[00:02] <|Anthony|> anyone have an idea when we can see this make its way to ubuntu? [00:02] <|Anthony|> https://cgit.freedesktop.org/drm/drm/?h=drm-fixes [00:02] <|Anthony|> being able to suspend/hibernate is kinda important lel [00:03] |Anthony|: what's the launchpad bug? [00:04] <|Anthony|> https://bugs.freedesktop.org/show_bug.cgi?id=110258 [00:04] Freedesktop bug 110258 in DRM/AMDgpu "Lenovo V110-15AST AMD A9-9410 AMD R5 Stoney hangs after waking after suspend. 5.0 onwards" [Major,New] [00:04] <|Anthony|> i'm not sure if it's in launchpad === Guest52205 is now known as argusbr === lotus|i5 is now known as lotuspsychje [00:08] <|Anthony|> should i just be looking into how to apply the patch myself? i'm so rusty i don't even know where to begin with that [00:09] |Anthony|: first, file a bug; if there's something to be fixed, it'd be nice to know about it.. [00:15] <|Anthony|> sarnold alright i'll file a bug report and reference the freedesktop.org bug and patch... for the right now issue, can you point me in the right direction for how to apply the patch myself? it would be nice to verify that the patch i'm about to reference in launchpad actually fixes the issue [00:16] |Anthony|: can you see if this affects you? https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1842785 [00:16] Launchpad bug 1842785 in linux (Ubuntu) "Freeze after waking up from suspend on a Ryzen 3000 based system" [Undecided,Confirmed] [00:17] |Anthony|: you can use apt source to download the source code of a package; and apt install build-essential ; apt install build-dep to install all the build dependencies of the package [00:17] <|Anthony|> lotuspsychje i'm reading it [00:17] <|Anthony|> sarnold thank you [00:18] |Anthony|: hopefully it's a quilt-managed source package.. in which case, add the patch to debian/patches/ and add the patch name to debian/patches/series -- add a new debian/changelog entry -- and then i *think* fakeroot ./debian/rules build -- (I only ever use built schroots with sbuild, but setting that up is a real pain) [00:22] <|Anthony|> sarnold thank you very much for the help. sounds like i should set aside a chunk of time to focus on this patch and bug report [00:23] |Anthony|: yeah, one's first modifications with debian packages can be a bit much :) it gets better.. [00:23] <|Anthony|> it's not my first rodeo... but it's been so friggin long that it might as well be [00:24] <|Anthony|> but to be fair, it was never a daily operation for me... patching kernel files [00:24] <|Anthony|> using git, yes, but not working on linux... just in linux [00:30] <|Anthony|> lotuspsychje i read the bug report you linked. it doesn't have a kern log attached which would tell me if it's the same issue. [00:33] <|Anthony|> lotuspsychje https://paste.ubuntu.com/p/7QZGPsPKnk kern log shows you the suspend process in good enough detail to see where it's failing [00:42] <|Anthony|> lotuspsychje i did comment on that BR though [00:43] |Anthony|: is your system similar, ryzen? === Sven_vB_ is now known as Sven_vB [00:52] So this ryzen issue does not apply to the 5.x kernel? [00:52] ohwait. [00:53] flog: the way around, alot of issues on 5 and higher [00:53] But i dont have any issues on my ryzen 3700u with vega on 5.x [00:53] flog: 4.19 kernel seems to have good influences on ryzen [00:53] flog: wich kernel are you? [00:54] 5.0.0.25-generic [00:54] -27 is out flog [00:54] I dont have the bandwith at work to download new kernels. [00:55] But maybe i never suspend my machine, just hibernating it. [01:03] flog: you might try to test hibernate, see if you can reproduce this [01:06] It resumes just fine from a shut lid. [01:06] But im no tsure if it is suspending or hibernating === zbenjamin is now known as Guest71466 === zbenjamin_ is now known as zbenjamin [01:08] Ubuntu 5.10 breezy badger [01:08] What log do i check to see what the kernel did during suspend? [01:09] flog: try a journalctl -f while you play around [01:11] lotuspsychje: it seems to resume just fin [01:11] e [01:12] flog: ok tnx for testing this, it might be a combo of the hardware too [01:12] it seems that apparmor is not allowing unbound to write to any other log file other than /var/log/syslog. Is this the intended behavior? [01:12] flog: maybe your system doesnt result into acpi errors [01:12] lotuspsychje: it is a brand new lenovo t495, ryzen 3700u vega10, 8GB soldered ram. [01:13] Dont know about the bios version or anything. the kernel version is as mentioned earlier 5.0.0.25 [01:13] flog: yeah the lenovo's are mostly problematic, but also due to bios firmware, updating fixes things [01:14] flog: you can check bios version in your dmesg, or inxi -F [01:14] Does the ubuntu version have anything to do with this? Using 19.04 [01:15] flog: well in most cases LTS is usually more stable, but depending on the problem it can be kernel specific related [01:17] lotuspsychje:trying to find it. [01:19] lotuspsychje: R12ET34W, released 2019-04-19. [01:19] Revision 1.4 [01:20] flog: compare to lenovo's website, or try gnome-software and fwupdate -l to see whats latest [01:21] So when I had Xubuntu 18.10 installed on my machine and ran `ubuntu-drivers autoinstall` or whatever it installed nvidia (for one graphics card PCs) instead of the bumblebee (for two graphics card PCs) as it says in the Debian wiki. So should I've uninstalled the nvidia driver and installed bumblebee? [01:21] But I dont seem to have any issues with issuing systemctl suspend. [01:22] oft_gegong: bumblebee is not adviced anymore, install the nvidia driver instead, so you can use nvidia-prime for optimus cards [01:22] flog: yeah, your hardware might not result into acpi issues, so case closed for you :p [01:22] :o [01:23] lotuspsychje: my settings, version, bios etc might be of interrest to someone else though. [01:23] flog: well if its working, its not reccomended to include into the bug [01:24] I would have thoguht that comparing to a working system with similar hardware is of interrest. [01:25] flog: in #ubuntu-discuss its surely interesting yeah [01:25] Alright. [01:39] hello, how can I install docker on ubuntu via apt install? [01:39] docker seems to be a system tray? [01:40] where is the docker container [01:40] !info docker.io [01:41] docker.io (source: docker.io): Linux container runtime. In component universe, is optional. Version 18.09.7-0ubuntu1~18.04.4 (bionic), package size 21179 kB, installed size 98282 kB [01:45] websearch turns up few and differing accounts of how well nvidia-prime works in 19.04. my use case is a dell 9560. any insight as to how much i'd need to tweak things in order to be able to use switchable gpu for e.g. cuda? === Slumlord_ is now known as Slumlord === Slumlord is now known as Slumlord_ === nt0_ is now known as nt0 [02:06] so my graphics card doesn't have optimus. so that's really confusing [02:06] yet I have two graphics cards [02:07] now I have to re-wonder if bumblebee is the right fit [02:07] yet bumblebee is support optimus. now I'm definitely confused [02:09] I only 3% care about this though [02:10] Does anybody know what this means when executing lsof :"can't identify protocol"? [02:10] Milijus: uuuuuuh. I'm 3% of sure that executing lsof with root privileges will clear those errors [02:11] but running lsof sounds scary to be honest [02:12] Milijus: very curious; guessing, you've got a socket that isn't tcp, udp, sctp... === sharpie_2 is now known as sharpie === zenguy is now known as coffeeguy [04:16] there's a shell script already on my EFI system partition that knows how to unlock my LUKS LVM. can I somehow make my initramfs run that instead of flooding errors and then dropping to an initramfs shell? most curiously, in that shell the ESP is not mounted (yet?). [04:30] looks like I want an initramfs-hook [04:46] can I use bashisms in /etc/default/grub? [04:48] Good evening [04:54] hi :) [04:54] hullo [04:55] I can't tell if I got IRC working or if nobody is saying anything [04:55] tpmccusker, you could ask #freenode ;) [04:56] they probably know channels that are very alive [04:56] Do people talk there? [04:57] I think the topic here also lists our talk channel. [04:58] nice, my boot process now has a few less errors. the most prominent is now "/sbin/cryptsetup: not found" [04:58] Sven_vB, this is a Ubuntu support channel; discussion/talk should be in -offtopic; discussions here should be on-topic [04:58] sorry Sven_vB, my comment should have gone to tpmcusker - apologies! [04:58] no prob! === [1]MrMobius is now known as MrMobius [05:37] Hi, I was trying to configure a vpn but "Configure VPN" option missing in menu? [05:37] Simple goggling shows this - https://askubuntu.com/questions/917896/configure-vpn-option-missing-on-menu [05:37] Tried that but still no configure option === im0nde_ is now known as im0nde [06:05] how long is this ppa going to be in testing? https://itsfoss.com/ubuntu-official-ppa-graphics/ [06:05] I need cuda 10 [06:34] Hello. [06:34] I would like an update of an old software being applied to the ubuntu repositories. How can I do that? [06:35] Could someone point me in the right direction? [06:35] DarkTrick, I'm not sure what you mean, but `sudo apt update` will update your software repository lists so your machine will know what packages can be updated; and `sudo apt full-upgrade` will upgrade all packages that can be upgraded [06:37] DarkTrick: which repo? [06:37] guiverc, I want to update the content inside the repo, that is on the server. [06:37] EriC^^, the ones, that are set up by default on ubuntu. [06:38] in particular: There is an update of the app "gjiten" available on github [06:38] I would like to apply that update to the repos [06:38] DarkTrick, are you asking how to become a MOTU? [06:38] (MOTU - master of the universe; or 'universe' repo) [06:39] !info gjiten [06:39] gjiten (source: gjiten): Japanese dictionary for GNOME. In component universe, is optional. Version 2.6-3ubuntu1 (bionic), package size 122 kB, installed size 388 kB [06:39] guiverc, thanks for the translation of MOTU [06:39] no, i don't [06:39] It would be enough, if I can ask a MOTU to apply the update [06:40] or at least... like ... pushing the update in that direction [06:41] I'm not yet firm with any process regarding that. But I'd like to take the first steps [06:43] DarkTrick, I'd request it via launchpad bug; PPA's are much quicker/faster as no review/process etc [06:49] guiverc, I'm not entirely sure what that means for me, but I will checkout launchpad and PPAs. [06:49] guiverc, Thank you [06:49] Hello, how can I install an old version cmake? [06:49] sorry i was unclear; trying to do multiple things... your answer was bug report. Because PPA's are easy (so easy I can upload to them), they are an alternative in the short term, ie. unrelated to your actual question, [06:52] DarkTrick, ^ [06:53] guiverc, d [06:53] (which is a "thumbs up" in ascii) [06:53] :) [07:04] Hello [07:05] Sir_Andrei, Hi, if you have a Ubuntu support question, please just ask it (try to keep to a single line, and be patient, people will reply when they can) [07:35] hey is there a way to reduce the dbm to increase roaming with intel wifi drivers ? [07:35] iwconfig sens does not seem to be permited [07:35] i need to have more agressive roaming [07:47] I want to run a GUI program remotely with xpra. How can I start xpra on ubuntu 18.04. I tried starting xpra with "xpra start :10". I got error message saying start xpra with