[01:16] <Guest1970> Hello, is there a channel for people hoping to learn to contribute code?
[01:17] <Guest1970> I have programming experience, I'm more so looking for a mentor
[01:18] <junebug> Guest1970, I dm'ed you
[01:29] <lantech19446> I'm running ubuntu in a hyper-v instance as a web server and it's seeing 30% processor usage at practically no load any ideas on decreasing processor usage
[01:32] <sarnold> lantech19446: try running perf top to see what it's doing
[01:36] <lantech19446> sarnold: having an issue with it it tells me to install perf from certain packages but then can't find those packages but thanks for the advice i'll get it sorted and do that asap
[01:37] <sarnold> lantech19446: maybe an apt update && apt upgrade and reboot, to get new kernel?
[01:43] <lantech19446> sarnold: it's on 5.13 and the linux-tools were supposed to be for 5.13 but i'm running ansible right now to do all the upgrades
[01:46] <sarnold> lantech19446: the perf tools are supplied by packages related to the exact kernel package you're running, eg linux-tools-5.4.0-91-generic or whatever
[01:47] <sarnold> lantech19446: my guess is your system booted a while ago, there's been kernel updates published, the packages for the kernel you're currently running have been removed from the mirrors because they've been replaced, and now the old ones aren't available any more..
[01:47] <lantech19446> right I'm running 5.13 and the linux tools were 5.13.0-053100 or something generic but it was definitely for the right kernel
[01:47] <sarnold> that's a funny version number
[01:48] <lantech19446> I don't remember precisely what the last 4 or 5 were but it was something like that
[01:48] <sarnold> it's about three digits more than I expected :) see eg https://launchpad.net/ubuntu/+source/linux/5.13.0-22.22  -- all those are 5.13.0-22 kinds of things
[01:50] <lantech19446> sarnold: after running all the upgrades it said to just get linux-tools-generic which are actually on version like 5.4 but it's installing now
[01:51] <wesh> hello?
[01:51] <sarnold> lantech19446: hrm... yeah, linux-tools-generic *should* do it, but if your kernel comes from a different package entirely, it might not be properly matched up
[01:51] <lantech19446> and now we're back to the other error. I don't have time to reboot it right now I leave in 4 minutes this will become a job for tomorrow
[01:52] <sarnold> gn8 :)
[01:52] <lantech19446> I did find out that I wrote my updates playbook correctly though
[01:52] <sarnold> \o/
[01:52] <lantech19446> night sarnold thanks as always for all the help
[03:01] <dqtjy> I'm using Ubuntu 20.04. All of the sudden when I go to a specific website, the browser just says "unable to connect" as if I gave a bad url. `dig sitedomain.com +short` shows 0.0.0.0,  and I can access the website just fine on another device in my network. Why is the url resolving to 0.0.0.0?  There is no entry for this domain in my hosts file.
[03:01] <dqtjy> `dig sitedomain.com @8.8.8.8 +short` shows two (correct) IPs.
[03:04] <dqtjy> The output of `sudo ss -lnp '( sport = :53 )'`  (  https://termbin.com/oaph )   shows systemd-resolve, and if I look up the PID, it's "systemd+"
[03:05] <sarnold> dqtjy: I suggest putting the domain name in https://dnsviz.net/ -- it does a great job of explaining a lot of common dns issues
[03:29] <webchat31> alright this has been driving me crazy for hours, how do I decompress a PLAIN .tar (no gzip or anything) using several cores?
[03:29] <webchat31> my problem is I have around 20 gigs of data composed of lots of tiny files that come packaged in .tar, when I try to extract them on a kaggle instance they take too long and the instance expires and fails.
[03:29] <webchat31> I can extract them and recompress them using another method on a live instance, but the "saved" instances expire too fast on regular .tar. So if there's some alternative format I can use for faster decompression that'd be good too.
[03:29] <webchat31> I can't put the data on Mega and use the megacmd tool too cause of the 5GB bandwidth limit
[03:39] <sarnold> webchat31: do you have the memory to decompress them to a shmfs mount rather than disk? if not, maybe try to add an eatmydata to it, to try to disable the sync syscalls from tar?
[03:39] <sarnold> webchat31: is this one single tarball with a million little files in it? or is it a million little tarballs with a handful of files in each?
[04:01] <webchat31> sarnold: It's one single tarball with lots of folders with lots of files in them. I wish to preserve the directory structure too.
[04:02] <webchat31> And no I don't think I have enough RAM for that
[04:02] <sarnold> webchat31: dang. then try the eatmydata approach
[04:03] <webchat31> Will do right now.
[04:03] <webchat31> The stupid thing is tar is only using 1 core anyways
[04:03] <webchat31> I'm not sure if something like 7zip will decompress faster because of multicore
[04:04] <sarnold> I'd honestly expect a plain tar to go at disk speeds
[04:05] <sarnold> even with just one core, that's all it has to do: read header off file; read that amount of data and write to disk. read header, read file, write file. read header, read file, write file. etc etc
[04:12] <webchat31> I see
[04:12] <webchat31> This might confirm my other suspicion that the process is actually limited by the upload speed of the saved instance, since I think the "output" folder actually uploads to some Google Cloud thing, which would explain the error message
[04:14] <dqtjy> I am still struggling to figure out why one particular domain fails DNS
[04:14] <dqtjy> lookup on my machine. I don't know if it's because of a recent update or
[04:14] <dqtjy>  what. `systemd-resolve --status` contains this line  `DNS Domain: ~.  `  and IDK if that's right
[04:14] <webchat31> sarnold: Any idea how to upload the files quicker (without having the throughput die due to there behind millions of files) while also not having to spend 10+ minutes on a GPU instance decompressing them every time I need to use the data?
