguard server" though, in network-manager it's just "add wg connection", and from there "add peers and routes"
[16:38] <five618> Hey all. So my server has been crashing lately and I wanted to figure out why from the logs. Strangely, the logs never contain anything about the crash. `journalctl -b 0` starts with a line "Sep 09: System clock time unset or jumped backwards, restoring from recorded timestamp...", and then it's this boot logs. But if I do `journalctl -b 1`, the
[16:38] <five618> latest logs are from Nov 22 (5 days ago!) even though my server has been running all the time in those 5 days. Any idea what's going on?
[16:53] <tomreyn> five618: i think you want    journalctl -b -1
[16:54] <five618> sorry that's what i used, just typoed here
[16:54] <five618> basically I think that journalctl drops my entire boot's logs if I do not clean shutdown
[16:54] <tomreyn> does    journalctl --list-boots    list the past boots though?
[16:55] <five618> no, it lists 14-22 nov, then 27-27 nov
[16:55] <five618> there's a 5 day gap that is definitely not true
[16:56] <tomreyn> if the system clock is broken, this could explain it
[16:57] <tomreyn> maybe the timesyncd logs in there will hint on a clock skew
[16:59] <tomreyn> oh, gone
[17:04] <five618> sorry tomreyn, my irc reconnected, ive missed any messages until now
 if the system clock is broken, this could explain it
 maybe the timesyncd logs in there will hint on a clock skew
[17:06] <tomreyn> but i'll afk for a bit now.
[17:06] <five618> worth noting, the "server" is a raspberry pi 4b, so it may not be able to persist a clock after crash
[17:07] <bG9s> five618: just curious if maybe /var/log/dmesg holds any dates from the missing range?
[17:08] <five618> I've disabled rsyslog in favor or journalctl, so that file is almost empty
[17:08] <five618> in favor of*
[17:08] <five618> as far as I understood, both journalctl and rsyslog were saving the same messages, one under /var/log/journal, the other uner /var/log/*
[17:08] <five618> so I disabled the latter as I prefer the flexibility of journalctl... was this a mistake?
[17:09] <bG9s> no
[17:09] <five618> strangely I just rebooted the system as a test, and my previous boot's logs are absolutely nowhere to be found
[17:10] <five618> the list boots shows the exact same data, and `journalctl -b -1` is sha256 identical to what i got when i ran the same command on previous boot
[17:10] <five618> and `journalctl -b 0` is fully different, it does not contain anything from previous boot
[17:11] <five618> yet my persistent logs are enabled, as I have logs from 5 days ago...
[18:06] <jhutchins> It's frustrating but true, if a system actually crashes, it ceases to write logs.  Logging is often the first thing to go, and sometimes is the thing that stops the rest of the system - sybsystems that won't run if they can't log.
[18:06] <jhutchins> It is theoretically possible for the system to shut down in an orderly manner, politely writing a crash dump and other diagnostics to the logs.
[18:07] <jhutchins> Of course, this assumes that the buffers are politely and successfully written to the disks, and the logs are not corrupt.
[18:07] <jhutchins> I have never actually seen such a polite, orderly shutdown, which we might call a preventative halt rather than a crash.
[18:08] <jhutchins> Logs are opened with read permissions, and often a lock that prevents other programs from trying to write to them simultaneously.
[18:08] <jhutchins> If these locks persist through the crash and are not cleared at boot, the system may appear to run normally but the logs will not be written.
[18:09] <jhutchins> Grr.  THey are opened with WRITE permission and often locked.
[18:22] <rrr> I cannot seem to instruct network-manager to create a hotspot, I suspect the adhoc ones work but cannot test that atm. I would like to know if I can make it work with as a normal hostspot tho.
[18:25] <rrr> I may have tried 100 different ways, including checking if hostapd works which do.
[18:34] <kubuntu> hi guys, someone who uses some tts in his distribution to read text?
[18:34] <kubuntu> Thx
[18:46] <five618> jhutchins: how could I troubleshoot this?
[18:46] <five618> I would have thought that logs are flushed to disk more often than at shutdown?