systemctl. [07:47] I tried start xpra with systemctl start xpra [07:47] but xpra still failed to start [07:47] what is the right way to start xpra, thx [07:57] xieyi: https://xpra.org/manual.html [08:04] lotuspsychje: thx for your reply. but ubuntu 18.04 doesn't accept the standalone way of launching [08:04] and I cant launch it through systemctl [08:05] I launch with "systemctl start xpra" [08:28] I was searching exactly what this site says: https://boopathi.in/blog/capturing-per-process-bandwidth-usage-using-nethogs/ But it doesn [08:28] t work for me. [08:29] Only "sudo nethogs -t" works without problems. [08:29] Any ideas? [08:40] Ublx: which ubuntu version, kernel version, are you working inside a VM or container? [09:14] tomreyn: Ubuntu 18.04, installed on a Dell Latitude E6540, newest update. [09:17] !pastebin | Ublx: can you show some of the failing commands on a pastebin [09:17] Ublx: can you show some of the failing commands on a pastebin: For posting multi-line texts into the channel, please use https://paste.ubuntu.com | To post !screenshots use https://imgur.com/ !pastebinit to paste directly from command line | Make sure you give us the URL for your paste - see also the channel topic. [09:39] tomreyn, ubottu : https://paste.ubuntu.com/p/Smz58Hvwqk/ [10:06] Hello! I am looking for a drop-down spell-checker app for general purpose (like browsing the web or writing in libre office) for an ESL friend who has trouble spelling. Any recomendations? [10:06] so when she types, the recomended words will drop down for her to selec [10:09] i dont think such a OS-wide spell checking tool exists [10:10] ayew: what about something like for libre? [10:10] to be clear you want something that brings up options for every word as they type? not just ones it marks incorrect? [10:11] yeah [10:11] so as they type it gives options, kind of like how it does in a smartphone [10:12] so really predictive text input, not a spell checker [10:12] ah yeah! that's the word I was looking for. Something like that [10:12] https://gitlab.com/bharadwaj-raju/TextSuggest [10:12] saw this a while back, never tried it === grindcrusher1 is now known as grindcrusher [10:13] Ublx: i was rather looking for a oneliner you run and the output it produces, showing an error message. [10:13] OS X has such a feature built in as well [10:14] ayew: thanks! I'll try it out! [10:14] Ublx: you said -t works fine, but that's what this script uses. [10:20] Is this hidden fallback to Google DNS still an issue? https://bugs.launchpad.net/ubuntu/+source/systemd/+bug/1449001 [10:20] Launchpad bug 1449001 in systemd "systemd-resolved: please do not use Google public DNS by default" [Undecided,New] [10:21] I have my router configured to get DNS from my ISP and I see Google DNS for IPv6 addresses... [10:22] It is totally possible I have manually configured something but I cannot find anything in the config files of my laptop. [10:25] I am running 18.04.3 [10:28] is there a difference between export PATH = /foo/bar:$PATH and just PATH = /foo/bar:$PATH ? [10:29] I think exporting affects child process as well? [10:40] Answer to myself: https://github.com/systemd/systemd/blob/013a87517d0a6682d7140321ce80d18c54fd1f57/meson_options.txt#L223 [10:40] Seems like fallback DNS is still hard coded to systemd. This is really sad. [10:40] hi [10:40] what is this telling me: https://ideone.com/CGCSzm ? [10:41] elias_a: it isn't it's part of systemd-resolved afaik [10:43] export: I don't understand. Please elaborate! [10:44] elias_a: see /etc/systemd/resolved.conf that should provide some insight, one could also probably disable the service entirely and configure something else but idk how stable that configuration would be. [10:44] julius_: You have probably manually installed some packages that are incompatible with the ones you are trying to install. [10:46] I upgraded from ubuntu 16.4 to 18.4 copied ny config files accross etc. postfix says it starts but status says: Active: active (exited) since Tue 2019-09-10 04:36:02 MDT; 8min ago - it does have port 25 open, but not working - how do I determine why it exited? [10:47] export: Ok. The problem is I have FallbackDNS= commented out and I get IPv4 DNS from my ISP quite correctly. For some reason IPv6 DNS points to Google DNS and I cannot understand why. [10:48] julius_: it tells you the Ubuntu release is no longer supported since libssl-dev 1.0.2g-1ubuntu4.15 is the lowest version in the 16.04 Xenial archives [10:48] !info libsll-dev xenial | julius_ [10:48] julius_: Package libsll-dev does not exist in xenial [10:48] !info libssl-dev xenial | julius_ [10:48] julius_: libssl-dev (source: openssl): Secure Sockets Layer toolkit - development files. In component main, is optional. Version 1.0.2g-1ubuntu4.15 (xenial), package size 1107 kB, installed size 5483 kB [10:48] elias_a: the default is to provide those services, from the arch wiki (not sure if ubuntu behaves the same way) but it says to set FallbackDNS without giving it any addresses. [10:48] jpmh: it's normal; my working postfix also shows "active (exited)", so look elsewhere (journalctl? /var/log/mail.log?) for why it's not working [10:48] export: If you look at the first link I sent you can see that there has been a hard coded setting to use Google DNS. This has overridden the settings in some test cases. [10:48] elias_a: https://wiki.archlinux.org/index.php/Systemd-resolved#Fallback [10:49] export: Let us talk about Ubuntu. [10:49] elias_a: we are talking about systemd. [10:49] mgedmin: I never saw th exited and do not see in on 16.4 system - and IT IS NOT working [10:50] elias_a: if you don't want to try it then the hardcoded defaults will obviously show, i just tried to help guide in a workaround as it is builtin to the service, disable or try to fix it. [10:50] well, for me it says "Process: 19649 ExecReload=/bin/true (code=exited, status=0/SUCCESS)" right below the "active (exited)" line [10:50] export: Yes. In order to know which version wee need to talk about version numbers to continue with. At that stage I am not interested in learning the version numbering of some other distro. [10:51] it's not surprising that /bin/true exits; it's a bit surprising that the postfix.service unit uses ExecReload=/bin/true [10:51] export: Ok. Now I see your point - thank you! [10:51] elias_a: nothing about versions, it was just something noted, from another distro, if ubuntu hasn't in fact compiled anything out and you haven't tried what i suggested then disable it and move on. [10:51] ah, postfix.service is a decoy unit, check systemctl status systemctl status postfix@-.service [10:52] export: I see this differently. Hard coded DNS fallback is a security issue and should be handled as such. [10:52] elias_a: per this link it shows if the option is not given it defaults to a compiled-in list. which could potentially mean "commented out" to mean not given, i'd suggest trying to provide nothing and see if you are capable of an override to the defaults. [10:53] elias_a: it could potentially, yeah. [10:53] julius_: did you run sudo apt-get update before you ran sudo apt-get install openssl libssl-dev zlib1g zlib1g-dev ? [10:54] julius_: which ubuntu release ( nc termbin.com 9999 < <(lsb_release -ds;cat /proc/{version,cmdline};echo "Session: $XDG_CURRENT_DESKTOP ($XDG_SESSION_TYPE)";echo Shell: $SHELL) ) is this? [10:55] export: We are talking about two separate problems here: 1) how to cope with the existing systemd and 2) how does the linux community react to some airhead doing potentially dangerous stuff like this. [10:55] elias_a: what does systemd-resolved.service's journal log show? That usually records the DNS servers it is using with "Switching to system DNS server ..." [10:56] elias_a: 1) workaround or disable 2) nobody wants to fork it or use something else [10:56] export: Both are important. I see this brainless Lennart thing really annoying. [10:56] please, no personal attacks, elias_a [10:56] TJ- but this is ubuntu xenial, how can it now know about packages that are for xenial? [10:56] elias_a: personal opinions are fine, but being upset won't change anything, acting will. [10:56] I believe systemd-resolve --status [10:57] will show you what DNS servers will get used [10:57] tomreyn lsb_release -a says "xenial" [10:57] tomreyn 16.04.3 [10:59] julius_: so that's an outdated 16.04 LTS installation. you can use apt(-get) to install the latest security and bug fixes. [10:59] ah, now its working. apparently i fucked with the sources.lsit. copied it over from another system and its wroking now [10:59] mind the language, please [10:59] sorry [11:00] so basically what happens is that one if the packages i want to install knew about libssl.1.0.0 1.0.2 but my system was outdated and 1.0.2 was not yet know to apt? [11:00] you need to keep systems up to date if security matters to you. [11:00] those are virtual environments that are seperated from the internet [11:00] yes, i think this is what happened. [11:00] but yes, youre right [11:01] but how can apt not know about it if a package already knew about it? [11:03] you had apt sources configured incorrectly, putting apt into a situation where one apt source made it aware that some package was available which depends on a certain version of another package but you did not have the apt source configured which informs apt how to get the other packages newer version [11:03] this is common when you have e.g. a working apt source for the base archive but not for security updates [11:05] julius_: my bet would be there was no entry in the broken sources.list for xenial-updates [11:16] hıw can i understand ip is static or dynamic ob ubuntu? [11:17] spinningCat: First learn what static and dynamic IPs are. [11:18] i know what static and dynamic ip are [11:20] TJ- there are only entries for latest main / latest stable in htere [11:20] spinningCat what was your question? [11:21] sec [11:21] spinningCat: The network manager should have configuration information for each network device. [11:21] this is server machine [11:21] i dont have gui [11:22] spinningCat: The network manager is accessible from the command line [11:23] spinningCat: "ip addr show" "dynamic" == DHCP assigned [11:23] ubuntu server 18.04 LTS and later default to systemd-network (manageable also through netplan) === ThreePhase is now known as walnut_burl [11:24] https://i.itsosticky.com/14n65l9.png [11:25] is this correct setting? [11:26] i still dont get what you were asking in the first place [11:26] i just set static ip for my system [11:26] i wanna make sure i did this correctly? [11:26] you did set a static ip for a interface, yes [11:27] did you edit that file by hand? [11:27] yesd [11:27] spinningCat: interface names are usually prefixed "en" (N) not "em" (M) [11:28] spinningCat: en == EtherNet [11:29] the screenshot you posted states that /etc/network/interfaces (the legacy network configuration file, as the comments on top of it indicate) is unwritable, you seem to have run nano without sudo [11:30] i configured it with sudo of course [11:36] tomreyn: If I would know my mistake I hadn't ask. I hoped for an idea/help. Hm, what would you do to solve it? [11:36] also according to that commend you have to sudo apt install ifupdown or the files in /etc/network/ will simply be ignored? [11:36] *comment [11:37] theres probably a tool in debian to configure that file, by hand is not always the best idea [11:37] nah, sudo vi is how debian people roll... regular people use system preferences -> network to configure static IPs for network-manager setups on desktops [11:38] of course it does work [11:40] Ublx: I'm still trying to know (without having to run it on a VM) what the very errors you are running into are. so right now i'm unable to help since i've not seen any error messages or malfunction. [11:42] hi folks, trying `sudo mount -t afp afp:... ` gives me `mount: /media/...: unknown filesystem type 'afp'.