[04:14] <webchat31> I think I might just try to split the dataset up into several extracted portions
[04:14] <webchat31> It'll be a pain but that's probably my only choice at this point
[04:15] <sarnold> webchat31: ooof, if you're unpacking to a directory that's actually just writing to a far away network filesystem, or fuse mount, or something, yeah, that could be pretty painful
[04:15] <webchat31> Yup
[04:16] <sarnold> webchat31: does the google cloud thing have a tar ingest mode or something? *maybe* there's an API tool you can use to feed it a gigantic blob of data in one go, to avoid doing millions of little things
[04:17] <sarnold> webchat31: at least I think I've seen various s3 APIs accept zip files. I think. That was ages ago, and amazon, but encouraging.. :)
[04:17] <webchat31> Actually I think if I find a file upload service with a size limit big enough I can use their dataset importing thing from a remote URL... Previously the dataset was only on GDrive and Mega which was incompatible with that service
[04:18] <webchat31> I'll try zipping it and uploading it somewhere
[04:18] <sarnold> maybe stuff it right into an s3 bucket?
[04:19] <sarnold> time for me to run, have a good night, day, etc :)
[04:19] <webchat31> Thanks for all the help, you too!
[04:19] <webchat31> These guys claim a 50gb upload limit https://jirafeau.kennesaw.edu/
[04:19] <webchat31> Time to put that to the test O:3
[04:19] <sarnold> hah, my guess.. not gonna go well :)
[04:20] <sarnold> have fun :)
[04:20] <webchat31> tyvm! you too
[05:02] <jorge> Good night! I am trying to compile the latest linux kernel in Ubuntu 21.10 and I an get an error message when I try the command "make menu config".The output is:
[05:02] <jorge> HOSTLD  scripts/kconfig/mconf
[05:02] <genii-core> !mainline
[05:20] <dqtjy> I removed the symlink at /etc/resolv.conf (it was replaced by Network Manager), disabled and stopped systemd-resolved, and added `dns=default`to /etc/NetworkManager/NetworkManager.conf, and then rebooted, and I one website still resolves to 0.0.0.0.  I have no idea why
[05:40] <klausfiend> personally (and i'm not recommending this necessarily) but i loathe all of those "smart" automatic DNS configurators; we just stood up anycast and use a resolv.conf that's controlled by config management
[05:48] <dqtjy> I don't know enough about it to have an opinion, but the feedback/general sentiment for systemd-resolved has been overwhelmingly negative.  In any case, I've disabled it and I'm still experiencing a dns failure for one specific domain.
[05:49] <dqtjy> IRC has always been my last resort so I'm at a loss
[05:52] <klausfiend> the fight to manage DNS resolver behavior on Linux is an especially crowded field. you have the traditional way — add static entries to /etc/resolv.conf and call it a day — but you also have resolvconf, NetworkManager, systemd, and DHCP clients all competing to deliver an incredibly basic configurable. if you're not worried about internal domains, i would just point things at 1.1.1.1 or 4.2.2.2 or 8.8.8.8 or another similar public 
[05:52] <klausfiend> S service and call it a day.
[05:55] <dqtjy> once I disabled resolved and removed the symlink at /etc/resolv.conf, it looks like Network Manager placed one there automatically. It contains `search home` as the first line. Should I remove that?
[05:56] <dqtjy> welp, tried that and restarted network manager, nothing
[05:57] <dqtjy> nvm restarting network manager overwrote that change 🤦
[06:04] <klausfiend> yeah, welcome to wrestling all the stupid crap that modifies your resolver config
[06:06] <dqtjy> but I don't get it, there's nothing in the resolver conf, just "search home \n nameserver $ROUTERIP "
[06:06] <dqtjy> the only thing that has been succesful is using dig w/ a public nameserver to get an IP addr, and putting that into /etc/hosts manually
[06:06] <dqtjy> but that's so dumb
[06:22] <DarkTrick> **Q** @ ubuntu-bug: Is there an automated way to change the username for the files being uploaded?
[06:28] <chiluk> uploaded via what ? apport-bug? ubuntu-bug?
[06:28] <chiluk> ignore me...
[06:29] <chiluk> @DarkTrick... probably not a quick way..  You could grep/sed... and submit manually I guess
[06:30] <DarkTrick> chiluk, ubuntu-bug
[06:30] <DarkTrick> chiluk, thank you!
[06:30] <chiluk> I'm surprised your username is being captured at all
[06:30] <DarkTrick> it's inside the logs
[06:31] <chiluk> yeah... that was the only way it would be getting captured.
[06:31] <DarkTrick> (I'd say this is actually a privacy problem...)
[06:31] <chiluk> maybe... but if you are logged in to launchpad to submit the bug.... people already know it was you who submitted it.
[06:32] <DarkTrick> chiluk, sure, but only if you launchpad name is your full name
[06:32] <DarkTrick> chances of having a full-name user name on your PC is higher, I gues
[06:32] <DarkTrick> s
[06:33] <chiluk> true.
[06:51] <alkisg> dqtjy: if it's just one domain that's wrong, then it's possible that the problem is in the dns server you're using, not in your resolver. Did you try dig with the nameserver you're using?
[06:53] <dqtjy> the name server specified on my machine is the router's IP.
[06:53] <dqtjy> omg wow
[06:54] <dqtjy> OK, so on the command line  I was using 1.1.1.1, which works successfully, but I just realized the router DNS is set to 1.1.1.3/1.0.0.3, which filters for malware/pr0n etc
[06:54] <dqtjy> but I just retried using 1.1.1.3 on the command line... and no, it doesn't work.
[06:54] <dqtjy> This... makes no sense
[06:54] <dqtjy> Then why does it work for other devices?
[06:55] <dqtjy> The DNS server on both my phone and router is set to 1.1.1.3/1.0.0.3, yet my phone browses that domain just fine.
[06:56] <alkisg> It may have a fallback mechanism to search e.g. google's dns when the primary one fails, dunno
[06:56] <alkisg> You'd need to do some packet capturing to be sure
[06:56] <dqtjy> this is driving me nuts
[07:00] <dqtjy> I think you might be right alkisg
[07:01] <dqtjy> I've been testing with some other domains and I think that's happening. Somehow my phone is just using another name server if 1.1.1.3 doesn't work
[07:02] <dqtjy> I'm so confused, ugh, then why even have the ability to set a ns if it won't even stick to it
[07:02] <alkisg> That's an android issue though, not an ubuntu issue ;)
[07:03] <dqtjy> * dqtjy has joined #android
[07:03] <dqtjy> lol
[07:11] <mncheck> is there a way to get the current theme name in vanilla 20.04 from the terminal?