[18:46] <five618> it sounds impossible to me that a full week of logs go missing during a crash
[18:47] <tomreyn> https://wiki.ubuntu.com/Kernel/Debugging
[18:47] <tomreyn> https://wiki.ubuntu.com/Kernel/CrashdumpRecipe
[18:47] <jhutchins> five618: Narrow down what's happening when it crashes.  See if you can find one app or one site or one feed that does it.  Monitor your hardware - hard crashes are often heat related.
[18:47] <tomreyn> https://ubuntu.com/server/docs/kernel-crash-dump
[18:47] <jhutchins> This assumes you get a clean dump.
[18:47] <five618> oh, i know what causes it, it's my torrent client. I download kali linux and let it seed overnight, 100% guaranteed crash by morning
[18:47] <five618> I reproduced this three times in a row
[18:48] <five618> but the thing is, I changed torrent client and even underlying filesystem, and it did not help, I suspect it's the Raspberry's USB controller
[18:48] <five618> but at the moment I seem to have an even bigger problem - my logs do not persist even during a clean shutdown
[18:49] <five618> how could I troubleshoot my journalctl logs not saving?
[18:49] <tomreyn> oh this is a raspi, good to know. what media are you using?
[18:50] <tomreyn> storage media
[18:50] <five618> a ton of WD Elements 8TB USB disks with BTRFS in Raid1
[18:50] <five618> but the problem occurred even with ext4 in mdadm raid1
[18:51] <tomreyn> oof, are you sure this hardware can handle this? is there even enoough power supply?
[18:51] <five618> they are all externally powered, no exception
[18:51] <tomreyn> and the usb?
[18:51] <five618> err the usb hub is not, but if each disk is, does the hub need power too?
[18:51] <tomreyn> which raspi hardware version is this?
[18:52] <five618> it's the latest 4b revision, the one with extra voltage rail that can support 2ghz out of the box
[18:52] <tomreyn> almost all hardware components need power to function
[18:52] <five618> (8gb ram)
[18:53] <tomreyn> well, i don't know then - might work, or not. you should probably read up on other peoples' experiments with a similar setup
[18:53] <five618> if each disk is externally powered, and the usb hubs are only aggregating the usb ports, isn't that extremely low power consumption?
[18:53] <five618> my usb hubs do have the option to be externally powered, i just thought it was completely unnecessary
[18:54] <tomreyn> it's not just power consumption that you need to sort out, there can be subtle incompatibilities between different chipsets and devices, espeically in SOC land.
[18:55] <tomreyn> i'm not at all surprised that a raspi would power cycle unexpectedly during heavy use.
[18:55] <tomreyn> thermal issues can be a cause there, too
[18:55] <tomreyn> if you want stable server operation, get proper hardware.
[18:56] <five618> it's definitely not thermal, I have it externally cooled and it runs below 40 degrees even at full stress
[18:56] <arraybolt3[m]> If you're running a bunch of disks in RAID1, I would definitely suspect power consumption right off the bat. The Pi is designed to work with four USB devices at once (I believe) - if you're using a USB hub, you're going beyond that. The I/O might be low power consumption, but enough "low power consumption" adds up to high power consumption. If you can power the hub, I would do that before anything else.
[18:56] <arraybolt3[m]> (Especially if you run the thing at full stress sometimes.)
[18:58] <arraybolt3> And the fact that it intially works and then crashes at some point overnight - I would guess you start getting some people downloading from you, all the disks start having I/O at the same time, causing both CPU and USB usage to spike, voltage drops, Pi glitches, crash.
[18:59] <arraybolt3> (Disk I/O causes the CPU to start working, so it's not just the hub drawing power, the CPU will start drawing more power too I believe.)
[19:02] <arraybolt3> With that in mind, if you can't power the hub, you might be able to get away with underclocking and undervolting the CPU and see if that gives things the bit of extra power they need to work right.
[19:08] <five618> Thanks for the feedback everyone. I will get some power supplies and power my hub as soon as I can!
[19:11] <arraybolt3> five618: Another idea in the mean time might be to take a disk or two out of the RAID array to reduce power usage. If BTRFS allows that, that is. (Make backups if you have critical data before doing this if you choose to do that.)
[19:14] <anders> are you all ready to hang xmas lights?
[19:14] <krytarik> !chat
[19:14] <krytarik> And no..
