[00:27] I'm running Ubuntu 20.04... Currently have JACK/PulseAudio setup working (and pretty well). However I'm thinking about moving to PipeWire so I can route PA apps and such like I do the JACK stuff. Any thoughts or comments? === scoobydoo_ is now known as scoobydoo === c10l0 is now known as c10l [01:00] hey all.. anyone here ever run ubuntu on macbook air and have success? I have a 2015 Air and couldn't get it to work. [01:03] live usb works but i couldn't seem to get it to install for some reason.. didn't want to touch onboard nvme drive so tried to install from liveusb to 512gb microsd in the SD slot. don't remember all the details on what happened. just trying to see if anyone else has done it successfully. I installed it on my Windows desktop machine on an unused drive and dual boot that just fine. [01:06] I did this on the Air with Linux Mint Cinnamon 20.2 just fine as well. === M4he is now known as mahe === keypushe- is now known as keypusher === cacti is now known as madmax === Pony is now known as Guest1413 === diskin is now known as Guest2895 === diskin_ is now known as diskin [06:05] vps-e42d72d2:14433 [06:31] hi all, looking to try the latest ubuntu on a macbook pro mid 2018 i've disabled the t2 chip security and am allowed to boot from flash drive. is there a good way to write thei so to a flash drive? [06:41] Hey QQ: I want to add an address to an network interface, does netplan apply interrupt traffic on that interface? [07:05] Disconsented: i've ran netplan apply over ssh without losing the connection [07:06] but you can use 'ip' to add an address [07:14] why is there no rar2fs package for ubuntu? [07:15] cheers ducasse [07:18] i'd love to have it on my media server but the install is a bit involved and i'd have to remember to maintain it [07:19] spagoote: what is rar2fs supposed to do? simply because there's no rar2fs at packages.debian.org/rar2fs [07:19] spagoote: this? https://github.com/hasse69/rar2fs [07:19] tarzeau_: yes [07:19] why do you still have rar archives? you get them or you create them? [07:19] i get them [07:22] spagoote: tried to build it, needs unrar source, which is non-free [07:22] spagoote: why don't you convert all your .rar to something that is free, and with better compression? [07:23] or even better use a live copressed filesystem like btrfs? [07:24] tarzeau_: that would be something i'd have to automate. rar2fs seems transparent in my use case [07:25] spagoote: should be easy to automated, i don't have a better solution sorry (except you create yourself a .deb, i started at sid.ethz.ch/debian/rar2fs) [07:33] tarzeau_: thanks! i'll look into it [07:45] hello [07:45] i am trying to improve my ubuntu skills/linux administrator skills [07:46] i made a shared folder, installed a web server, created a user, gave permissions to files/directories with chown, installed openssh [07:46] any ideas of what else i can do? [08:04] wwwi: master ACLs, learn how NFS works and the pitfalls of independent systems using NFS shares :) [08:06] wwwi: set up a virtual machine with libvirt/kvm [08:38] hello [08:38] ? [09:04] TJ-, and ducasse thank you [09:04] TJ-, what are the ACLs and the NFS? [09:04] ducasse, why set up a virtual machine with kvm instead of oracle's virtual box? [09:05] wwwi: ACL == Access Control Lists, see "man acl" after "apt install acl" or http://manpages.ubuntu.com/manpages/focal/en/man5/acl.5.html [09:06] TJ-, I see [09:06] wwwi: Virtualbox uses out of tree kernel modules that do things weirdly in the network layer; Qemu/KVM uses the kernels own facilities, properly [09:06] TJ-, I see. [09:06] wwwi: there is a management tool for GUI "virt-manager" to make working with virtual machines easy [09:10] Greetings [09:11] I would need help with my ubuntu laptop 20.04 zsys enabled , it doesn't boot anymore , booting from the 20.04 livecd, zpool import shows rpool and bpool [09:12] is it possible that grub messed things up ? [09:12] is it better to boot from a 21.04 (more recent zpool version ?) [09:13] i never made a zpool upgrade by myself fyi [09:14] would you have any idea how to check and fix if this is a grub issue ? [10:52] Which program would i use when i wanted to feed keystrokes to an interactive program in a script? Something like "yes", but with custom strings. [10:53] faekjarz: you can use just about anything, if the program reads standard input. [10:53] faekjarz: e.g. printf '%sn\' 'string one' 'string two' y Y n wq | cmd [10:54] old fashioned way is using xdotool [10:54] !info xdotool [10:54] xdotool (1:3.20160805.1-4, hirsute): simulate (generate) X11 keyboard/mouse input events. In component universe, is extra. Built by