` [11:42] any suggestions [11:42] i am on 18.04 [11:43] tomreyn: It does nothing. That's so confusing. No output, no error. [11:45] Ublx: have you tried to just run nethogs itself, without the script? does this work fine? i'm wondering whether this '{...}' you're passing to the shell is correct syntax. [11:47] gebbione well what kind of filesystem is afp? [11:48] https://en.wikipedia.org/wiki/Apple_Filing_Protocol [11:48] o can open it through nautilus/file but it wont mount on terminal [11:48] the kernel doesn't support that; wikipedia says there's a fuse-based implementation [11:48] nautilus doesn't use kernel-level mounts, it uses gvfs [11:49] there's gvfs-fuse so you should be able to access the files in a dot directory in your home somewhere, if you've mounted the thing with nautilus [11:50] https://en.wikipedia.org/wiki/GVfs [11:50] ah it's in /run/user/1000/gvfs/ these days [11:51] thanks for pointing that [11:51] it works :) [11:52] tomreyn: nethogs works like I wrote. But this code (source website I sent above, too) does not work. [11:53] Has anyone actually got pulseaudio's airplay implementation to work? I have appletv built into my samsung smart tv. I can see it on my iphone as an audio target for airplay. Where do i even begin on ubuntu? A quick google indicates airplay support has been baked into pulseaudio for a while now [11:54] Is there a GUI interface for searching for airplay devices, or even seeing them? My ubuntu and tv are both on the same LAN [11:58] Most misleading error message ever (from apt-get source systemd): whilst using squid-deb-proxy on the LAN "Could not open file systemd_237-3ubuntu10.28.dsc - open (13: Permission denied) [IP: 2a02:8011:2007:0:ba27:ebff:fe18:6c 8000]" [12:00] oooh squid-deb-proxy, I still have scars [12:01] jenkins deb repo redirects to a random mirror; new mirror shows up, is absent from my squid config, apt says that jenkins.org 403 forbidden when it's actually _my_ squid that's giving the 403 [12:02] TJ-, what's your current directory ? can you download it in your home dir ? [12:06] ioria: that was the thing; I was chasing the DENIED on squid since it was logging TCP_MISS_ABORTED but the proble was on the client - no write permissions to the directory [12:06] TJ-, i see [12:07] mgedmin: yes, you have to keep a tight rein on those mirror-dstdomain.acl definitions [12:08] mgedmin: I'm currently trying to figure out one to cope with the raspbian.org mirrors that generate FQDNs outside the set [12:15] Hey folks [12:29] Ublx: http://manpages.ubuntu.com/manpages/bionic/man8/nethogs.8.html states that nethogs defaults to interface eth0 if not specified. your script at https://paste.ubuntu.com/p/Smz58Hvwqk/ does not specify an interface. on ubuntu 18.04, there is usually no interface named eth0 (unless you undid the configuration for predictable network interface names) [12:31] Ublx: also ensure that $HOME/.nethogs/log/ exists and is writable by the process trying to write it, and unless you have passwordless sudo configured, consider removing the "sudo" off line 9, since this script appears to be running as root anyways. === Wryhder is now known as Lucas_Gray === grumble is now known as \emph{grumble} === EriC^ is now known as EriC^^ === DarkTrick_ is now known as DarkTrick [13:31] I'm really struggling to get pulseaudio to see my airplay device on the network. My understanding is avahi is responsible as a daemon for finding bonjour services etc. Is there any way i can debug this? [13:32] I installed an npm package bonjour-browser and ran that, and i can see the device. my iPhone can also airplay to it just fine so it's there and working and ready [13:35] mdns-scan? avahi-browse -a? [13:35] avahi-browse _airplay._tcp -v -r can see it [13:36] mgedmin: so how exactly does pulseaudio find it? through avahi? [13:36] I've no idea, I've never had an airplay device [13:36] :) [13:36] avahi is the canonical mDNS implementation on linux, so probably? [13:37] deanc: have you installed pulseaudio-module-raop? [13:38] yes meg [13:38] yes mgedmin [13:38] pacmd list-sinks only finds my sound card [13:38] and the paprefs step mentioned in https://superuser.com/questions/331126/how-can-you-stream-music-from-ubuntu-to-airplay-speakers-on-your-network-with-ub ? [13:39] yes ive done both things there mgedmin [13:39] google also tells me there are two versions of the airplay protocol [13:39] https://hfujita.github.io/pulseaudio-raop2/ [13:39] do you know which version your device uses? [13:39] mgedmin: if you read the README of that repo it says it's merged into pulseaudio 12 [13:39] or something else, either way im on 12.2 [13:43] does pacmd list-modules show the raop-discover module wasloaded? [13:44] yes [13:46] I'm out of ideas [13:46] same :) [13:48] hi [13:48] mysql-5.5.5-10.0.38-mariadb-0ubuntu0-16.04.01 <- does this mean mysql or mariadb is installed= [13:56] mariadb from a ppa would be my guess [13:58] ubuntu itself has no package called 'mysql' or 'mysql-5.5.5'; it has packages called 'mysql-server-5.7' and 'mariadb-server-10.0' [14:12] I set up my ubuntu box years ago and I don't remember how I configured /tmp. If I don't see it in the output of mount, is it in the / partition? [14:12] I'm running 19.04. [14:15] waltman: Only mount points are listed. If a path is not seen in the output of mounts, that would indicate that the path is not part of a mount definition. [14:15] df -h /tmp [14:15] findmnt /tmp [14:17] Thanks. That's what I thought but I wanted to make sure I wasn't missing something dumb. [14:17] by default ubuntu doesn't use a separate /tmp partition [14:18] gotcha [14:20] some of the files which used to be stored in /tmp are nowadays stored in /run though, which is a tmpfs by default [14:23] I'm dealing with a separate problem and someone suggested /tmp filled up. / has 14 GB free so it seems unlikely. [14:24] waltman: unless root-fs ran out of inodes [14:24] waltman: ^^^ only happens with LOTS of small files though [14:27] EriC^^ yo [14:29] waltman: df -hi /tmp [14:29] It's got 1.2 million inodes free. [14:30] should be enoug for now ;) [14:30] do i create ntfs partition for windows? [14:31] fengway: go to ##windows for help with installing Windows === ianychoi_ is now known as ianychoi [14:41] hi [14:41] ifup for ubuntu 18? [14:41] network is down :/ [14:41] I need it for ISO minimal install with package download [14:41] I am in console, ip addr [14:42] phobosoph, what exactly is the problem? [14:43] wired lan or wifi? [14:43] lordcirth_: 1) Ubuntu ISO installer 2) Setup console 3) ip tells me that ens3 is down. [14:43] LAN [14:43] DC [14:43] phobosoph, please keep your messages on one line. Did the installer say it set up networking? [14:44] ahhh! this can be the reason [14:52] i am on live ubuntu usb now and want to resize partition to install windows. But i can not resize my partition, it may be because it is encrypted. What can I do now? [14:54] fengway: decrypt it first then resize the /dev/mapper device [14:54] maybe [14:54] don't think I've never done it [14:54] do i decrypt it while i am on live usb_ [14:54] yes [14:54] how [14:55] Hi all, I'm trying to install ubuntu on my thinkpad t440 using minimal iso. However, it seems I can't detect the wireless connection. When I run `ip a`, it doesn't return the wireless interface at all. I've tried looking around, and several guide / posts suggested to use ethernet cable instead. Do you have any idea why the wireless interface doesn't show up? [14:55] You should be able to use the cryptsetup tools to shrink it, I think? [14:55] fengway, * [14:55] lordcirth_ how do i use that_ [14:55] lamurian, What Ubuntu version? Does the wifi card show up under 'lspci'? Is it disabled by physical switch? [14:55] fengway: how did you encrypt the drive in the first place? [14:56] fengway, 'man cryptsetup' [14:56] leftyfb ubuntu asked about it when i was installing ubuntu [14:56] fengway: https://unix.stackexchange.com/a/399625 [14:56] I'm using 18.04 LTS version, lordcirth_ [14:56] lordcirth_ i dont know what that means [14:56] fengway: of course, before you're doing any of this, you have backups right? [14:57] haven't tried `lspci` yet, wait a moment [14:57] fengway, the 'man' command brings up the manual for other commands. It is useful. [14:57] !man | fengway [14:57] fengway: The "man" command brings up the Linux manual pages for the command you're interested in. Try "man intro" at the command line, or see https://help.ubuntu.com/community/UsingTheTerminal | Manpages online: http://manpages.ubuntu.com/ [14:58] why must everything be so complicated with linux [14:58] It isn't. You'll learn soon enough, if you chose to. [14:58] leftyfb i don not know how to follow that [14:59] fengway: then you shouldn't be attempting to resize an encrypted drive [14:59] fengway: encryption isn't meant to be easy [14:59] Well, it should be. But it isn't. [14:59] holy shit, ubuntu could have warned me about encryption before it told me to set it up [15:00] fengway, it didn't "tell you" to set it up. It's an option, which is off by default. [15:00] And if it's a laptop, it should be encrypted. [15:00] encryption should be easy to turn off, what kind of backwards encryption is this [15:01] Why would block device encryption be easy to just "turn off"? [15:01] if i have password i should be able to turn it off [15:01] fengway: that's not how encryption works [15:01] fengway: maybe you should research thing before playing with them [15:01] yea should have sticked with windows [15:02] or bought mac, macOS is best anyway [15:02] fengway: you can't just turn off full disk encryption in windows, either [15:02] or osx [15:02] lordcirth_ : I've run `lspci`, network controller listed as Intel Corporation Wireless 7260 [15:02] because it would have to re-write every single file to the disk [15:02] can anyone guide me throught this resizing_ [15:02] i need it for work by tomorrow [15:02] fengway: hey [15:02] EriC^^ Thank god, i need your help man [15:03] what's up [15:03] He's trying to shrink a LUKS partition so he can dual-boot Windows. [15:03] EriC^^ I am on live usb but i can not resize my partition because it is encrypted [15:03] oh crap [15:04] cant help much there man, never used luks or done it before [15:05] do you not have any space at all? can you paste "sudo parted -ls | nc termbin.com 9999" again? [15:05] fengway, have you identified the partition you want to resize? [15:05] fengway, backup your files to an external drive, and start over [15:05] When I tried `ip link show`, the wireless interface is not listed there. I wonder what's wrong here. Alternatively, I've also tried booting into ubuntu-mate and other distro live CDs, and it can connect to the internet without a problem [15:05] lordcirth_ i think so [15:06] EriC^^ it say 0 free space on gparted, maybe because it is encrypted_ [15:06] NorthwestVegan, starting over shouldn't be needed. [15:06] fengway: well, i guess it's a learning process, maybe lordcirth_ can