[19:54] <arraybolt3> Silly question most likely, but I'm trying to basically make something that says "Open every file in every directory of the current folder recursively." Something like "for i in $(ls -R); do vim $i; done;". Trouble is that "ls -R" has this weird, human-readable output that vim won't take as input at all. Is there some way to do a recursive directory listing so that I get an entire path for every
[19:54] <arraybolt3> file (absolute or partial, doesn't matter as long as I can pass it as an argument to vim)?
[19:55] <arraybolt3> (Er, the format of ls -R is a weird human-readable format who's lines can't be passed as a command-line argument at all.)
[19:55] <rbox> find
[19:55] <arraybolt3> ...ok that was ridiculously easy.
[19:55] <bG9s> find is a good one
[19:56] <bG9s> vim -p `find . -type f`
[19:56] <bG9s> or vim -p `find . -type f -iname '*.c'`
[19:56] <arraybolt3> I just did "for i in $(find); do vim $i; done" and it worked.
[19:56] <arraybolt3> (Careful, don't use the above loop unless you want to get stuck in Vim, possibly for a good long time)
[19:56] <bG9s> killall vim
[19:57] <arraybolt3> Yeah but then the loop just opens a new one.
[19:57] <bG9s> poweroff
[19:57] <arraybolt3> while true; do killall vim; done
[19:57] <arraybolt3> Ctrl+C out when it works.
[19:59] <arraybolt3> bG9s: Your "vim -p `find . -type f`" lets me get out way easier.
[19:59] <bG9s> yes I know I'm a genius
[19:59] <bG9s> /s
[19:59] <arraybolt3> lol
[19:59] <bG9s> gT and gt
[20:00] <bG9s> move between tabs
[20:00] <bG9s> vim
[20:00] <arraybolt3> Nice.
[20:01] <arraybolt3> (Gotta love how Linux has some awesome tool for just about every single arcane usecase out there. I can't believe the answer to my question was a single command with no arguments, and then there were even other fancier ways of doing it.)
[20:01] <arraybolt3> rbox, bG9s: Thank you guys!
[20:02] <bG9s> o7
[20:02] <rbox> arraybolt3: unix has been around for a little bit...
[20:20] <karjala> Does anyone else experience this problem in Ubuntu 22.04? When I click on http&https links in emails in Thunderbird, nothing happens (the browser doesn't launch, doesn't open a page, etc)
[20:20] <karjala> I run Kubuntu
[20:20] <karjala> No computer I've ever installed Kubuntu 22.04 on works correctly in this regard
[20:21] <karjala> How do we solve this?
[20:25] <matsaman> karjala: maybe: https://superuser.com/questions/1650046/kde-plasma-5-reload-kdeglobals-or-set-default-browser-via-command-line
[20:25] <karjala> Thank you matsaman
[20:25] <Guest5121> Hi =)  Hi guys, would installing 'sudo apt install default-jre' expose my computer to more exploits, java based exploits?
[20:29] <mybalzitch> don't download and run untrusted java code?
[20:29] <matsaman> or any untrusted code
[20:29] <matsaman> applies to any system, not just java
[20:48] <Guest5121> Hi guys
[20:48] <Guest5121> Im seeking generic hardening advice for Linux ubuntu 22.04
[20:48] <Guest5121> desktop user
[20:51] <arraybolt3> Guest5121: Ubuntu 22.04 is already quite secure out of the box. Just keep the system updated and don't run untrusted code.
[20:51] <bG9s> yep, it uses apparmor out of the box right?
[20:52] <bG9s> and that's pretty good already, default policies enabled etc
[20:52] <arraybolt3> If installing Snaps, make sure that the publisher is safe.
[20:52] <arraybolt3> (Yep, pretty sure AppArmor is installed and running out of the box.)
[20:52] <arraybolt3> Guest5121: I would recommend enablin full disk encryption during installation, and if you can, use Secure Boot.
[20:53] <arraybolt3> Those are really the only two things I can think of right off the top of my head, and Secure Boot hardly adds anything and you can leave it off if you want/need to.
[20:54] <bG9s> I might add I believe in the compilation is really where the hardening effect is baked in, so we as end-users/desktop users get the pre-built hardened binaries
[20:54] <arraybolt3> (I have Secure Boot off on all of my systems. It *might* help in some instances, but those instances are rare.)