[07:14] <Kuleshov> HNY!!!
[07:18] <mncheck> like gsettings get org.gnome.desktop.interface gtk-theme
[08:43] <ducasse> does that not work?
[08:44] <ducasse> what about org.gnome.desktop.wm.preferences theme?
[10:08] <sdfgsdfg> hi
[10:46] <haroon> Hello I was trying to get a netplan config working, where I want to assign one ip static and the others through a match. however my approach doesn't seem to work. can someone tell me if im just doing it wrong or this is not possible? https://pastebin.com/raw/idsMPXfR
[10:54] <beuys> Hello! I just noticed that on Ubuntu, I don't have a "create from xsane" optin in Gimp to scan documents. How do I get it back?
[10:56] <beuys> Yay, "apt install xsane" brought it back!
[11:27] <beuys> Unfortunately it f*cks up the colors.
[11:30] <Maik> mind your language please beuys
[11:31] <beuys> Maik: Do you feel pain when you see the 4 chars "f*ck"?
[11:34] <Maik> just cut it will ya
[11:34] <Maik> and behave... there's nothing more to it.
[11:35] <beuys> Maik: I'm just trying to understand you. So I can take care of your feelings apropriately.
[11:40] <Maik> beuys: you're in a ubuntu support channel, there are rules and guidelines to apply to. I suggest you read them and do so as was kindly told. If you feel like trolling, please don't bother me and go elsewhere.
[11:40] <Maik> Enough said.
[12:41] <Thomas25> Can I boot ubuntu desktop via iGPU (not using NVidia) and run a Win11 KVM with NVidia passthrough and also another OS KVM with shared iGPU?
[13:02] <hans> why the fawk doesn't `add-apt-repository`  imply `apt-get update` ? so that adding multiple in a row will be faster or something? (eg if you need to add 10 in a row, doing 9 useless updates might be annoying)
[13:03] <hans> oh, it does (in 20.04 at least)
[13:03] <hans> nvm
[13:20] <BluesKaj> Hi folks
[13:24] <POQDavid> Hi i been trying to mount a nfs storage for backup and when i transfer large files it hangs and stops responding i managed to get it a bit better by lowering rsize and wsize by half the default size but still doesnt look stable how i can fix the problem with my mount
[13:25] <Thomas25> grub seems very slow, like 500ms latency
[13:26] <Thomas25> when i look at the EFI partition from other systems, it seems different. Grub files seems to be not populated. Is that normal?
[13:26] <EriC^^> Thomas25: basically just shimx64.efi and grub.cfg
[13:26] <Thomas25> when i look from other OSes, that's what i see
[13:27] <Thomas25> i.e. i wanted to edit grub config from another os
[13:28] <EriC^^> Thomas25: if you mean the menu, it's actually in /boot/grub/grub.cfg , that grub.cfg in the efi partiiton is just to point to the rootfs and set $prefix to find the menu
[13:28] <Thomas25> so the extremely minimal grub mounts the actual root and reads the config from there
[13:28] <EriC^^> Thomas25: you're not supposed to edit it by hand though as changes will go, you want to edit /etc/default/grub or /etc/grub.d/40_custom and run update-grub
[13:29] <EriC^^> yeah
[13:29] <EriC^^> it sets prefix and finds the modules and whatnot, otherwise you land in grub rescue>
[13:29] <Thomas25> i had win11 + mac installed on this, booting via clover bootloader.
[13:30] <Thomas25> after installing ubuntu that's gone. the system won't even boot unless i specificall select ubuntu efi from BIOS boot selector
[13:30] <Thomas25> then in the grub edit, i can point to the efi of either win11 directly and it boots
[13:30] <EriC^^> probably it used some efi file, is there nothing in the efi partition anymore?
[13:30] <Thomas25> or load clover, a second bootloader, and then i can boot mac
[13:31] <Thomas25> EriC^^ no they are all there
[13:31] <EriC^^> oh ok
[13:31] <EriC^^> so you want clover as the default efi loaded
[13:31] <Thomas25> everything is there, it's just bios can't pick one and it can't boot
[13:31] <Thomas25> EriC^^ actually yeah
[13:31] <EriC^^> Thomas25: sudo efibootmgr -v | nc termbin.com 9999
[13:31] <EriC^^> it should still be in the list, all you have to do is change the bootorder
[13:32] <Thomas25> https://termbin.com/fzxy
[13:33] <EriC^^> seems gone
[13:33] <Thomas25> no it's there
[13:33] <EriC^^> try "sudo ls -lR /boot/efi | nc termbin.com 9999"
[13:33] <Thomas25> i've booted it :)
[13:33] <EriC^^> i mean in the list
[13:34] <Thomas25> https://termbin.com/fpb3
[13:35] <Thomas25> so now, by default bios tries to load /EFI/BOOT/BOOTX64.efi
[13:35] <EriC^^> ok, last command "df -h /boot/efi"
[13:35] <Thomas25> it was working fine from what i've understand and it was loading clover before
[13:35] <Thomas25> but now it's confused
[13:35] <EriC^^> just get the device name e.g /dev/sda1 or whatever
[13:35] <EriC^^> Thomas25: no worries
[13:35] <Thomas25> nvme1n1p1
[13:36] <Thomas25> this is the ssd where my actual EFI lives. (ubuntu is in nvme0)
[13:36] <EriC^^> sudo efibootmgr -c -w -l \\EFI\\CLOVER\\CLOVERX64.efi -L "Clover" -p 1 -d /dev/nvme1n1
[13:37] <EriC^^> that should add it to the uefi list
[13:37] <EriC^^> try "sudo efibootmgr -v | nc termbin.com 9999" to confirm
[13:38] <Thomas25> albeit typo https://termbin.com/26mo
[13:39] <EriC^^> that typo wont fly :D
[13:39] <Thomas25> yeah i just noticed the critical typo
[13:39] <Thomas25> :)
[13:39] <EriC^^> sudo efibootmgr -B -b 0000   to delete
[13:39] <Thomas25> cannot efit/
[13:39] <Thomas25> edit?