help as well [15:06] EriC^^: we've already posted instructions [15:06] " leftyfb i don not know how to follow that" [15:06] yea i do not know [15:07] what's the link? [15:07] lordcirth_can you follow me through this quick [15:07] EriC^^: https://unix.stackexchange.com/a/399625 [15:07] EriC^^: my suggestion is to decrypt and then resize [15:07] fengway, First, you need to open the encrypted drive so you can resize the filesystem. [15:07] im just saying that backing up a few files and reinstalling may be less time than anything else [15:08] fengway: https://help.ubuntu.com/community/ResizeEncryptedPartitions#Reduce_an_encrypted_partition [15:08] but if the existing system is highly configured and theres lots of things he doesnt want to redo idk [15:08] lordcirth_ how_ [15:08] fengway: i can walk you through if you want, first you'll need to boot an ubuntu live usb [15:08] EriC^^ i am on live usb now [15:09] also, you might want to backup your important files in case something goes wrong, often with fs resizing and whatnot you never know [15:09] yea sure [15:09] EriC^^: I think those instructions missed a step. I don't think the /dev/mapper devices show up until after you decrypt the volume [15:09] leftyfb: yeah it's further up top as you said [15:09] ah right, missed that [15:09] fengway: type "sudo parted -ls | nc termbin.com 9999" if you're good to go [15:10] EriC^^ done [15:10] paste the link here [15:10] EriC^^ https://termbin.com/vmu5 [15:11] fengway: type "sudo cryptsetup luksOpen /dev/nvme0n1p3 crypt3" [15:12] you seem to have enough space on the nvme [15:12] oh i need to decrypt it, let me set up keyboard on my language here so i can type pw [15:12] Is there any other specific information I need to provide here? lordcirth_ [15:12] OerHeks: that's the encrypted part [15:13] lamurian, what is your kernel version? 'uname -r' [15:14] my xenaial Notebooks can communicate perfectly over my USB uart-to-uart cable using this command: "stty -F /dev/ttyUSB0 sane raw -echo && rlwrap socat -d -d GOPEN:/dev/ttyUSB0 STDIO" but when I use my bionic live session as one side, it doesn't see any data, and the only thing I can receive on the xenial side is one Unicode invalid character replacement character per each line that I send from bionic. what's wrong? [15:14] why is there only limited languages i can set up for keyboard on live usb [15:15] fengway, probably to save disk space. [15:15] limited space [15:15] fengway, you can install more packages though. [15:15] seems like its working now [15:15] EriC^^ done [15:16] lordcirth_ It is 4.15.0-20-generic [15:16] fengway: type "sudo vgscan --mknodes" [15:16] fengway: then type "sudo vgchange -ay" [15:16] lamurian, if you run 'locate iwlwifi' does anything show up? [15:16] EriC^^ done [15:17] lordcirth_ : It seems the live USB doesn't have the command `locate` [15:17] When I issued the command, it returns "locate: not found" [15:17] lamurian, ah ok. "cat /etc/modprobe.d/iwlwifi.conf" - anything? [15:18] no iwlwifi.conf in the directory [15:18] fengway: type "sudo parted -ls | nc termbin.com 9999" again [15:19] EriC^^ Error: /dev/mapper/crypt3: unrecognised disk label [15:19] no worries, what's the link [15:19] EriC^^ https://termbin.com/05r5 [15:20] fengway: type "sudo e2fsck -f /dev/mapper/ubuntu--vg-root" [15:21] EriC^^ should i fix all the inodes? [15:21] haha, hilarious. I'm using said bionic live session to recover a xenial. when I try the same command inside the /target chroot (where the to-be-recovered xenial is), it works. nice! [15:21] fengway: yeah [15:21] so really what does the bionic live session do to screw that up? [15:21] EriC^^ done [15:21] fengway: how big do you want to make it? it's currently 510GB [15:22] 120gb [15:22] ok, type "sudo resize2fs -p /dev/mapper/ubuntu--vg-root 120g" [15:22] is this the windows partition? [15:23] no that's the ubuntu one [15:23] oh no [15:23] i want the windows one 120gb [15:23] ah [15:23] ok, type "sudo resize2fs -p /dev/mapper/ubuntu--vg-root 390g" [15:23] lamurian, ok. I think you need a newer kernel. "sudo apt update && sudo apt full-upgrade && sudo apt install linux-generic-hwe-18.04 && reboot" [15:24] it's doing its stuff now, do you guys learn this stuff by experience or study? [15:24] lordcirth_ : I didn't know I could update the kernel of live USB iso. I'll try doing so [15:25] lamurian, oh, you are still in live? Rebooting won't work then [15:25] lamurian, in that case you need to download the 18.04.3 ISO, not the original 18.04 [15:25] fengway: mostly experience for me [15:25] fengway, you run into a problem, you research how to fix it, then you know. [15:25] I compared the output of "stty --all --file=/dev/ttyUSB0" inside and outside the chroot, they're identical according to diffuse. [15:26] Hey, when I do apt upgrade I get 0 upgraded, 0 newly installed, 0 to remove and 1 not upgraded. [15:26] I followed the internet and tried apt dist-upgrade, but that didn't help either [15:26] MindSpark, run apt full-upgrade [15:27] Does anyone know what I need to do? [15:27] or dist-upgrade, should work too [15:27] lordcirth_ is it the one listed here? https://help.ubuntu.com/community/Installation/MinimalCD [15:27] I would like to confirm tho, what was the problem that my network interface is not detected and how does it relate with my kernel version? [15:27] both don't do it :/ They show the exact same message [15:27] what package is it? apt list --upgradable # should show it [15:28] lamurian, hmm, there might not be a mini.iso with the HWE kernel [15:28] OerHeks, that show nothing. It's docker-ce and docker-ce-cli. [15:28] i A docker-ce-cli Breaks docker-ce (< 5:0) [15:28] EriC^^ done [15:28] MindSpark, oh.. i dunno about that, please mention such important ifo at start [15:28] lamurian, when drivers don't work, especially on newer hardware, a kernel update is usually a good thing to try. [15:29] waste of time [15:29] lordcirth_ : I see.. so it might be better to install ubuntu-mate instead of the mini iso? [15:29] fengway: type "sudo e2fsck -f /dev/mapper/ubuntu--vg-root" [15:29] Is it possible to later declutter all packages into just the barebone installation? lordcirth_ [15:30] EriC^^ done [15:30] lamurian, It's also possible that 4.15 is new enough, but the mini.iso just doesn't have your driver. [15:30] OerHeks, sorry and thanks [15:30] lamurian, that depends how much you want to "declutter". If you just want to save space, there's only a handful of large packages you could remove. [15:30] ls -la /lib/firmware|grep wifi-7260 | nc termbin.com 9999 lamurian please show me the appearing link/url [15:30] even nicer. seems like the chrooted xenial calibrated something correctly. now, magically it works in the bionic live session as well. [15:30] MindSpark, you might want to reask in #docker [15:31] fengway: type "sudo lvdisplay | nc termbin.com 9999" [15:31] EriC^^ https://termbin.com/e54v [15:31] OerHeks, alright, will do, thanks [15:34] lordcirth_ : I'm not sure how it works in ubuntu, in my arch box I only need base and base-devel as my default installation. I've done some searching, it sesems that base-devel is equivalent to build-essential [15:34] Is it possible to uninstall all the packages and just leave the base and build-essential package? [15:35] lamurian, there isn't really a 'base' metapackage in Ubuntu, no. If you really want to be minimal, you could use debootstrap to install Ubuntu yourself? That's more work, though. [15:35] ubuntu server + build-essentials [15:35] how do I check if my kernel has support for ttyUSB0 serial? [15:36] at boot time I mean [15:36] Sven_vB: if you're running a supported version of ubuntu, then it does [15:36] Yeah, the ubuntu-server package is probably the closest. [15:36] leftyfb, thanks! [15:36] Or just use the server iso [15:37] where can I see which console= kernel parameters I currently use? [15:38] found /proc/cmdline [15:38] lordcirth_ : I see. I need to read more about debootstrap then, thanks for your help [15:38] currently I don't have any console=. now if I add my serial, do I need to add my screen as well somehow or will it always print to the screen as well? [15:39] fengway: trying to figure out the sizes, i'm not 100% sure resize2fs means gigabyte or gebibyte by the "g" 1 sec [15:41] i think it means gebibyte by "g" according to the manpage [15:42] EriC^^ doesnt't matter that much? like if i get little more or less space, or is it more important for something else? [15:42] Can anyone suggest a good thumbnail previewer they've used? [15:43] jym, often that function comes with a programm, like nautilus, darktable, and such. [15:44] preview thumbnails over network is utterly slow [15:45] OerHeks: I saw gThumb, seems okey but no keyboard control. There's pheonixslides for osx, which i like [15:47] fengway: can't really find any consistent pages about it, anyways type "sudo lvreduce -L -83G /dev/ubuntu--vg/root" if it turns out to use a different unit we can still tune it later [15:47] EriC^^ Volume group "ubuntu--vg" not found Cannot process volume group ubuntu--vg [15:48] OerHeks: Example https://blyt.net/phxslides/ [15:48] fengway: try "sudo lvreduce -L -83G /dev/ubuntu-vg/root" [15:49] EriC^^ Logical volume ubuntu-vg/root successfully resized. [15:51] fengway: nice, let's see what stuff looks like [15:51] fengway: "sudo lvdisplay | nc termbin.com 9999" [15:52] if resize used GiB, then hopefully lvreduce also used GiB and you have 391GiB lv === sveta_ is now known as Sveta [15:52] EriC^^ https://termbin.com/g0l8 [15:54] EriC^^ not good? [15:55] let's see how big the filesystem is [15:57] EriC^^ https://termbin.com/g0l8 ? [15:58] does anyone know what the holdup is on packaging the 435 nvidia drivers for 19.04? [16:00] fengway: nah, sorry, try to mount it, "sudo mount /dev/ubuntu--vg-root /mnt" [16:00] EriC^^ mount: /mnt: special device /dev/ubuntu--vg-root does not exist. [16:01] fengway: my bad, nevermind, type "sudo blockdev --getsize64 /dev/mapper/ubuntu--vg-root" [16:02] EriC^^ 420650942464 [16:02] fengway: ok, the filesystem is currently 420gigabytes [16:03] so i guess we need to resize it more, i tbh im kinda pissed at the whole computer world with the gigabyte gibibyte thing, geez [16:04] doesn't it mean i have 110 gb for windows? [16:04] no, it's more like 89gigabytes for windows [16:04] oh [16:04] fengway: anyways, lets resize the filesystem again [16:04] EriC^^ cool [16:05] this time we want to give it in GiB [16:05] ok, type "sudo resize2fs -p /dev/mapper/ubuntu--vg-root 360g" [16:06] EriC^^ done [16:06] fengway: ok, just to verify [16:06] fengway: type "sudo blockdev --getsize64 /dev/mapper/ubuntu--vg-root" [16:06] EriC^^ 420650942464 [16:07] hmm [16:07] that's odd [16:07] fengway: try mounting it [16:08] type "sudo mount/dev/mapper/ubuntu--vg-root /mnt" [16:08] EriC^^ mount: /mnt: special device /dev/ubuntu--vg-root does not exist. [16:08] i think there's a typo in the command [16:08] ah my bad again xD [16:08] type "sudo mount /dev/mapper/ubuntu--vg-root /mnt" [16:08] i used the one you gave me before, it gives that error [16:09] ah, use the one right above [16:09] EriC^^ mount: /mnt: /dev/mapper/ubuntu--vg-root already mounted on /mnt. [16:11] fengway: ok, are you sure the resize2fs did something a sec ago? [16:11] EriC^^ https://pastebin.com/n8HevmKm [16:12] that looks good [16:12] fengway: ok, try "df -H /mnt" [16:12] note the capital H not h [16:12] EriC^^ Filesystem Size Used Avail Use% Mounted