[20:54] <bG9s> then apparmor adds the security policies to protect during usage
[20:56] <Guest5121> I have secureboot On, and disk encryption
[20:56] <Guest5121> I update daily, have firewall on, ufw , and disabled webrtc and webgl in browsesr.
[20:56] <bG9s> then as a desktop user this is already quite good, the rest is handled by ubuntu crew when building the binaries
[20:56] <arraybolt3> Disabling webrtc and webgl sounds overkill to me but otherwise that looks great.
[20:57] <Guest5121> its jsut the beginning
[20:57] <bG9s> the best is yet to come?
[20:57] <Guest5121> the best is yet to com-pile
[20:57] <bG9s> ++
[20:58] <arraybolt3> Guest5121: Be careful - you could cause more problems than you fix if you try to do a whole bunch to your system security-wise. Ubuntu is built to be usable in enterprises, it is designed to be secure by default. Security is more about how you use is and less about what you do to it.
[20:58] <rob0> arraybolt3++
[20:59] <Guest5121> I mean
[20:59] <arraybolt3> s/use is/use it/ (why do I hit "s" rather than "t" so often...)
[20:59] <Guest5121> Yeah thats true I still want to harden it
[21:00] <arraybolt3> Guest5121: If you want to go really overkill like I do sometimes, install GNOME Boxes and install a Lubuntu VM into that. Then if you have to do any web browsing with sites you don't trust, do it in the VM rather than on the main system.
[21:00] <arraybolt3> (Another tip would be to install an adblocker into your web browser. Adblocker Ultimate works good.)
[21:01] <Guest5121> yeah I have an adblocker
[21:01] <Guest5121> What do U think of Ublock origin
[21:01] <Maik> Guest5121: please stick to Ubuntu support related questions
[21:01] <arraybolt3> Never used it. I've seen some stuff about it that was worrying, and some stuff about it that was good.
[21:02] <Guest5121> what was worrying about it
[21:02] <Guest5121> pycurious hi, do you know where to buy a raspberry pi by any chance ha ha
[21:02] <arraybolt3> Guest5121: Stuff about it was leaking info or something like that, check the reviews for it in Firefox. Dunno if those are accurate, though.
[21:02] <Maik> Guest5121: please take it to #ubuntu=offtopic
[21:03] <Maik>  #ubuntu-offtopic
[21:03] <arraybolt3> Maik: Most of this was Ubuntu support related I thought, except for the Raspberry Pi.
[21:03] <Maik> that was what i was aming at
[21:03] <Maik> arraybolt3: besides that they were trolling in #linux
[21:04] <Guest5121> I wasn't trolling at all
[21:04] <Guest5121> I was asking the same questions as here
[21:05] <arraybolt3> Maik: Ah. I left #linux some time ago so I didn't notice.
[21:05] <Maik> Guest5121: thus crossposting too which isn't allowed. Anyway lets move on
[21:06] <Guest5121> omg mods are the biggest barrier to learning
[21:06] <Guest5121> Asking two seperate rooms isnt spam or whatever
[21:06] <Guest5121> Its two seperate rooms
[21:06] <rbox> lol
[21:06] <bG9s> incredible, it's as if the age could be seen in text
[21:07] <arraybolt3> It's one of the rules we have for the good of the user. Sad that it seems to be offensive so often.
[21:08] <bG9s> they'll learn, and they'll come to appreciate it
[21:08] <bG9s> I like a good mod
[21:08] <arraybolt3> (TL;DR: Asking the same question in two rooms will probably get you two different sets of instructions to do the same thing. Following both at once may cause havoc. In the event someone reads the logs, here's at least part of why crossposting is a bad idea.)
[21:10] <polip> Is linux better than windows?
[21:11] <rbox> polip: no, you should stick with windows
[21:11] <arraybolt3> Depends on the usecase.
[21:12] <polip> but I formatted my Win10 hdd today, now I installed ubuntu.
[21:12] <polip> I am too layz to go back.