xdotool. Size 41 kB / 122 kB [10:57] hmm, ok, thank you - i need to feed cryptsetup with a list of password permutations, because i can't remember exactly what it was *sigh* quite annoying, but i'm the only one to blame === fireat is now known as el-x-cutie [11:09] @faekjarz, looking at the man-pages for cryptsetup I found this [11:09] --key-file, -d name [11:09] Read the passphrase from file. [11:10] If the name given is "-", then the passphrase will be read from stdin. In this case, reading will not stop at newline characters. [11:10] so you do not have to deal with an interactive prompt but can feed the permutations in directly. [11:18] Hi, my hardware has been replaced on 20.04, I'm in the rescue system and need to change the file containing the MAC address. Where is that file? [11:21] crst: can you elaborate? What file are you searching for? I don't know of anything that would block your booting based on the network card mac address... === diskin is now known as Guest8690 === diskin_ is now known as diskin [11:23] alkisg: not sure either, support told me: Due to a hardware fault, we have replaced your server while taking over the hard drives. Your server has been booted into our rescue system. Please modify your system configuration to accommodate the new MAC address. [11:23] When I reboot I can't ssh into the server anymore [11:24] OK I guess then you have some form of network connection that is based to the mac address [11:24] You don't need to boot it into a system rescue cd, a normal booting (via local console, not ssh) should also do, [11:25] but if this is a VPS, it's probably using netplan, which I'm totally ignorant about [11:26] Check in /etc/netplan, and wait for other's feedback [11:27] alkisg: thanks, yeah there's no MAC entry in /etc/netplan, the provider is Hetzner and it's a dedicated server if that helps. [11:28] Other places to look for are /etc/network/interfaces or interfaces.d, and /etc/NetworkManager/system-connections [11:28] (for ifupdown and network-manager) [11:29] server? /etc/systemd/network/ [11:30] crst: are you sure there is such a hard-coded entry? Have you actually triedto reboot the server normally? [11:31] TJ-: I just know that I can't ssh into the machine when booting, and have the statement from support as above. [11:31] crst: what name/address are you using to connect to it with ssh? [11:33] crst: you need to connect to the server's remote console to watch and oversee the boot process [11:40] TJ-: I don't think that's possible here to watch that unfortunately, and it looks like the server doesn't boot properly. I can ping it but nmap returns "Host seems down" [11:40] Either it doesn't boot or there's a problem with the network connection [11:40] crst: no remote console!?!? move hosts! [11:41] It's a dedicated server, not a vm [11:43] So? even more reason for KVM/iLo/iDRAC - all ours have it [11:43] how else are you supposed to work on a broken headless remote server without some kind of out-of-band access ? [11:46] Sorry there is an option https://docs.hetzner.com/robot/dedicated-server/virtualization/vkvm/ that's news to me [11:47] crst: that's what you need! :) [11:47] crst: actually no, that isn't what I meant! [11:50] crst: Under "Maintainance" at https://docs.hetzner.com/robot/dedicated-server/ [11:51] crst: looks like they have a few options there [11:51] crst: looks like you need https://docs.hetzner.com/robot/dedicated-server/maintainance/kvm-console [11:57] TJ-: awesome, just ordered one [11:59] crst: looks like you have 3 hours to fix it if it is broken :) [12:00] thye dont always turn it off right away [12:00] and they also have some livecds you can use to boot and log with a ssh key [12:07] helu :) [12:08] vuurdraak: if you have a ubuntu support question just ask, casual chit chat is in #ubuntu-offtopic. Thanks [12:08] sorry my apoligies , i was just redirected from freenode to here, as i didnt know everybody went here [12:13] TJ-: This is what I get https://ibb.co/L5p48CC [12:15] crst: hardware failure [12:16] crst: if they changed the hardware out from under you, it's their issue [12:17] Looks like it... told support, thanks so much for your help in this everybody! I keep you updated. They replaced my "allegedly broken" hardware with a "faulty one" it seems. [12:19] why would someone create a virtual machine? [12:33] wwwi, because we can? [12:39] @ wwwi , you realy want one if you want scammers to connect to your machine and let them hack you like Kitboga does [12:45] :) [13:04] /j #zfsonlinux [13:04] oops [13:05] what is the nfs? I googled it, but the help there didn't help me [13:06] interesting, bing shows me a lot [13:06] www