[13:39] <EriC^^> nope
[13:40] <Thomas25> holy sh1t, that's insane, why? :)
[13:40] <EriC^^> probably the way uefi firmware thingie is
[13:40] <POQDavid> is rsize=32768,wsize=32768 too low ?
[13:41] <EriC^^> would be nice to edit though
[13:41] <Thomas25> hmm now it looks missing in the efibootmgr -v
[13:41] <Thomas25> i wanted to verify the number
[13:41] <EriC^^> great, rerun the -c command to add it back
[13:42] <Thomas25> looks good i think https://termbin.com/y0sx
[13:43] <EriC^^> yup looks good
[13:43] <Thomas25> shit didn't boot again
[13:43] <Thomas25> let's see if the entry is still there
[13:44] <Thomas25> when i just let it try to boot, it resets.
[13:44] <Thomas25> i have to get into the boot mgr of bios to boot ubuntu
[13:44] <Thomas25> clover is gone from the list again
[13:44] <Thomas25> do we have to commit the changes somehow?
[13:44] <EriC^^> aha
[13:45] <EriC^^> i think you've switched the microsoft efi files so that it boots ubuntu instead?
[13:45] <Thomas25> i was switching MS boot files because otherwise clovre wouldn't boot
[13:45] <Thomas25> that was a trick i was doing before installing ubuntu
[13:46] <EriC^^> i see
[13:46] <EriC^^> ok, seems the file is gone now
[13:46] <Thomas25> maybe while we are at it we can just try to run win11 via KVM? (looking forward to trying that) hehe
[13:46] <Thomas25> what file?
[13:46] <EriC^^> try 'sudo cp /boot/efi/efi/clover/cloverx64.efi /boot/efi/efi/microsoft/boot/bootmgfw.efi'
[13:46] <EriC^^> bootmgfw.efi , it's gon
[13:46] <EriC^^> *gone
[13:46] <Thomas25> i `mv` it to bootmg something .original
[13:46] <Thomas25> i can boot it manually via grub, i tried it
[13:47] <EriC^^> yeah
[13:47] <Thomas25> but note that efibootmgr does not list clover again!
[13:47] <EriC^^> is this an hp laptop?
[13:47] <Thomas25> is that normal?
[13:47] <Thomas25> nope, it's a desktop
[13:47] <EriC^^> ok
[13:47] <Thomas25> but i must note that i cannot enter the BIOS screen right now, don't know why. it seems frozen just a black screen
[13:47] <EriC^^> why not replace bootmgfw.efi with the clover one so it boots
[13:47] <Thomas25> luckily i can get into boot manager via F8
[13:48] <Thomas25> we could but isn't it kinda weird? i don't know how EFI firmware runs these
[13:48] <Thomas25> but moving it to another folder and all of its config in another folder
[13:48] <EriC^^> you were doing it before though no?
[13:48] <Thomas25> nope
[13:48] <Thomas25> i was just making windows not bootable
[13:48] <Thomas25> so bios would load clover
[13:49] <Thomas25> and from there i can chose win or mac
[13:49] <EriC^^> oh
[13:49] <EriC^^> and how does grub find windows
[13:49] <EriC^^> you added a custom entry?
[13:49] <Thomas25> nope
[13:49] <Thomas25> it found and added it when installing i think
[13:49] <Thomas25> not sure
[13:49] <EriC^^> cause its used to looking for bootmgfw.efi
[13:49] <Thomas25> yeah probably
[13:50] <EriC^^> what i'd do is replace stuff as they were
[13:50] <Thomas25> so what do you think about clover efi entry is going missing?
[13:50] <Thomas25> as they were? i.e. remove grub from efi?
[13:50] <EriC^^> on my laptop, it also wont boot ubuntu, before i used to switch the files bootmgfw.efi and change grub to look for a different filename
[13:50] <EriC^^> but later i found if you disable the entry via bootmgfw.efi it'll boot ubuntu no problem
[13:51] <Thomas25> via efibootmgr you mean?
[13:51] <EriC^^> yes, sorry
[13:51] <EriC^^> this is what mine looks like, https://termbin.com/dcsf
[13:51] <EriC^^> if there's no * next to Bootxxxx it means its disabled
[13:51] <EriC^^> works like a charm on this laptop, might work for you too
[13:51] <Thomas25> ok
[13:52] <EriC^^> if i ever need to remove ubuntu, just re-enable the entry, and everything else is stock
[13:52] <Thomas25> yeah i could use that for removing windows update problem
[13:52] <Thomas25> but i don't know why it didn't boot when we made that change
[13:52] <Thomas25> because it looked lefit
[13:52] <Thomas25> legit*
[13:52] <EriC^^> yeah kinda odd it removed the entry, maybe you have to 'trust' the file manually from the bios or something?
[13:53] <Thomas25> so efibootmgr communicates with BIOS to tell commands to EFI for boot entries? it is not on hard drives whatsoever right?
[13:53] <hans> why doesn't p7zip-full include p7zip-rar ?
[13:53] <EriC^^> acer's have that, have to set a bios pass, then you can trust efi files
[13:53] <EriC^^> Thomas25: exactly, the uefi list is saved on the nvram in the motherboard
[13:53] <Thomas25> i doubt because this was an efi that i was using as my primary boot from that very same location
[13:54] <Thomas25> feels like there's something wrong with my bios then
[13:54] <Thomas25> entries going missing
[13:54] <Thomas25> + i cannot enter bios screen
[13:54] <EriC^^> yeah i dont get why it removed it, and ubuntu is there
[13:54] <EriC^^> i'd switch the windows file with clover one then
[13:55] <EriC^^> and you can edit grub to look for the new windows file that you name, it wont matter for anything else itll work as usual
[13:56] <EriC^^> Thomas25: 'sudo cp /boot/efi/efi/clover/cloverx64.efi /boot/efi/efi/microsoft/boot/bootmgfw.efi'
[13:56] <Thomas25> i believe that'd work
[13:56] <EriC^^> yeah
[13:56] <Thomas25> but not pretty!