on/dev/mapper/ubuntu--vg-root 380G 200G 161G 56% /mnt [16:12] that should give the size in gigabytes [16:12] ok, 380gigabytes, not bad [16:12] yea [16:13] it's fine [16:13] looks good i guess, now we still have to shrink the lvm [16:13] we want the lvm just slightly bigger than it [16:14] fengway: what's the current lvm size? type "sudo lvdisplay | nc termbin.com 9999" [16:14] tomreyn: Great tipps! Thanks a lot!! [16:14] i think a while ago it took the input as GiB not GB [16:14] let me scroll up for a sec [16:14] EriC^^: just use --resizefs surely (with lvresize) ? [16:14] EriC^^ https://termbin.com/lfik [16:15] it says GiB on the log [16:15] TJ-: yeah, that's a good idea, we were kinda following an ubuntu tutorial and unfortunately didn't [16:16] fengway: ok, so we need the lvm to be 354GiB === tucked is now known as disi [16:18] fengway: try "sudo lvreduce -L -37G /dev/ubuntu-vg/root" [16:18] @er [16:18] EriC^^ done [16:19] fengway: type "sudo lvdisplay | nc termbin.com 9999" to confirm [16:19] EriC^^ https://termbin.com/jn3b [16:20] FYI: lvm tools use base 2 values (G = gibibyte = 1024x1024x1024) [16:20] fengway: looks good [16:21] fengway: ok, continuing with the tutorial now, type "sudo pvdisplay | nc termbin.com 9999" [16:21] EriC^^ https://termbin.com/x5h7 [16:24] fengway: looks like around 120gb is "free PE", which is great [16:24] EriC^^ Are we done? :D [16:25] fengway: anyways lets unmount the root while we're at it, sudo umount /mnt [16:25] fengway: not yet, last couple steps [16:25] i will never encrypt again [16:25] EriC^^ unmount done [16:25] fengway: lol [16:26] fengway: that is the wrong way to think about this [16:27] fengway: type "sudo pvresize --setphysicalvolumesize 356G /dev/mapper/crypt3" [16:27] we might need to make it a little larger if there's not enough free PE [16:28] EriC^^ https://pastebin.com/MjuMrzVB [16:29] fengway: lets try bigger i guess, sorry i dunno exactly cause the numbers aren't adding up here https://termbin.com/x5h7 [16:30] ah wait i think they are 1 sec [16:30] i have no idea what's going on anyway [16:30] forgot to divide by 1024 and not 1000 1 sec [16:31] ok great so 120GiB is actually free [16:33] hmm 356 should have worked [16:33] fengway: try "sudo pvresize --setphysicalvolumesize 360G /dev/mapper/crypt3" [16:34] EriC^^ sudo pvresize --setphysicalvolumesize 360G /dev/mapper/crypt3/dev/mapper/crypt3: Requested size 360.00 GiB is less than real size 475.72 GiB. Proceed? [y/n]: y WARNING: /dev/mapper/crypt3: Pretending size is 754974720 not 997660672 sectors. /dev/mapper/crypt3: cannot resize to 92159 extents as later ones are allocated. 0 physical [16:34] volume(s) resized / 1 physical volume(s) not resized [16:37] fengway: i think it's cause the stuff is fragmented, the guide mentions it can happen [16:37] fengway: try "sudo pvresize --setphysicalvolumesize 370G /dev/mapper/crypt3" maybe we get lucky [16:37] EriC^^ same error [16:38] i think the next step is to figure out how to see where it has everything allocated, and maybe delete the swap or 'manually defrag it' as the guide suggests [16:41] fengway: try "sudo pvdisplay -m | nc termbin.com 9999" [16:42] EriC^^ https://termbin.com/t4vv [16:42] fengway: ah great, finally we got lucky [16:43] it's just the swap that's in the way [16:43] fengway: type "sudo lvremove /dev/ubuntu--vg/swap_1" [16:44] EriC^^ Volume group "ubuntu--vg" not found Cannot process volume group ubuntu--vg [16:44] ubuntu-vg [16:44] ^ [16:45] fengway: type "sudo lvremove /dev/ubuntu-vg/swap_1" [16:45] Logical volume "swap_1" successfully removed [16:45] ok great [16:45] fengway: type "sudo pvresize --setphysicalvolumesize 357G /dev/mapper/crypt3" [16:46] sudo pvresize --setphysicalvolumesize 357G /dev/mapper/crypt3/dev/mapper/crypt3: Requested size 357.00 GiB is less than real size 475.72 GiB. Proceed? [y/n]: y WARNING: /dev/mapper/crypt3: Pretending size is 748683264 not 997660672 sectors. Physical volume "/dev/mapper/crypt3" changed 1 physical volume(s) resized / 0 physical volume(s) not [16:46] resized [16:46] great [16:48] fengway: ok now to recreate the swap [16:48] fengway: sudo pvchange -x y /dev/mapper/crypt3 [16:48] sudo pvchange -x y /dev/mapper/crypt3 Physical volume "/dev/mapper/crypt3" is already allocatable. Physical volume /dev/mapper/crypt3 not changed 0 physical volumes changed / 1 physical volume not changed [16:49] fengway: ok, type "sudo pvdisplay | nc termbin.com 9999" to see the free pe [16:49] EriC^^ https://termbin.com/qepn [16:51] which file system is better to use qemu/libvirt images that store on ? ext4/xfs and any mount option? [16:51] fengway: sudo lvcreate -L 2.2G -n swap_1 ubuntu--vg [16:52] fengway: brb in 1min [16:52] EriC^^ Logical volume "swap_1" created. [16:53] back [16:53] fengway: sudo mkswap -L swap_1 /dev/ubuntu--vg/swap_1 [16:54] fengway: it should mention a UUID, select it and press ctrl+shift+c to copy [16:54] EriC^^ Setting up swapspace version 1, size = 2.2 GiB (2365583360 bytes)LABEL=swap_1, UUID=82e9f7e7-4108-408b-af4b-dd4cfff4d941 [16:55] fengway: type "sudo mount /dev/mapper/ubuntu--vg-root /mnt" [16:56] ubuntu--vg or ubuntu-vg? [16:56] i think it's --vg in that [16:56] mount: /mnt: wrong fs type, bad option, bad superblock on /dev/mapper/ubuntu--vg-root, missing codepage or helper program, or other error. [16:57] What would be a good graphic card for OpenCL under ubuntu ? Looking for a cost efficient pcie x16 card with actually working OpenCL drivers [16:59] fengway: type "sudo e2fsck -f /dev/mapper/ubuntu--vg-root" [17:00] EriC^^ e2fsck 1.44.1 (24-Mar-2018)The filesystem size (according to the superblock) is 94371840 blocksThe physical size of the device is 92998656 blocksEither the superblock or the partition table is likely to be corrupt!Abort? yes [17:00] https://unix.stackexchange.com/questions/540048/kvm-qemu-libvirt-image-path-suggestion-file-system-and-mount-options [17:03] fengway: hmm, it's 703MB off [17:04] fengway: that's odd, the filesystem was 380G according to df -H, we made the lvm slightly bigger at 381GB [17:06] EriC^^ HMM [17:07] fengway: this is odd, 94371840 if the block is 512bytes, is 48GB, if 4096 is 386GB [17:07] i mean we could make the lvm for root a little bigger again, taking up 1GB from the space, if it makes it happy i guess [17:07] feng [17:08] fengway: delete the swap again, "sudo lvremove /dev/ubuntu--vg/swap_1" [17:08] actually i think anyways we'd be screwed [17:08] nevermind [17:08] cause we did mkswap, so it would have written stuff in that area [17:08] okey.. [17:09] fengway: try to do a fsck, maybe it's just kind of confused, it did say 380GB in df, and it does have like 381GB of lvm [17:09] i tried before you said that, but got this: mkswap: cannot open /dev/ubuntu--vg/swap_1: No such file or directory [17:09] oh [17:09] nah but you did mkswap, i remember you pasted the uuid and everything [17:10] fengway> EriC^^ Setting up swapsp........ [17:10] maybe it's with one dash [17:10] when I boot my xenial notebook with kernel param "console=ttyUSB0,115200n8", I see nothing in my serial terminal and the notebook screen shows just an excessively rapidly blinking "_" shape cursor. what am I doing wrong? [17:10] fengway: yeah [17:10] fengway: i think it's pretty much the same [17:10] fengway: try to run the fsck again, but dont abort, let's see if it can get stuff in order and mount properly and all is good [17:11] fengway: type "sudo e2fsck -f /dev/mapper/ubuntu--vg-root" [17:12] EriC^^ https://pastebin.com/GBmWi07S [17:12] fengway: hmm [17:13] EriC^^ /dev/mapper/ubuntu--vg-root: ***** FILE SYSTEM WAS MODIFIED *****/dev/mapper/ubuntu--vg-root: 776617/23592960 files (0.9% non-contiguous), 50563815/94371840 blocks [17:13] fengway: try "sudo mount /dev/mapper/ubuntu--vg-root /mnt" [17:13] mount: /mnt: wrong fs type, bad option, bad superblock on /dev/mapper/ubuntu--vg-root, missing codepage or helper program, or other error. [17:14] adding console=tty1 at least brought the screen back. [17:14] fengway: this sucks [17:15] what happened? [17:15] fengway: i'm just glad you backed up at this point [17:15] i have no idea [17:15] is my ubuntu fucked? [17:15] we mounted it fine earlier, df said 380GB space, we had more in the lvm, we resized the lvm leaving it slightly larger than 380gb, somehow it broke, i have no idea why [17:16] i rechecked the numbers, they were all fine, we removed 37GiB from 391.76 of the lvm, leaving like 381GB for root [17:18] are we screwed? [17:18] hey, i have a quick question on the syntaxing of displays in the grub confiugration, lets say i have HDMI-1 1920x1080+3840+0 how would i append that to grub properly? [17:18] fengway: yeah i wouldn't trust it anymore, seems it had some stuff in those last blocks 93...... that it couldnt write and i guess that's where it puts some filesystem stuff [17:18] something like display=eDP-1 display=eDP-1:1920x1080+1920+0@60 append="video=HDMI-1:1920x1080+1920+0@60" [17:18] fengway: if you run fsck again what does it say at this point? [17:18] fengway: try "sudo e2fsck -f /dev/mapper/ubuntu--vg-root" [17:19] EriC^^ mount: /mnt: wrong fs type, bad option, bad superblock on /dev/mapper/ubuntu--vg-root, missing codepage or helper program, or other error. [17:19] fengway: the e2fsck command, not mount [17:21] EriC^^ https://pastebin.com/FcUMAPjJ [17:23] fengway: ok, let's try something, to delete the swap [17:23] and enlarge the lvm of root, it might work, maybe mkswap only puts in a few bytes in there as a header or something [17:24] fengway: try "sudo lvremove /dev/ubuntu-vg/swap_1" [17:24] EriC^^ done [17:26] anyone got any idea why `xdg-open` has stopped opening pdf files correctly for me (it opens them in an ebook viewer). I think it might have something to do with the fact that `xdg-mime query filetype anyfile.pdf` doesn't print anything, but I'm not sure how that works. [17:26] for example: [17:26] anyone got any idea why `xdg-open` has stopped opening pdf files correctly for me (it opens them in an ebook viewer). I think it might have something to do with the fact that `xdg-mime query filetype anyfile.pdf` doesn't print anything, but I'm not sure how that works. [17:26] % xdg-open x.pdf [17:26] xdg-mime: mimetype argument missing [17:26] Try 'xdg-mime --help' for more information. [17:26] libpng warning: iCCP: known incorrect sRGB profile [17:26] libpng warning: iCCP: known incorrect sRGB profile [17:27] fengway: type "sudo lvextend -l +100%FREE /dev/ubuntu-vg/root" [17:27] sorry for the spam, folks [17:28] EriC^^ Logical volume ubuntu-vg/root successfully resized. [17:28] fengway: ok try the e2fsck again [17:28] fengway: try "sudo e2fsck -f /dev/mapper/ubuntu--vg-root" [17:29] EriC^^ same error [17:29] for the record, `xdg-mime query filetype $file.pdf` exits with a zero exit status despite not printing anything [17:30] and evince opens the pdf file just fine [17:30] and file(1) can tell that it's a pdf doc just fine [17:30] EriC^^ is the new partition screwed or ubuntu one? [17:30] does anhyone know what determines the route order in a netstat -nr ? [17:31] rogpeppe, is that ebook viewer in your %path%? [17:31] hey, is there an ipv6 issue with the ubuntu repositories at the moment? E: Failed to fetch http://security.ubuntu.com/ubuntu/pool/main/s/systemd/libnss-systemd_240-6ubuntu5.7_amd64.deb 404 Not Found [IP: 2606:4700:30::6818:75d1 80] [17:31] OerHeks: i guess so. i'm not sure what the executable is. let me check. [17:32] if so, right-click