[21:13] <rbox> oh well
[21:13] <polip> lazy
[21:13] <rbox> you can just throw it out i guess
[21:13] <arraybolt3> !ot
[21:23] <Diagon> 22.04 / Thinkpad T530 / optimus with nvidia driver.  Any suggestsions how to get the Mini-DP video out to work?  Doesn't even appear with xrandr.
[21:38] <oerheks> Diagon, after a quick search; '  BIOS set to integrated graphics, ubuntu doesn't see an external monitor.  change to discrete graphics (NVS 5400M), it works. '
[21:38] <oerheks> The mini-DP and VGA ports are routed through the dedicated-GPU (dGPU)
[22:10] <Sabotender> having trouble installing exfat-fuse, running ubuntu 22.04
[22:12] <oerheks> exfat is in the kernel now.
[22:12] <oerheks>   exfatprogs (version 1.1.3-1) is present and installed.
[22:13] <oerheks> oh sorry, not standard
[22:15] <Diagon> oerheks / I know that.  I'm running nvidia, as `prime-select query` tells me.
[22:16] <Diagon> oerheks / still, no mDP.
[22:16] <Sabotender> just tryin whatever I can find in  google. Ijust want this to work
[22:17] <arraybolt3> Sabotender: Are you trying to make an exFAT partition?
[22:17] <arraybolt3> If so, you probably want exfatprogs, not exfat-fuse.
[22:18] <Sabotender> no trying to read from one.
[22:18] <arraybolt3> Hmm. Is the drive plugged in? This should just work.
[22:18] <rrr> Hello again, while using hostapd and dnsmasq - I managed to connect to it but without internet traffic, I also had to set a static ip. Is anyone familiar with this?
[22:18] <arraybolt3> If the drive is plugged in, the problem is probably with the filesystem or drive, not the driver, in which case we need to do some more digging.
[22:19] <Sabotender> yes its plugged ut get unknown file system type exfat
[22:19] <arraybolt3> Sabotender: What command are you trying to use to mount it? Or are you just trying to open it with the file manager?
[22:19] <arraybolt3> (Also is this a clean installation of 22.04, or did you upgrade from an earlier version of Ubuntu?)
[22:20] <oerheks> !info exfatprogs
[22:20] <oerheks> not sure you need those to rw
[22:21] <rbox> you need the mount helper
[22:21] <rbox> to mount
[22:21] <oerheks> standard if it would not mount, filesystem errors?
[22:22] <Sabotender> this is from an image of ubuntu. and I want to mount it on the desktop, I want to download a file from gitt, make and install it from there so I can have a compiled copy separate from the file system ecause it has very limited spacce
[22:23] <Sabotender> and I dont want to do this again in case there is another disaster
[22:25] <oerheks> exfat and ubuntu image, i do not get the connection
[22:25] <Sabotender> I **hate* compiling drivers when it should ust work, imo. I only want to do this once
[22:27] <snowhawk> hello!
[22:28] <snowhawk> if/when i install Ubuntu, can I wipe the partition that it makes for a different distro if I choose?
[22:29] <rbox> unhappy with the answer you've already gotten?
[22:29] <rbox> rofl
[22:29] <snowhawk> thought this was a diff room
[22:29] <rbox> so you're just goign to ask the same thing everywhere?
[22:30] <snowhawk> ya
[22:30] <arraybolt3> snowhawk: Technically yes you can probably install whatever distro you want into the existing partitions, deleting the old one in so doing.
[22:31] <arraybolt3> (Deleting the old OS but keeping the same partitions.)
[22:31] <Sabotender> I am running ubuntu on an odroid which has no ether nett connection, so I need to compile drivers for a  RTL8811CU
[22:31] <arraybolt3> Sabotender: Are you using odroid's Ubuntu image?
[22:31] <arraybolt3> If so, that may explain the problem. odroid may have applied customizations to their version of Ubuntu that we don't know about and aren't able to support.
[22:32] <Sabotender> I am currently tethering my iphone for internet access, so I can install the required dependences and utilities
[22:32] <arraybolt3> (The Raspberry Pi is supported here since Ubuntu makes their own Raspberry Pi images. But we don't make odroid images - odroid does that, I believe. Which may explain why you're running into this problem - perhaps they removed exfat support from their kernel.)