[13:56] <EriC^^> gotta do what you gotta do :D
[13:57] <Thomas25> haha
[13:58] <Thomas25> it did work
[13:58] <Thomas25> i'm surprised clover didn't find ubuntu automatically though
[13:58] <Thomas25> not a big problem with the primer you've given i think i'll figure the exact setup that i want
[13:58] <EriC^^> for grub, either add a custom entry to /etc/grub.d for bootmgfw-orig.efi or if you want you could edit '/usr/lib/os-probes/mounted/efi/20microsoft' and replace bootmgfw.efi to bootmgfw-orig.efi in the middle
[14:00] <EriC^^> alright, sounds good!
[14:00] <Thomas25> god dammit, when i wanted to upgrade bios firmware, win11 bsodded me
[14:00] <EriC^^> yikes
[14:31] <hans> how do you enable pam_limits on 20.04? this thing, http://manpages.ubuntu.com/manpages/bionic/man8/pam_limits.8.html
[14:33] <hans>  libpam-modules is installed, but it doesn't seem to do anything, the directive "@hans   hard nice     -20" appears to be ignored in /etc/security/limits.conf
[14:36] <hans> and `cat /proc/modules | grep -i limit` doesn't list pam_limits.. just  a bunch of netfilter limits
[14:37] <LarsErikP> Hi! Question about the netboot-kernel of the focal legacy-installer.. The kernel was last updated 01.02.21; will that ever be updated again? I would really like a kernel that includes this: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1893956
[14:41] <hans> LarsErikP: hmm that's a though one, i remember having a similar issue with Debian netinst, and someone on ##debian-next showed me an easy-ish way to build the netinst from source with newest kernel, unfortunately i don't remember how it was done, but it actually wasn't difficult with the guidance of someone who knew how to do it..
[14:43] <hans> LarsErikP: anyway, last time i faced an issue like that, i ended up building the netinst from source (but it was debian, not ubuntu)
[14:43] <LarsErikP> hans: would have been really helpful :P
[14:44] <hans> LarsErikP: have you tried booting the ubuntu installer from netboot.xyz ? it might boot a newer kernel than the legacy isos, https://netboot.xyz/
[14:45] <LarsErikP> oh
[14:45] <hans> (it doesn't use the isos at all)
[14:45] <LarsErikP> worth a try!
[14:45] <hans> and it ask you if you want the legacy installer or the .. modern one, at least for 20.04
[14:46] <LarsErikP> hans: do they publish they actual kernel files?
[14:46] <hans> netboot.xyz ? well it list the url to the kernel it loads, right before loading it
[14:51] <hans> LarsErikP: btw netboot.xyz currently has a bug with DHCP, if your network doesn't auto-configure DHCP, you'll have to use a bunch of kernel parameters to boot the ubuntu installer (but almost all wifi's have DHCP enabled :P )
[14:52] <LarsErikP> hmm.. i kind of just wanted to grap the kernel and initrd.gz for focal..
[14:52] <LarsErikP> not actually pxe-booting directly from them :P
[14:54] <hans> maybe http://archive.ubuntu.com/ubuntu/dists/focal/main/installer-amd64/current/legacy-images/hd-media/
[14:54] <LarsErikP> too old :(
[14:54] <hans> doh kk
[14:57] <hans> LarsErikP: if you wanna try the build-it-yourself-route, this might be helpful, https://wiki.debian.org/DebianInstaller/Build
[15:00] <LarsErikP> thanks!
[15:17] <stealthrecon> I have setup Maas and I am wanting to write a commissioning script that will add a list of additional applications I need installed.  Does anyone have an example script that does something similar?
[15:17] <leftyfb> stealthrecon: try #maas
[15:18] <stealthrecon> Thanks.
[15:42] <yates> running lubuntu 18.04, after changing the default text application using Preferences/Default Applications of LXSession i'm getting "KDEInit could not launch 'konsole'" (even after rebooting)
[15:47] <yates> anyone have any suggestions?
[15:48] <yates> i also see a message box that popped up titled "Update lxsession database" which states "The database is updating, please wait" but it never completes.
[15:56] <krisstern> yates: Maybe try changing the hostname?
[15:57] <yates> krisstern: huh?
[15:57] <krisstern> Like this: https://forum.kde.org/viewtopic.php?f=66&t=132466
[15:58] <krisstern> For the "KDEInit could not launch" error...
[15:58] <yates> ok, thanks - i'll try
[15:59] <krisstern> Cool :-)
[16:13] <Guest816> Why so much hate for Ubuntu?
[16:23] <Kel> I feel like every distro gets a bit of hate from fanboys of other distros, hehe.
[16:23] <Kel> I use Ubuntu, Debian, and Manjaro for different things, depending on my needs *shrugs*
[16:23] <Kel> They all have a purpose in my view.
[16:23] <Maik> offtopic chatter imho
[16:24] <Kel> oh he left anyway
[16:24] <Maik> yep
[16:24] <Kel> I have joins/parts/quits filtered, hehe - oh well.
[16:25] <llanhmod> I dont need to restart after enabling a ua subscription, do I ?
[16:36] <llanhmod> well its already enabled so perhaps not.
[16:55] <MarcUK> Hi all i got a radio controller it needs a program called Eepe i found the 64 bit deb package but its complaining about dependencies is there a way to install it and get it to do the dep automatically
[16:56] <MarcUK> radio controller as in turnigy 9xr for RC aircraft
[16:57] <Syntax-> Try gdebi that usually gets it right
[16:57] <rfm> MarcUK, "apt install package.deb" will chase the dependencies
[16:58] <MarcUK> didnt work
[16:58] <MarcUK> sudo apt install ./eepe-amd64.deb
[16:58] <Syntax-> whats the output?
[16:58] <MarcUK> epe : Depends: libqt4-gui but it is not installable
[16:59] <MarcUK> Depends: libqt4-network but it is not installable
[16:59] <MarcUK> Depends: libqt4-core but it is not installable
[16:59] <MarcUK> Depends: libqt4-xml but it is not installable
[16:59] <Syntax-> try "sudo apt install libqt4*"
[16:59] <Syntax-> and then try installing the package again
[17:00] <MarcUK> nope complained about dependancies
[17:00] <Maik> the question is, is libqt4 still supported....