the file , click the Open With tab and select the application you want to always open the file and click Set as default… [17:32] manually there are a lot of mimetypes to pdf to edit [17:32] fengway: k well, df screwed us over basically [17:33] fengway: the error you put said the filesystem according to the superblock was 94371840, which works out to 386GB, [17:33] OerHeks: yup, it's /usr/bin/ebook-viewer [17:34] EriC^^ so only new partition is screwed? [17:34] OerHeks: if I right-click in file viewer, it shows the correct app (Document Viewer) [17:34] fengway: i remember at one point we did resize with 360GiB, which works out to 386GB, so i believe it, so basically when we checked the size using "df -H" it said 380GB when it was in fact 386GB, so we made it smaller than it when we made it like 381GB [17:34] OerHeks: xdg-open opens it in a different app than the default there [17:35] fengway: i'm not too sure about how lvm works completely, if it were a normal partition, currently all that would be lost is whatever we did with the mkswap operation, which i think is just a few bytes, let me try to mkswap a file and see what that does actually, 1sec [17:36] OerHeks: FWIW `xdg-mime query default application/pdf` prints "evince.desktop", which is what I'd expect (i.e. the default seems to be set up correctly) [17:38] OerHeks: but `query filetype FILE` seems to print nothing, regardless of what kind of file I pass it, which seems like it might be the source of the issue [17:38] OerHeks: sorry, "xdg-mime query filetype etc" [17:39] * rogpeppe just realises that `xdg-mime` is just a shell script and has a go at debugging it. [17:41] fengway: looks like mkswap writes around 20bytes [17:42] https://paste.ubuntu.com/p/2W24PGh6JN/ [17:42] not much i guess [17:43] fengway: we can try enlarging the PE again, and the lvm, and see how the filesystem fairs, i dunno if that would work cause of stuff lvm has to write or what, but it's worth a shot i guess [17:43] hmm, looks like the problem is that xdg-mime expects `gio help info` to succeed, but it fails [17:44] ah! it's a clash in names [17:44] EriC^^ Do you think it will work? I am considering just installing windows over whole HD [17:44] with the gioui stuff! [17:44] got it! [17:48] fengway: it might, we're pretty much there, just need to expand the PE and run that lvextend command a final time [17:48] * EriC^^ getting the command [17:48] EriC^^ ok lets try [17:51] fengway: type "sudo pvresize --setphysicalvolumesize 375G /dev/mapper/crypt3" [17:52] EriC^^ done [17:55] fengway: type "sudo lvextend -l +100%FREE /dev/ubuntu-vg/root" [17:55] EriC^^ Logical volume ubuntu-vg/root successfully resized. [17:56] fengway: finally try "sudo e2fsck -f /dev/mapper/ubuntu--vg-root" [17:56] fengway: the swap is deleted right? does it show up in "sudo lvdisplay" at all? [17:57] EriC^^ /dev/mapper/ubuntu--vg-root: 776617/23592960 files (0.9% non-contiguous), 50563815/94371840 blocks [17:57] EriC^^ https://pastebin.com/jQG32cUR [17:58] fengway: are you able to mount it? [17:58] fengway: try "sudo mount /dev/mapper/ubuntu--vg-root /mnt" [17:58] EriC^^ didin't get any message on that command [18:01] fengway: nice, it mounted [18:01] fengway: does everything look still there? try "ls /mnt" [18:02] bin dev initrd.img lib64 mnt root snap tmp vmlinuzboot etc initrd.img.old lost+found opt run srv usr vmlinuz.oldcdrom home lib media proc sbin sys var [18:02] fengway: looks good [18:02] ok, so what's currently left for windows? [18:03] try "sudo pvdisplay" [18:03] https://pastebin.com/rakxtUqM EriC^^ [18:06] fengway: ok, let's get rid of the swap temporarily from fstab, so you can boot up properly, we'll create it later, you can always make a swap file pretty easily too [18:06] fengway: type "sudo cat /mnt/etc/fstab | nc termbin.com 9999" [18:07] fengway: we just have to shrink the crypt container and i think you should have the space to install windows [18:07] EriC^^ https://termbin.com/ha5h [18:07] around 110GB or something for windows [18:07] how do I check whether my processor really needs these files? "W: Possible missing firmware /lib/firmware/i915/kbl_guc_ver9_14.bin for module i915" [18:08] fengway: ok, type "sudo -H gedit /mnt/etc/fstab" and put a "#" at the start of the line /dev/mapper/ubuntu--vg-swap_1 none swap sw 0 0 [18:08] oh that seems to be about graphics only [18:08] fengway: then save and exit [18:09] EriC^^ done [18:10] fengway: ok, unmount the fs, sudo umount /mnt [18:10] EriC^^ done [18:11] fengway: type "sudo cryptsetup status crypt3" [18:11] what does it say next to offset? [18:11] EriC^^ umount: /mnt: not mounted. [18:12] fengway: i think you ran the other command [18:13] EriC^^ offset: 4096 sectors [18:13] fengway: ok [18:16] hi [18:16] does 16.04 ship with mariadb or mysql by default? mysql is still installed as command line client somehow [18:18] julius: ubuntu ships with both, you choose which one you want to use [18:19] on the work pc the version was something like mysql-5.x.y-10.x.y mariadb [18:19] EriC^^ so? [18:21] fengway: still trying to figure out what to put, tutorial is kind of vague also manpage [18:21] basically this is the command we're going to use sudo cryptsetup -o 4096 -b resize crypt3 [18:22] fengway: our PE is 383996MiB [18:23] just to be safe we'll add like 1gb on that for the crypt [18:23] so let's say 385000MiB [18:25] fengway: type "sudo cryptsetup status crypt3 | nc termbin.com 9999" and paste the link here [18:26] EriC^^ https://termbin.com/fd5d9 [18:30] fengway: ok i think this is it [18:30] EriC^^ we are done? [18:31] fengway: sudo cryptsetup -o 4096 -b 788480000 resize crypt3 [18:31] EriC^^ https://pastebin.com/7Xx01zL4 [18:32] fengway: ok, try without the offset i guess, the tutorial mentions in future cryptsetup it's deprecated [18:32] fengway: sudo cryptsetup -b 788480000 resize crypt3 [18:32] EriC^^ done [18:32] fengway: nice [18:33] fengway: type "sudo vgchange -an" [18:33] then "sudo cryptsetup luksClose crypt3" [18:33] EriC^^ 0 logical volume(s) in volume group "ubuntu-vg" now active [18:33] fengway: what does "sudo parted -ls | nc termbin.com 9999" give? [18:34] EriC^^ https://termbin.com/lb58 [18:35] ah ok, last step, we need to resize the partition [18:35] so, crypt should be 385000MiB now so we need something slightly larger than that as a partition [18:35] fengway: type "sudo parted /dev/nvme0n1 unit s print | nc termbin.com 9999" so we have a backup of the partitions sector locations [18:36] EriC^^ https://termbin.com/j26n [18:38] fengway: ok, now you need to delete the 3rd partition, then recreate it at the same starting sector, but at a smaller ending sector, 1sec [18:38] first type "sudo gdisk /dev/nvme0n1" [18:39] Found valid GPT with protective MBR; using GPT. [18:39] alright, press "p" to list the stuff [18:40] Number Start (sector) End (sector) Size Code Name 1 2048 1050623 512.0 MiB EF00 EFI System Partition 2 1050624 2549759 732.0 MiB 8300 3 2549760 1000214527 475.7 GiB 8300 [18:40] fengway: press "d" [18:40] Partition number (1-3): [18:44] fengway: choose "3" [18:44] EriC^^ done [18:44] fengway: ok, now press "n" [18:45] EriC^^ done [18:45] when it asks for starting sector put it [18:45] 2549760 [18:45] when it asks for ending sector put in 790577152 [18:45] EriC^^ Value out of range [18:45] for starting sector [18:46] fengway: it might be some gdisk discrepancy between versions of gdisk [18:47] fengway: what's the starting sector it offers [18:47] EriC^^ Partition number (3-128, default 3): [18:48] ah, press "3" [18:48] First sector (34-1000215182, default = 2549760) or {+-}size{KMGTP}: EriC^^ [18:49] fengway: use the default [18:49] Last sector (2549760-1000215182, default = 1000215182) or {+-}size{KMGTP}: EriC^^ [18:51] fengway: put in 790577152 [18:52] Current type is 'Linux filesystem'Hex code or GUID (L to show codes, Enter = 8300): [18:52] ef00? [18:52] EriC^^ [18:52] no [18:55] fengway: 1 sec, searching for luks type hexcode [18:56] for LUKS LVM, should my initramfs include /etc/crypttab? because it doesn't. also no /sbin/cryptsetup [18:57] Sven_vB: thoswe will only be included if the root-fs is listed in /etc/crypttab [18:57] IIRC, by default it will scan for LUKS devices and prompt for password. I think you only need crypttab if there's settings or a keyfile [18:57] EriC^^: there isn't one, I generally use 'Linux file-system' [18:57] When is the new ubuntu lts? [18:58] TJ-: ah, thanks [18:58] I have 18.04 just installed on a vps [18:58] !lts | mandude [18:58] mandude: LTS means Long Term Support. LTS versions are supported for 5 years on the desktop and server. The latest LTS version of Ubuntu is !Bionic (Bionic Beaver 18.04). Ubuntu !flavors may have different support durations, check their release notes for information. [18:58] fengway: keep the default, 8300 [18:58] I wonder how many years before I have to upgrade to the new lts? [18:58] Sven_vB: lordcirth_: without a crypttab entry the cryptsetup tooling won't be installed in the initrd [18:58] EriC^^ Changed type of partition to 'Linux filesystem' [18:58] !boinic [18:58] mandude: every LTS is supported for 5 years [18:58] !bionic [18:58] Ubuntu 18.04 LTS (Bionic Beaver) is the 28th release of Ubuntu and the current LTS release. Download at https://www.ubuntu.com/download - Release Notes: https://wiki.ubuntu.com/BionicBeaver/ReleaseNotes [18:58] mandude, every even .04 is LTS, so 20.04. [18:58] fengway: alright, press "w" to write the partition table [18:58] mandude: https://wiki.ubuntu.com/LTS [18:58] every 2 years? [18:58] TJ-, I do have a crypttab entry in the chroot whose GRUB I'm trying to recover. [18:58] ok [18:59] EriC^^ The operation has completed successfully. [18:59] TJ-, or rather, the LUKS now. GRUB works again. [18:59] Oh so the new LTS is coming around soon [18:59] 8 months or so [18:59] fengway: ok, type "sudo partprobe /dev/nvme0n1" [18:59] TJ-, when I boot it, errors about missing /sbin/cryptsetup and then initramfs shell [18:59] Do people upgrade from one lts to another? [18:59] EriC^^ done [18:59] using apt [18:59] fengway: "sudo parted -ls | nc termbin.com 9999" [18:59] mandude, Many do. I usually reinstall because my installs are more complex. [18:59] mandude: try #ubuntu-offtopic for non-supported related discussion [19:00] Sven_vB: right, and that'll be because the initramfs-tools cryptsetup hooks didn't figure out that the root-filesystem is LUKS [19:01] EriC^^ https://termbin.com/3qx4 [19:01] fengway: looks good, there's around 110gb for windows [19:02] I do not want to have to setup my configs again. All I need is this: setup new vps, install nginx, git, couple other things like screen, .screenrc, some home configs, nginx vhost file, ssl certs for ones not using letsencrypt and that's it. [19:02] I currently use git and git init inside /etc and /var and /home/user/ and add whatever configs I need [19:02] mandude: you should look into ansible [19:02] Then on new vps I pull from git and voila. [19:03] saltstack is better than ansible, but yes, config management is good. [19:03] People keep telling me to use ansibles or whatever. ID unno [19:03] mandude: do you have a support issue? [19:03] EriC^^ cool, more stuff left to do? [19:03] Is my git way faster or ansibles faster? [19:03] I guess ansible automates the process [19:04] I mean. I wonder how ansible is differnet than running a bash script after spinning up a vps. [19:04] \ [19:04] I mean, if you spin a new vps, you still have