[22:33] <arraybolt3> Basically it would probably be easier to find someone who can fix the problem by looking into the support channels odroid offers. We don't support it here, not because we don't want to, but because we have no idea how to.
[22:33] <arraybolt3> Sabotender: ^
[22:33] <Sabotender> its silly because peeps use these things for NAS setups
[22:34] <Sabotender> im not asking for support, I just want to install exfat-utils
[22:35] <arraybolt3> Sabotender: Does "sudo apt install exfatprogs" work?
[22:35] <arraybolt3> Or "sudo apt install exfat-fuse"?
[22:35] <Sabotender> let e try just a moment please
[22:42] <Sabotender> I was able to install that but still unknow file type exfat
[22:43] <Sabotender> *file system
[22:44] <romanparish> you trying to mount a exfat?
[22:44] <romanparish> on a PI?
[22:44] <Sabotender> no, an Odroid
[22:45] <romanparish> ah
[22:54] <rrr> I have resolved the dns server issue I had but my hostapd wifi connection is not sharing the internet for some reason.
[22:55] <rob0> is the wireless client connecting to the AP?
[22:55] <rrr> yes
[22:56] <rrr> oh, the AP? what's that mean. I think it do.
[22:56] <rob0> are you running a DHCP server on the AP? Is the client getting an IP address?
[22:57] <rrr> yes, I am running isc-dhcp-server and also seem to have dnsmasq
[22:58] <rob0> pick only one of those
[22:58] <rob0> dnsmasq is easiest and also provides DNS
[22:58] <rrr> ok.
[23:04] <rob0> Either one (dhcpd or dnsmasq) will log when a client gets an IP address.
[23:04] <rrr> dhcpd seems to be working better, it must be configured more correctly but no internet -- I also want it to use a vpn connection if I can, as that was the goal.
[23:06] <rob0> you can give your client 8.8.8.8 or similar as nameserver; you don't have to run your own.
[23:07] <rob0> (dhcpd can be configured to assign nameserver addresses)
[23:11] <rrr> I have that configured for now, I plan to try to use my own VPN and openvpn server with it's own dns.
[23:14] <jwash> how do i start a program so that input from the f5 key is ignored? it is constantly pressed in error.
[23:16] <rbox> you can't... you can remap f5 probaly
[23:17] <rob0> rrr, then it's probably a routing issue. Where is the vpn client running?
[23:30] <snowhawk_> Hey!
[23:30] <snowhawk_> I am in my boot menu.
[23:31] <snowhawk_> OS boot Manager or USB Hard Drive?
[23:31] <snowhawk_> Which should I select?
[23:33] <rbox> try one, if it doesnt work, try the other
[23:33] <kirrim> A while ago I updated Ubuntu 22.04 LTS and now it won't boot to my DE, it gets stuck with the kernel boot log on TTY1 and I can switch to TTY2 and login. Networking is available, so I ssh'ed into that PC through my laptop so I can copy log files. Since the DE won't start and it's all TTYs, I tried running `startx` and it quit with an error, pointing at a log file. Should I upload the log file to the ubuntu pastebin for help?
[23:36] <rob0> snowhawk_, we can't know what you need. Are you wanting a USB stick to boot from?
[23:38] <snowhawk_> Needed to know what boot option to use
[23:38] <snowhawk_> It's booting now...but it had a few errors
[23:41] <kirrim> for "startx", here's the error log : https://paste.ubuntu.com/p/NfrjZHR7TD/
[23:42] <morgan-2004Dell> crash to black screen and white cursor. How can I find out what happened?
[23:44] <kirrim> morgan-2004Dell , what helps me is force-restarting the computer (try googling "reisub") and running "journalctl -b-1" to view the error log from the last session.
[23:45] <kirrim> The journalctl log contains time stamps, so search for entries slightly before the time of the crash
[23:46] <Bashing-om> morgan-2004Dell: ^ there is also X's log file: ~/.xsession-errors .
[23:49] <Sabotender> is it appropriate to as questtions about problems with making and installing drivers in ubuntu?
[23:50] <Sabotender> I think the linker is having problems
[23:51] <Sabotender> perhaps I should ask in #c?
[23:52] <Sabotender> hmm
[23:54] <kirrim> I would start with asking the developers of the drivers, what is their list of supported systems