[17:01] <MarcUK> You might want to run 'apt --fix-broken install' to correct these.
[17:01] <MarcUK> tried that too didnt work
[17:01] <Syntax-> you could try to unzip the .deb file and unzip data i think and run the binary file directly
[17:02] <rfm> It's probably built for a different release of Ubuntu -- libqt4 doesn't seem to be in anything since bionic
[17:02] <Maik> that^
[17:03] <Maik> even a apt search on 21.10 doesn't show it
[17:03] <Maik> MarcUK: ^^^
[17:04] <MarcUK> yep but its needed for upgrading the firmware on this rc controller
[17:04] <Maik> too bad then i guess since libqt4 isn't supported anymore
[17:06] <MarcUK> terrible that you have to go find a windows machine just to upgrade stuff lolz
[17:06] <MarcUK> i unpacked it it didnt run
[17:07] <Syntax-> Package: eepe
[17:07] <Syntax-> Priority: extra
[17:07] <Syntax-> Section: checkinstall
[17:07] <Syntax-> Installed-Size: 1384
[17:07] <Syntax-> Maintainer: erazz
[17:07] <Syntax-> Architecture: amd64
[17:07] <Syntax-> Version: 362-1
[17:07] <MarcUK> maybe i can donload windows version and run it on wine
[17:07] <Exterminador> hello guys! sorry for the stupid answer. on the following command, which is the package that install the IDE? `apt install openssl sqlite3 libwxgtk3.0-gtk3-0v5 lame libtag1v5 libtagc0 libmp3lame0 libmp3lame-dev libogg-dev libvorbis-dev libsndfile1 libsndfile1-dev libavcodec-extra libavformat58 libcurl4-openssl-dev libmpg123-0 libresample1 libncurses5 libphysfs1 libpcre3 libprotobuf23 libmysqlclient21 libfaac0 libopus0 libloudmouth1-0
[17:07] <Exterminador> libdbus-glib-1-2 libmuparser2v5 libsoxr0 build-essential`
[17:08] <hans> MarcUK: i have a Windows XP VM for use with old shitty Dell DRAC/iDRAC systems that only run shitty java web applets/flash clients
[17:08] <hans> MarcUK: and yeah it is shitty
[17:09] <Syntax-> Depends: libqt4-gui ,libqt4-network ,libqt4-core ,libqt4-xml ,avrdude
[17:09] <Syntax-> apparently these are the dependencies you need for the package
[17:09] <Syntax-> if you can find maybe .deb packages for those or their source code?
[17:10] <MarcUK> avrdude is installed becasue i just used it to upgrade firmware on two programmers
[17:10] <Syntax-> ah
[17:10] <Syntax-> https://askubuntu.com/questions/766615/how-to-install-libqt4-core-and-libqt4-gui-on-ubuntu-16-04-lts
[17:10] <Syntax-> here might be something usefull
[17:11] <hans> MarcUK: your best bet is probably to run it on a Ubuntu 18.04 VM (maybe try virtualbox?)
[17:11] <hans> oh.. 14.04 then.. dang
[17:12] <MarcUK> ok i will look into it all im trying to do is upgrade the firmware on this flight controller to the last firmware that was made for it as its now discontinued
[17:13] <hans> is the 14.04 repositories still up?
[17:41] <Walex> hans: there is an archive of old repos
[17:41] <Walex> hans: http://archive.ubuntu.com/ubuntu/dists/
[18:46] <morantron> hi! i'm using an nvme drive and trying to install grub but i'm seeing just /dev/nvme1 and /dev/nvme1n1 (no /dev/nvmen1pX devices), did i do something wrong when partioning the disk?
[18:55] <rollappuser> sup
[18:55] <leftyfb> morantron: "lsblk" should show you your drive and partitions
[18:56] <lotuspsychje> welcome rollappuser
[19:08] <tomreyn> morantron: did you find out, yet? "no /dev/nvmen1pX devices" suggests that there are no partitions on this nvme storage.
[19:08] <tomreyn> morantron: is this on an already installed ubuntu system? which ubuntu release, which kernel?
[19:09] <tomreyn> are you booting in legacy bios or uefi mode, is it an MBR or GPT partition table on the nvme?
[19:39] <Sven_vB> hi! where do I post feature requests for fbpanel? I can't find a bug tracker in the man page
[19:40] <tomreyn> apt should know the developers' website.
[19:40] <tomreyn> that's if it is an ubuntu package
[19:41] <genii> You can also do ubuntu-bug fbpanel but label the bug Wishlist or Feature Request
[19:42] <tomreyn> apt-cache show fbpanel | grep ^Homepage:
[19:46] <Sven_vB> thanks!
[19:49] <EldonMcGuinness> Any libvert users around that wouldn't mind helping me wrap my head around a few things?
[19:49] <kenwoodfox> Heyo! Im having some trouble packaging a snap im trying to make https://github.com/FRC-1721/signstatus/tree/frame-snap
[19:49] <kenwoodfox> it does build but it keeps saying its using python2, witch i think is causing my issue where it cant run the app? Im using python 3.9 on the host, im doing all my testing with 3.9, idk why its using 2.7 in the snap :c https://cdn.discordapp.com/attachments/506631005189177355/920401970806800425/unknown.png
[19:49] <EldonMcGuinness> Having the darndest time understanding backing disks, snapshots, and how they play together.
[19:52] <sarnold> kenwoodfox: this is a wild guess, but try replacing 'python' in the command: with 'python3'
[19:52] <kenwoodfox> sarnold: couldent hurt, trying that now
[19:53] <kenwoodfox> sarnold: That did it! now i got a new error but yayyy https://cdn.discordapp.com/attachments/506631005189177355/920402975707496508/unknown.png
[19:53] <sarnold> kenwoodfox: woot!
[19:54] <sarnold> $ apt-file search libGL.so.1
[19:54] <sarnold> libgl1: /usr/lib/x86_64-linux-gnu/libGL.so.1
[19:54] <sarnold> kenwoodfox: try adding libgl1 to the stage-packages: stanza?