to install ansible and play a playbook. [19:04] I just don't get what people are trying to push on me all the time. [19:04] fengway: maybe try to boot into ubuntu real quick to make sure it's working right, also remember it lost some bytes when we had the whole mishap with df and mkswap and stuff so you know, there's no knowing what got affected could be nothing serious or not who knows [19:04] im an old man. :) [19:05] EriC^^ will do now [19:05] mandude: Cheerio to being old man! [19:06] Well my beard is almost white. [19:06] still not as long as gandalf. Anyway. [19:21] TJ-, then why would it include conf/conf.d/cryptroot and conf/conf.d/cryptsetup in my initrd? [19:21] TJ-, what does it use them for if not for my LUKS? [19:22] xenial 64bit btw [19:25] Sven_vB: sounds like the hook scripts are broken if parts are copied in but not the binaries... unless the binary is broken of course [19:26] TJ-, how can I check? [19:27] Sven_vB: grab a verbose log of the initrd.img build, with "update-initramfs -vu -k $VERSION |& tee /tmp/initrd.log" then "grep -n crypt /tmp/initrd.log" [19:28] Sven_vB: you should see stuff like "Copying /sbin/cryptsetup" and so on [19:31] where do the applications get installed when installing using snap? [19:32] TJ-, I guess -k is redundante because I have just one kernel. will do. [19:33] Sven_vB: indeed [19:33] moo [19:35] TJ-, lots of matches. http://paste.debian.net/plainh/b94f2697 [19:36] Sven_vB: that looks broken totally, there should be loads more, and it is missing a lot of parts [19:36] TJ-, maybe I forgot to install some package? [19:37] Sven_vB: what release of Ubuntu is that? that doesn't look like Ubuntu at all [19:38] TJ-, Ubuntu 16.04.6 LTS chrooted into a Bionic live session [19:39] I think I meant "into" the other way around. [19:40] Sven_vB: the "Calling hook cryptroot" should be followed by a lot of activity, such as writing ./conf/conf.d/cryptroot, copying various crypto modules, /sbin/cryptsetup and its supporting libraries, and more [19:42] Sven_vB: this is what to expect: http://paste.ubuntu.com/p/5K6NWHgKCb/ [19:43] TJ-, should I check which package provides the cryptroot hook and reinstall it? [19:44] Sven_vB: "cryptsetup" === CodeMouse92__ is now known as CodeMouse92 [19:45] TJ-, thanks! I rather meant whether it would make sense to reinstall it. well I guess it won't hurt either. [19:46] Sven_vB: then on calling update-initramfs the cryptsetup hook scripts under /usr/share/initramfs-tools/hooks/ will read /etc/crypttab and if they detect the root-fs relies on an entry there, will trigger the creation on the initrd.img:/conf/conf.d/cryptroot and copying in the required binaries, libraries, and support scripts [19:46] I am messing around with compiz, but I cant seem to find anything regarding workspaces. What could I be missing? Running 19.04 [19:46] TJ-, so can I conclude from the fact it did copy conf/conf.d/cryptroot, that it correctly detected my /etc/crypttab? [19:47] Sven_vB: that looks like a leftover and I wouldn' [19:47] ... expect the package is currently installed [19:48] Sven_vB: i'd guess that was left behind after cryptsetup package was removed but not purged [19:48] TJ-, aptitude said "The following packages will be REINSTALLED:¶ cryptsetup¶ 0 packages upgraded, 0 newly installed, 1 reinstalled, 0 to remove" so I assume it was installed. [19:48] Sven_vB: /etc/initramfs-tools/conf.d/cryptroot is NOTHING to do with the /conf/conf.d/cryptroot in the initrd, which is generated by cryptsetup hook scripts from the /etc/crypttab [19:49] anyway it is now reinstalled so I'll try again. [19:50] Sven_vB: I'd bet a new initrd.log will show it installing now === LabMonkey is now known as Mechanismus [19:53] TJ-, grep output is exactly the same [19:53] http://paste.debian.net/plainh/942204ef [19:53] Sven_vB: are you running update-initramfs from within the chroot of the target? [19:54] EriC^^ it is screwed [19:54] fengway: what happened? [19:54] EriC^^ will not load after i write encryption pw [19:55] just loads forever [19:55] fengway: the ubuntu with dots screen? [19:55] EriC^^ yea [19:56] fengway: did you try pressing Esc ? it should show what it's doing behind the scenes [19:56] EriC^^ did not try that [19:56] fengway: couple of other things to know/do ... [19:56] give it a shot, it might be something simple like waiting for swap or something maybe [19:57] TJ-, yes inside [19:57] fengway: ... 1) tap Esc key at boot-time to get the GRUB menu and edit the boot entry manually to remove "quiet splash" so you can see what is going on in text console [19:58] Sven_vB: and have you mounted devtmpfs to the target's ./dev/ so it can find the devices ? [19:59] Sven_vB: as in "mount --rbind /dev /target/dev" if /target/ is the chroot [20:00] Sven_vB: and of course ensure the entries in /target/etc/crypttab are valid especially if UUID, using blkid [20:00] TJ-, seen from the outside I have udev on /target/dev type devtmpfs (rw,nosuid,relatime,size=6076068k,nr_inodes=1519017,mode=755)¶ devpts on /target/dev/pts type devpts (rw,nosuid,noexec,relatime,gid=5,mode=620,ptmxmode=000) [20:01] Sven_vB: good, so then check /target/etc/crypttab and /target/etc/fstab entries for the root-FS are correct [20:06] TJ-, inside chroot: "# readlink -m /dev/disk/by-uuid/$(grep -oPe 'UUID=\S+' crypttab | cut -d = -sf 2)¶ /dev/sda5" and the fstab entry for / is … wrong. :D [20:07] I guess I need to update my chroot script to reflect the new LVM names [20:07] Sven_vB: looks like you'v found the problem then [20:07] yeah. thanks! [20:07] Is it sad that I know this stuff, including what the hook scripts do, without needing to think about it!? :) [20:09] yes. it means there's so much work left to do to make Ubuntu even better. [20:10] in an ideal world we'd have luks-lvm-diagnose on the Live CD and it would have told me. [20:10] ha not so much, but I've worked on cryptsetup/initramfs-tools for the last 12 years or so [20:11] written quite a few bits of tooling for it [20:11] also a chroot helper program whom I just tell the partition whose /etc/fstab could tell it how to establish my chroot. [20:11] nice, thanks for making it better! [20:12] Generally the only time these configs go wrong is when local admins/operators break them, so there's no much call for such things [20:14] So i've got an issue with rebooting a PXE/iSCSI setup. It looks like 18.04 does not log out from the target when shutting down. The target is therefore locked/unaccessible when the system reboots and iPXE tries to log in. Thoughts? [20:16] Online_: check the journal/logs to see what is going on... there've been a long list of race-conditions in systemd's reboot logic so you may have found another [20:17] my thought, you set it up, but a reboot is a disaster for you...you'll need to reconfigure or rethink how you set things up and try something different [20:19] EriC^^ https://imgur.com/a/6qbwZgh [20:19] if a reboot brought down your entire stack of cards....re architecture or redo is justified.....okay i'm done i promise [20:21] I'm actually doing "target disk mode" over ethernet using a Debian USB stick with dnsmasq and tgt that exports my Ubuntu laptop drive to a dekstop computer and netboots the raw device over the wire. I even WOL-start my desktop after the target is up. Everything works perfectly except rebooting. [20:21] fengway: do you remember what the last one was? a start job is running for dev-mapper what was the rest? are you in the live usb righ tnow? [20:22] EriC^^ Was something about swap, i am on live usb now [20:22] fengway: oh ok great [20:23] fengway: type "sudo parted -ls | nc termbin.com 9999" [20:23] https://termbin.com/z4h5 [20:24] EriC^^ [20:24] fengway: Id' recommend doing an emergency start, with "systemd.unit=emergency.target" on the kernel command-line, from GRUB menu editing, so you can control and investigate the start-up sequence [20:24] TJ-: it seems to be waiting for the swap, though we had commented it out in fstab [20:25] i will try out systemd.unit=emergency.target [20:26] EriC^^: where do you see the swap reference, in the photo? [20:27] TJ-: the last line dev mapper, i asked him about the rest, he mentioned it said swap [20:27] fengway> EriC^^ Was something about swap, i am on live usb now [20:27] EriC^^: hmmm, from earlier, was there a swap LV ? [20:28] yeah [20:28] EriC^^: were there any other LVs beside that and root-fs? [20:28] no [20:29] EriC^^: what makes me concerned is the list of failed units (which are system units) coming after some --user unuts, such as gnome-calculator [20:29] EriC^^: or is that 'just' snaps ? [20:33] EriC^^: re the swap, if commented out in fstab that would suggest systemd-fstab-generator wouldn't write a unit for it to /run/systemd/generator/ - so maybe there's a swap.unit somewhere in addition ? [20:34] yeah, maybe [20:34] fengway: any luck? [20:35] EriC^^ I forgot to take picture of the command lol, but took this picture https://imgur.com/a/ToDeVGj [20:35] fengway: ok, type "sudo cryptsetup luksOpen /dev/nvme0n1p3 crypt3" [20:37] EriC^^ done [20:39] fengway: type "sudo mount /dev/mapper/ubuntu--vg-root /mnt" [20:39] EriC^^ done [20:39] fengway: "sudo cat /mnt/etc/fstab | nc termbin.com 9999" [20:40] EriC^^ https://termbin.com/bx6m [20:40] fengway: ah it's not commented out [20:41] you told me to do this right? # /dev/mapper/ubuntu--vg-root / [20:41] fengway: no, i meant for the last one, swap [20:41] damn [20:41] no worries [20:43] oops [20:43] was that why everything was screwed?= [20:44] fengway: let's hope so [20:44] xD [20:44] how do you edit it again back [20:47] fengway: sudo -H gedit /mnt/etc/fstab [20:48] do i remove # from the line i added it to EriC^^ ? [20:48] fengway: yeah [20:48] fengway: or "sudo sed -i '/swap/ s/.*/# &1/' /mnt/etc/fstab " [20:49] EriC^^ https://termbin.com/3na6 [20:49] TJ- is that for adding comment? [20:49] fengway: looks good but there's still a "#" in the /root line [20:50] EriC^^ https://termbin.com/enzn [20:50] fengway: yes, if you want to avoid running an editor [20:52] fengway: looks good, try rebooting [20:52] to normal ubuntu? [20:52] fengway: likewise for removing the stray # you could do something like "sudo sed -i '/#.*0.