[19:54] <kenwoodfox> sure, it looks like it warns me im missing a lot of things in my stage-packages
[19:54] <kenwoodfox> ill add that
[19:54] <EldonMcGuinness> I tried to setup a vm, using the 20.04 Ubuntu server image as a backing file and can get a server up. I can take snapshots and revert between them, but when I try to follow a blockcommit command that *should* allow me to squash snapshots it gives me an error saying the snapshots are not in the chain.
[19:59] <kenwoodfox> Well thats a weird errror https://cdn.discordapp.com/attachments/506631005189177355/920404709184012398/unknown.png
[19:59] <matsaman> EldonMcGuinness: might want to talk to #kvm or #qemu
[20:00] <kenwoodfox> i think i gotta squash out all the libs that snapcraft says may be missing
[20:00] <matsaman> snap =P
[20:00] <EldonMcGuinness> Cheers @matsaman
[20:00] <kenwoodfox> Im not enjoying snap so far :/ sry
[20:00] <kenwoodfox> i just wanna use my ubuntu frameeeeee
[20:01] <sarnold> kenwoodfox: oh yikes :/ that one's beyond my guesses, try #snapcraft ? I think that's the place..
[20:02] <kenwoodfox> it does work if i do it manually https://cdn.discordapp.com/attachments/506631005189177355/920405270205710356/unknown.png
[20:02] <kenwoodfox> oh cool ill hop over there, thanks sarnold
[20:02] <sarnold> oh that feels so close..
[20:02] <kenwoodfox> I knoowwwww
[20:02] <kenwoodfox> aahh
[20:02] <kenwoodfox> Its fun tho
[20:24] <EldonMcGuinness> Welp guess I'll just leave that up there and maybe post it in the forums too. Would love to be able to consolidate some of these lab machines as VMs heh.
[20:55] <jhutchins> EldonMcGuinness: What are you using to manage the VMs?
[20:57] <EldonMcGuinness> just going cli with virsh
[20:57] <EldonMcGuinness> this is a headless server
[21:00] <EldonMcGuinness> Well, I figured out how to just create a fullbackup of a VM, that will do for now, but would rather get a better handle on the snapshot bit as I'm sure it is something stupid I'm not getting heh
[21:00] <EldonMcGuinness> fullback while live that is
[21:03] <EldonMcGuinness> What I have gotten working is, instead of using a disk with the ubuntu cloudimg as a backer, just copy and resize the ubuntu cloudimg and use it directly.
[21:04] <EldonMcGuinness> This results in the blockcommit working the way I would think it would, however, can't seem to get how to squash snapshots together. Perhaps I'm just misunderstanding the idea and what it can do
[21:05] <jhutchins> EldonMcGuinness: As I understand snapshots, they are cumulative, so merging them really isn't a valid concept.
[21:07] <EldonMcGuinness> I was thinking of them like git commits, one could squash them into a single commit, as long as they are temporally ordered.
[21:07] <EldonMcGuinness> ok cool, thanks for the info.
[21:08] <EldonMcGuinness> now just have to figure out if I can use a single backing file for many vms without commits changing the basefile
[21:08] <EldonMcGuinness> At least I got the back working! *small victories*
[21:10] <sarnold> I think qcow2 allows you to have multiple layers of backing files in case you really want to get creative
[21:16] <EldonMcGuinness> @sarnold: yep it sure does. What I was hoping I could do would be to make successive snapshots as I'm making changes and if I like the end result squash all of said changes into a single snapshot.
[21:17] <EldonMcGuinness> I mean there's nothing mission critical I just hated seeing a bunch of snapshots, though I guess I could just delete the previous ones.
[21:18] <sarnold> EldonMcGuinness: yeah afaik you can have multiple snapshots, but squashing them together wouldn't necessarily make sense
[21:18] <EldonMcGuinness> Yeah I think that's the issue is just my brain isn't quite grabbing it yet
[21:18] <EldonMcGuinness> Again I think I'm comparing it to commits on GitHub when this is a different beast altogether
[21:20] <EldonMcGuinness> Question, let's assume I make four changes and four subsequent snapshots after each change. If I go back and delete the second snapshot, what I still be able to revert to the first or the third?
[21:24] <sarnold> EldonMcGuinness: that's a good question. I sure hope so, otherwise it wouldn't feel super useful to me.
[21:27] <rya> Is anyone aware of an apt repo, or upgrade guide to get Pulseaudio 15 on Ubuntu 20.04?
[21:30] <leftyfb> rya: PA15 was released after the latest release of Ubuntu. You'll more than likely have to wait till 22.04
[21:31] <sarnold> it looks like impish does have pulseaudio 15 https://launchpad.net/ubuntu/+source/pulseaudio
[21:31] <genii> They may backport it at some point
[21:31] <sarnold> I'm not actually sure if you can get there from 20.04 though, since 20.10's been retired
[21:35] <rya> Thanks -- if I manage to get it working (instead of the more likely scenario of having broken audio for some period of time), I'll report back on it.
[21:35] <mwhudson> you definitely _can_ update from 20.04 to 21.10 directly (*cough*) but it's not supported in the same way
[21:38] <sarnold> mwhudson: oh? I figured do-release-upgrade on its own would fail getting to 20.10, and do-release-upgrade -d would jump to jammy directly (probably a bad idea right now)
[21:38] <mwhudson> sarnold: hmm i managed to do it somehow or other
[21:40] <tomreyn> quoting /etc/update-manager/release-upgrades: #  normal - Check to see if a new release is available.  If more than one new release is found, the release upgrader will attempt to upgrade to the supported release that immediately succeeds the currently-running release
[21:43] <EldonMcGuinness> @sarnold: after I'm finished cooking dinner I'm going to spin up a machine and see what happens when I try to do that, I'll ping you if you want when I know for sure what happens
[21:44] <tomreyn> so next *supported* release that is available after 20.04 (if Prompt=normal) would be 21.04 / hirsute. then you'd need to do to impish (and then, optionally, to jammy(
[21:45] <tomreyn> i'm not sure what Prompt=lts with -d do, though
[21:48] <sarnold> EldonMcGuinness: cool cool, please do, if you remember, still want to, etc :)
[22:08] <sinhue> Hello. I want to upgrade ubuntu 14 to ubuntu 16 and still keep PHP version 5.6. How would I go about this? It always upgrades my php to 7
[22:09] <jeremy31> sinhue: Ubuntu 14 and 16 are both EOL/unsupported except for ESM
[22:10] <sinhue> jeremy31, I know this.
[22:10] <sinhue> If it were any other version, is this possible? to somehow exclude some package from being updated?
[22:11] <sarnold> the upgrade will fail
[22:11] <jeremy31> sinhue: apt mark hold PACKAGE
[22:12] <sarnold> I'm not sure if you're better off running an ancient unsupported system as a whole, because that's nice and obvious to everyone; or if you'd be better served to upgrade the system, uninstall php, build the old one from sources yourself, and run that -- that way at least it confines the "old and busted" to php..
[22:12] <jeremy31> I am sure Synaptic Package Manager can force a version of a package, but it will cause issues if something else requires the newer version
[22:12] <sarnold> .. but it won't be obvious to everyone at your business that you're running unsupported software, so that's not fantastic
[22:12] <rya> sinhue, can you run php from a container?
[22:14] <EldonMcGuinness> sarnold: okay just tried it out, it looks like each snapshot is relative to the current base, even though if you do list the snapshots there is a switch called tree which shows them as being children of each other, so I'm not sure what the deal is with that.
[22:14] <EldonMcGuinness> Just for clarification I created four snapshots each time going in and creating a file. Deleted the third snapshot reverted to the fourth snapshot and it still contained the contents that would have been in the third snapshot.
[22:15] <EldonMcGuinness> So it looks like these are not incremental snapshots in the normal sense, instead they are incremental from a common base image
[22:16] <sarnold> EldonMcGuinness: cool, thanks :)
[22:50] <Guest83> hello
[22:51] <Guest83> what means the error in nmcli: Error: Failed to scan hidden SSID: Scanning not allowed immediately following previous scan.
[22:52] <Guest83> the wifi network is visible
[22:55] <Guest83> i don't understand it. in windows the wifi network works problem-free
[22:57] <Jeremy31> Guest83: What wifi device?
[22:59] <mithrin> Is it possible to perform upgrade of ubuntu from one lts version to another with a package on hold?
[23:01] <Guest83> no fresh installation, but wireless packages was updated after installation
[23:01] <Guest83> its a Qualcomm Atheros AR242x / AR542x Wireless Network Adapter
[23:02] <Guest83> Fujitsi Lifebook e8310
[23:04] <Jeremy31> Guest83: That is an older wifi device, any chance the wifi router is using TKIP encryption?
[23:04] <Guest83> TKIP encryption?
[23:07] <Jeremy31> Guest83: in terminal check>  iwlist scan | egrep -i 'ssid|cipher' | nc termbin.com 9999 #post URL from terminal
[23:08] <Guest83> its not listed
[23:09] <Jeremy31> Guest83: you are using a hidden wifi access point?
[23:10] <Guest83> only found CCMP networks
[23:10] <Guest83> Jeremy31no
[23:11] <Jeremy31> Guest83: is your wifi set to auto channel?  You might need to put it on a channel between 1-11 as Atheros uses a code on the chip to determine what channels are allowed in Linux
[23:11] <Guest83> Jeremy31: no
[23:11] <lupulo> Guest83, ip link
[23:11] <lupulo> for the device
[23:14] <Guest83> actual control channel 60 in the wifi AP(in router)
[23:15] <Jeremy31> Guest83: what about the 2.4 GHz channel?  I am not sure if that old card supported 5GHz channels
[23:16] <Guest83> upps, yes i have seen at the 5G network xD
[23:16] <Guest83> control channel 13
[23:18] <Guest83> wait, i check Windows
[23:29] <Guest83> okay, why that works under windows problem-free?
[23:31] <Jeremy31> Guest83: Use Windows to change to 2.4 GHz channel to 11
[23:31] <Guest83> AR242x / AR542x Wireless Network Adapter, driver=ath5k driverversion=5.11.0-27-generic
[23:32] <Jeremy31> Guest83: change the wifi access point to use channel 11 rather than 13 or auto
[23:34] <tomreyn> 5.11.0-27-generic sounds like out of date, if this is ubuntu
[23:34] <Guest83> it is setted to auto
[23:34] <Jeremy31> Guest83: change it to use channel 11
[23:34] <Guest83> yes, fresh installed
[23:37] <tomreyn> core2du, ddr2, 2gb installed, says the vendor website
[23:38] <Guest83> only 36, 40, 44, 48, 52, 56, 60, 64, 100, 104, 108, 112, 116, 132, 136, 140 or Auto is choosable in Router
[23:39] <Jeremy31> Guest83: go into 2.4 GHz settings
[23:40] <Guest83> oh, there is 1-13 choosable
[23:40] <Jeremy31> choose 11 or lower, Guest83
[23:44] <Guest83> many thanks :D
[23:45] <Guest83> but why 12 and 13 isnt supported in driver?
[23:49] <Guest83> with linux-backports-modules-karmic should it work, but i think that is outdated?
[23:51] <tomreyn> !karmic
[23:52] <Jeremy31> Guest83: I don't think karma is still supported
[23:52] <tomreyn> looks like, slightly
[23:52] <Jeremy31> Guest83: I know how to change the channel allocations in the ath module, but it isn't worth the effort
[23:53] <Guest83> why has the driver support deteriorated so much?
[23:54] <Guest83> Jeremy31 how does it work? i like to hack xD
[23:56] <Jeremy31> Guest83: what country are you in?
[23:56] <Guest83> germany
[23:56] <Jeremy31> Guest83: in terminal, try> sudo iw reg set DE
[23:57] <Jeremy31> you might have to reboot in order to try channel 13
[23:58] <Guest83> ehh, why are the channel allocations region locked? o.O
[23:58] <Jeremy31> Guest83: post URL from terminal for> dmesg | grep ath | nc termbin.com 9999
[23:59] <Guest83> https://termbin.com/z9jg