*1$/ s/^#//' /mnt/etc/fstab " === sveta_ is now known as Sveta [20:58] EriC^^ On my normal ubuntu now :) [20:58] great :) [20:58] what now? [20:59] time to install windows i guess? [21:00] EriC^^ When ubuntu was not logging on i tried to install it from the live usb and got error "no device drivers were found" [21:00] EriC^^ Maybe the usb is not correctly set up? [21:01] fengway: odd, never seen that one before, i wouldn't worry though [21:02] the live usb seems to be working pretty well [21:02] EriC^^ i mean the windows live usb, i can't install windows because of that error [21:04] fengway: first things first checksum the .iso [21:04] EriC^^ how? [21:04] is it windows 10? [21:05] fengway: seems like some people online had a similar problem and using a usb 2.0 port helped instead of 3.0 [21:06] EriC^^ yea, but when i put live usb on usb 2.0 my motherboard doesn't give me option to boot from it [21:06] fengway: is the windows cd actually booting? or it's like a bios message you're getting? [21:07] EriC^^ it's booting, i get that fancy windows menu and shit, i first choose my language also [21:08] fengway: where did you get the .iso from? [21:08] @er [21:08] EriC^^ https://www.microsoft.com/nb-no/software-download/windows10ISO [21:09] official [21:12] fengway: seems like this site has checksums https://tb.rg-adguard.net/public.php [21:13] put your info there and get the checksum [21:13] is the .iso file accessible from your ubuntu session right now? [21:13] yea [21:13] nice [21:14] not sure about all option on that site tho [21:14] fengway: i guess you picked the may 2019 version in the windows download site? [21:14] yea [21:15] what was the filename? [21:15] Win10_1903_V1_English_x64.iso ? [21:15] Win10_1903_V1_English_x64.iso [21:16] ok cool, the md5sum should be 8ba0e81b276d9052e8538deb0cf6c7d0 [21:16] in ubuntu type "md5sum /path/to/Win10_1903_V1_English_x64.iso" [21:16] Folks, I'm getting a permission denied error for a mount during "snap install lxd". Could it be related to the fact that /home on my system is a symbolic link? [21:17] EriC^^ 8ba0e81b276d9052e8538deb0cf6c7d0 Win10_1903_V1_English_x64.iso [21:17] fengway: looks good, so it's not a corrupted iso [21:18] fengway: you know, i think you could even boot the windows usb from grub, if it was really a usb port issue [21:18] The "snap install lxd" is run during "apt-get install lxd" [21:19] EriC^^ you mean could not*? [21:19] fengway: i think if it's some weird issue, you could ultimately use a vm like virtualbox, to boot the iso and install it to your actual harddisk from there (to your free space) and then boot it normally as a dual-boot [21:19] fengway: no you can [21:20] fengway: try to put it in the usb2.0 port, when you get grub, press "c" you'll get a command line [21:20] EriC^^ can i install on main hd from VM? [21:20] fengway: yeah you can, using virtualbox's rawdisk image trickery [21:21] i have vmware installed with w10 [21:21] fengway: can you boot into grub and talk here at the same time from a phone or something? [21:22] nevermind, it's really straightforward anyways [21:23] in grub after pressing 'c', type "ls" it should list all disks and partitions, you likely will see (hd1,gpt1) (hd1,gpt2) (hd1,gpt3) or similar (those are gpt partitions) [21:23] i think you'll see those, and another (hd2,gpt1) that has only 1 gpt, that should be your windows usb with its single fat32 [21:24] okey will try [21:24] fengway: you'll want to do "ls (hd2,gpt1)/" and see if it's windows usb, then do "chainloader (hd2,gpt1)/efi/boot/bootx64.efi" [21:24] okey [21:25] to boot the .efi file, i think that should be the name, either that or probably (hd2,gpt1)/efi/Microsoft/Boot/bootmgfw.efi" [21:25] /join #rantradio [21:26] you can use "ls (hd2,gpt1)/efi/" to browse around and stuff, once you find your file and do the chainloader command, just type "boot" and it should boot [21:27] will try it out now [21:40] EriC^^ I don't get it: https://imgur.com/a/tKU80ug [21:41] fengway: you have to add a "/" at the end or else it will just give the filesystem info [21:41] e.g "ls (hd0,gpt1)/" [21:41] i did it on the third, but how does that info help me [21:41] fengway: try "chainloader (hd0,gpt1)/efi/Boot/bootx64.efi" [21:42] then "boot" [21:42] okey [21:42] i think that should do it [21:47] EriC^^ same error: https://imgur.com/a/VBT5YBh [21:49] fengway: reading here https://www.drivethelife.com/windows-drivers/fix-media-driver-your-computer-needs-is-missing.html [21:54] EriC^^ i will try to replug it [22:04] EriC^^ no luck........................................#"¤"#%!"# [22:05] fengway: one of the ways has to work, it seems like a common problem [22:06] fengway: i had seen that some guy downloaded rufus, and used that to make the bootable usb and it worked for him [22:06] rufus is on linux? [22:07] i think on both windows and linux [22:07] fengway: did you try the method 3 stuff, usb3 bios settings? [22:07] fengway: maybe windows wants ahci instead of raid or something too [22:09] EriC^^ not yet, but one tragic thing is that when i was trying out switching usb to other ports i tried to switch my printer to usb 3.0 port because last time i used it (over a year ago) i remebered that i had to connect it to this usb port. Now it seems that VMWare is getting contact with it. But my VMWare got corrupted somehow so trying to [22:09] reinstall it [22:10] ah [22:11] EriC^^ you have been amazing tho, and i don't know if vmware will work anyway. And i am so tired today, will try out vmware first if it doesn't work then more on w10 tomorrow. Thanks so much for the help man, you have been on this for hours [22:12] i have to be on work in 8 hours and need to get some sleep [22:12] fengway: no problem man, have a good night [22:15] fengway: you did a lot for today, and you're pretty close, get some rest and get at it later and you should be good i think, you're almost there with it, anyways see ya [22:19] EriC^^ Yea we should get there, should i have my swap on now? [22:19] fengway: no dont enable it yet [22:20] is it memory swap? i don't think vmware can work without it? [22:21] fengway: you could create a swap file real quick if you want [22:21] fengway: type "sudo fallocate -l 1G /swapfile" [22:22] fengway: "sudo chmod 600 /swapfile && sudo mkswap /swapfile" [22:23] fengway: then type echo "/swapfile swap swap defaults 0 0" | sudo tee -a /etc/fstab [22:23] EriC^^ /swapfile swap swap defaults 0 0 [22:24] actually "/swapfile swap swap sw 0 0" [22:24] fengway: ^ [22:25] EriC^^ sudo: /swapfile: command not found [22:26] fengway: echo "/swapfile swap swap sw 0 0" | sudo tee -a /etc/fstab [22:26] EriC^^ /swapfile swap swap sw 0 0 [22:28] fengway: you have to put "echo" before it [22:28] EriC^^ i did [22:29] echo "/swapfile swap swap sw 0 0" | sudo tee -a /etc/fstab/swapfile swap swap sw 0 0 [22:30] fengway: type "cat /etc/fstab | nc termbin.com 9999" [22:30] EriC^^ https://termbin.com/ezxu [22:31] fengway: type "sudo -H gedit /etc/fstab" and delete the extra one that has defaults in it [22:31] delete this one /swapfile swap swap defaults 0 0 [22:32] EriC^^ https://termbin.com/59yl [22:32] looks good fengway [22:33] "sudo swapon -a && swapon -s" [22:33] it should show up there if it's all good [22:33] sudo swapon -a && swapon -s [22:33] Filename Type Size Used Priority/swapfile file 1048572 0 -2 [22:34] all good [22:34] cool [22:56] EriC^^ it showed up on vmware bcs it's on other port now, i'm so dumb...... [22:56] fengway: ah great [22:57] EriC^^ will try out other software for label and see if it works. If everything works, can you help to make ubuntu partition take whole HD again lol? Maybe like tomorrow or something [22:59] fengway: can't promise anything, if i'm on then sure [22:59] yea of course, thanks again dude, it's amazing that you kept on for so long [23:00] is my swap normal now? EriC^^ [23:00] 1 GB is enough? [23:00] np, yeah it should be good [23:00] one gig swap is enough for most uses, yeah [23:00] cool [23:01] brain farting this hard hurts physically [23:01] initially you had 1gb too, so it should be good i guess [23:01] ah nice [23:03] fengway: did you give the bios settings stuff a shot from that guide? the method 3 one? [23:03] it seemed like it had potential [23:03] EriC^^ i didn't, is it about having usb 3 on auto or something? I think should maybe also disable fast boot, but can't find settings for it [23:04] yea i think actually that may be it [23:04] i can try out and see if it works even if vmware work if you wanna know, but if vmware work then i will just try it and cancel it if it works :) [23:06] nah it's ok, thought you wanted to get windows on there long run or maybe other label software is hard to come by on ubuntu [23:06] can i install on ubuntu lts 18.04 the newest kernel or will the great dependency issues? [23:07] the label software send info to printer, i just want to see if all that works on vmware. If it does then I don't want windows seperately, because it's easier to just turn on vmware than reebot [23:07] !mainline | seven-eleven [23:07] seven-eleven: The kernel team supply continuous mainline kernel builds which can be useful for tracking down issues or testing recent changes in the Linux kernel. More information is available at https://wiki.ubuntu.com/Kernel/MainlineBuilds [23:07] problem was before i couldn't get vmware to recognize printer, but it's wierd how it doesn't work at all on usb 2 but does on usb3 [23:08] seven-eleven: those kernel images are not supported here (other than for tracking down issues) though [23:08] fengway: yeah [23:09] tomreyn, gotcha [23:09] EriC^^ good night, bro. I will try to get few hours of sleep now. Thanks again a lot! [23:10] tomreyn, i assume i wont have dependecy issue, rather software might stop working probably because of bugs [23:11] !hwe | seven-eleven: not usally, no. what Is supported, though, is this [23:11] seven-eleven: not usally, no. what Is supported, though, is this: The Ubuntu LTS enablement stacks provide newer kernel and X support for existing LTS releases, see https://wiki.ubuntu.com/Kernel/LTSEnablementStack [23:11] one could install !hwe over LTS [23:11] hwe is enabled by default on a fresh 18.04.3 [23:15] tomreyn, awesome, thank you ! [23:16] you're welcome! [23:31] Hey guys, I was following some instructions to set up a letsencrypt config file and it had this line to be placed in cli.ini file: config-dir=~/.certbot/config [23:31] But this seems to create a folder named '~' inside my /var/www [23:33] it's quite possible whoever wrote the instructions never tested them [23:36] @sarnold Oh I think I know what happened. I run certbot renew cron as www-data user and IO guess /var/www is the home directory? [23:36] So I probably needed to have something like config-dir=/var/www/.certbot/config [23:37] yeah that will probably do the trick. ~ is something *programs* need to handle if they are going to use it to mean "whatever is in $HOME" or "whatever getent says is this user's home directory" .. [23:38] .. so most programs won't actually do anything about it, and *maybe* they'll have the shell do those things for them, but that's complicated, because that usually introduces security or reliability problems [23:40] @sarnold, thanks. I should probably look a little more closely before I just implement stuff I see online [23:40] destinydriven: it certainly never hurts to give things a mental dry run before doing things at the console :) [23:41] destinydriven: but if my hunch on this is right, well.. that comes from experience and seeing tooooons of software, some better than others, hehe :) [23:41] Yup, but to be fair this was from a Certbot engineer from the official certbot community forum [23:42] hmm.. that sounds like a pretty good starting point :) [23:43] yup === mandude is now known as EmergencyMedical [23:49] hey everyone [23:49] wb gimpnixon [23:49] I was here with a question the other day about 19.10. I was directed to #ubuntu+1 but no one has spoken a word in there for three days. [23:49] ow [23:50] So i was wondering if maybe i could mention my question again and see if mayeb someone has popped in that wasnt in the other day and maybe has an idea [23:51] sarnold, hey sarnold, how has everything been? [23:52] gimpnixon: not bad not bad [23:53] gimpnixon: well it's fine by me, dunno the others, hehe :) [23:53] sarnold, i still havent got anywhere with my issue, but i have really enjoyed the recommends thing we talked about. it's nice not bringing in so much or at least having a choice now about what i bring in on most installs [23:53] if the other channel's too idle to be useful then it feels useful to try again here [23:55] OK I'll shout it out and see if anyone bites