Network File System [13:06] !nfs [13:06] nfs is the network file system. See https://help.ubuntu.com/community/SettingUpNFSHowTo for information on installing and configuring NFS. [13:08] TJ-: "As requested, we have briefly checked your drives and cannot find a critical issue. Furthermore the BIOS is already up-2-date." They say all is ok, lol [13:12] crst: did you show them the screenshot with the EDAC report? [13:12] Sure I did [13:12] They boot into rescue and say all is ok [13:13] crst: well, as it broke after they changed hardware not sure what else you can do. what kernel version is Rescue compared to the installed ? [13:14] TJ-: Not really sure I'll try to note that next time when in rescue, it's supposed to be a default Linux-64 whatever that means. [13:16] What hardware could be broken according to the screenshot? [13:16] Right now the output it just a never ending VVVVVVVVVVVV... [13:18] crst: see https://bugzilla.kernel.org/show_bug.cgi?id=210593 [13:18] bugzilla.kernel.org bug 210593 in Platform Specific/Hardware "EDAC amd64: Error: F0 not found, device 0x1650 (broken BIOS?) with Ryzen 5600x & B450 chipset" [Normal, Resolved: Patch_already_available] [13:23] Hi folks [13:26] Could I have any fault in this because I use the standard latest kernel (nothing custom just regular apt upgrade)? I'd say no. [13:27] crst: check which kernel versions are being used between rescue and installed, as I said [13:28] Will do, where can I find the kernel version of the installed? [13:28] crst: well, from rescue you can acess the installed system and check what's in /boot/ [13:32] TJ-: rescue: Linux rescue 5.13.13 #1 SMP Thu Sep 2 05:38:34 UTC 2021 x86_64 GNU/Linux [13:32] installed: looks like config-5.4.0-84-generic [13:34] crst: right, so it's possible the Ubuntu kernel doesn' contain the EDAC patch [13:35] crst: and they've changed the mobo to one that triggers it which didn't happen on the earlier mobo/CPU [13:36] crst: best thing to do would be to use a rescue chroot to upgrade the Ubuntu kernel to one of the HWE releases [13:36] thank you oerheks [13:36] crst: that'd get you a v5.11 kernel on Ubuntu 20.04/Focal [13:38] Just today before that happened I've ran `apt update` and `apt upgrade`, so nothing should be outdated in theory. Do I need to customize the kernel version now? Shouldn't the provider support that version? [13:38] crst: yes, you'll need the HWE versions of the kernel [13:38] !hwe | crst [13:38] crst: The Ubuntu LTS enablement stacks provide newer kernel and X support for existing LTS releases, see https://wiki.ubuntu.com/Kernel/LTSEnablementStack [13:40] crst: e.g. linux-generic-hwe-20.04 [13:40] !info linux-generic-hwe-20.04 [13:40] linux-generic-hwe-20.04 (5.11.0.34.36, hirsute): Complete Generic Linux kernel and headers. In component main, is optional. Built by linux-meta. Size 2 kB / 18 kB. (Only available for amd64, armhf, arm64, powerpc, ppc64el, s390x.) [13:42] 5.13.13 comes from the mainline ppa ? [13:42] so it is done manually. [13:49] I've never upgraded a kernel via rescue or so, can you help me with that please? [13:52] interesting, I am still on the GA kernel, I guess due to having upgraded onward from 14.04, I am wondering if I should also install HWE stack and/or if this might break things already installed, I think I had a problem once going to HWE but I can not remember what it was [13:52] if you run 5.13, you surely upgraded to a testing kernel. [13:57] oerheks: 5.13 is the rescue env provided by Hetzner [13:57] crst: do you have a shell in the rescue environment ? [13:58] yes [13:58] crst: does it have Internet access so we can collect some logs? [14:00] TJ-: It does have internet access and I've mounted the disk containing the installation already. [14:01] crst: great; show us this: " (lsblk; blkid) | nc termbin.com 9999 " [14:02] https://termbin.com/7rje [14:03] crst: OK, just to match the pattern used by the installer, and to humour me, can we change the mount-point to something else ? as in "umount /mnt; mkdir /target; mount /dev/md0 /target" [14:04] crst: My fingers are used to typing /target :) [14:04] Sure thing https://termbin.com/thgs [14:06] crst: OK, lets motor! Are you able to copy/paste the commands I give here into it? [14:06] Sure [14:06] crst: even better :) [14:08] crst: it's not clear whether that system boots using UEFI or BIOS/Legacy [14:08] crst: I don't see any sign of a UEFI system partition, and p2 looks like it may be a BIOS Boot partition [14:09] crst: show us " cat /target/etc/fstab | nc termbin.com 9999 " [14:09] https://termbin.com/1jz54 [14:10] I also think it doesn't use UEFI, can't say for sure though [14:11] crst: looks unlikely. Let's get things prepared. I'll assume you're 'root' [14:11] correct [14:12] crst: " mount -t proc proc /target/proc; mount -t devtmpfs udev /target/dev; mount -t devpts devpts /target/dev/pts; mount -t sysfs sysfs /target/sys " [14:12] done [14:13] crst: then " chroot /target apt list --installed 'linux-image*' | nc termbin.com 9999 " [14:13] crst: this is to check what kernel packages are currently installed [14:13] Afternoon all. I'm having some trouble with the HP servers we have in that the SD card reader is enumerated as /dev/sda, then the RAID array as /dev/sdb. The HP documentation suggests adding `rd.driver.pre=smartpqi` to the kernel command line at boot time to ensure the RAID array is enumerated first, and this works fine for our RHEL installations, however that's a dracut option which isn't going to work for Ubuntu. Does anyone know [14:13] what the equivalent option to `rd.driver.pre` is for Ubuntu? `rd.modules_load` looks promising however I'm not sure this will load the driver early enough. [14:13] TJ-: https://termbin.com/ur1he [14:14] crst: let's install HWE: " chroot /target apt install linux-generic-hwe-20.04 |& tee /tmp/fix.log " [14:15] crst: this will capture the log /tmp/fix.log. When the command ends, paste the log with "cat /tmp/fix.log | nc termbin.com 9999 " [14:16] https://termbin.com/v6i8 [14:16] crst: oh dear, we have name resolution problems (DNS) [14:17] Yeah it's using the default ones from Hetzner, let me check if I can change the DNS in the rescue [14:18] crst: great; we'll need it [14:18] check current with "resolvectl status" [14:32] No success changing the DNS here, any ideas? Changing /etc/resolv.conf [14:33] doesn't work [14:34] crst, late to the party what is your problem again? [14:35] crst: can you show "resolvectl status" [14:35] crst: hopefully the rescue is using systemd-resolved [14:36] https://termbin.com/jo5g [14:36] but you need to restart it afair [14:36] I normally edit those on my Hetzner machines but this is rescue... [14:36] crst: can the system reach the primary? "ping 185.12.64.1" [14:37] yes: 6 packets transmitted, 6 received, 0% packet loss, time 5008ms [14:37] crst, systemctl restart systemd-resolved [14:37] crst: you shouldn't directly alter /etc/resolv.conf; that should be a sym-mlink [14:37] I know, not sure what else to do [14:37] crst: that's good - so does this work: "resolvectl query de.archive.ubuntu.com" [14:38] https://termbin.com/k2q8 [14:39] crst: looks t be working right now, lets rush! [14:39] crst: " chroot /target apt install linux-generic-hwe-20.04 |& tee /tmp/fix.log " [14:40] sucks https://termbin.com/r06j [14:40] I hate Hetzner so much right now... [14:40] crst, can you ping google.com ? [14:40] yes [14:41] crst: ha! I can understand that [14:41] It's not the first time their DNS made me trouble [14:41] So bad [14:41] crst, are you in chroot env ? [14:41] I'm in rescue as root [14:42] crst, cat /proc/cmdline [14:42] crst: does this work? "ping 141.30.62.25" [14:42] crst: we can manually add an /etc/hosts record to work around this if so [14:42] It does work, yes [14:43] crst: hmmm, before we do that, I wonder if this is just the target. what does "ls -l /target/etc/resolv.conf" report? [14:44] crst: if that is a symlink it'll be broken if the /run/ tmpfs isn't in place, as is to be expected in a chroot [14:44] lrwxrwxrwx 1 root root 39 Apr 22 17:58 /target/etc/resolv.conf -> ../run/systemd/resolve/stub-resolv.conf [14:44] crst: aha! fix: "mount --bind /etc/resolv.conf /target/etc/resolv.conf" [14:45] mount: /target/etc/resolv.conf: mount point is a symbolic link to nowhere. [14:45] crst: ahh! [14:45] crst: we need to put this back later, so don't forget. "rm /target/etc/resolv.conf" [14:45] Will remember, thanks [14:45] cant you just bind mouont /ron ? [14:46] *run [14:46] crst: "mount --bind /run/systemd/resolve/stub-resolv.conf /target/etc/resolv.conf! [14:46] or that 🙂 [14:46] ogra: it's not helpful since it mixes the rescue env with the chroot's so services/processes start dropping their files in it [14:46] mount: /target/etc/resolv.conf: mount point is a symbolic link to nowhere. [14:47] crst: errr, didn't you just delete it!? [14:47] Oh ok, sorry. Now we get: mount: /target/etc/resolv.conf: mount point does not exist. [14:48] crst:oh, true! "touch /target/etc/resolv.conf" then redo the command [14:48] touch it before bind mounting [14:48] TJ-: done :) [14:48] crst: now try again! [14:49] crst: " chroot /target apt install linux-generic-hwe-20.04 |& tee /tmp/fix.log " [14:49] https://termbin.com/nbjk [14:50] crst: finally! now we undo what we did and you try a reboot. [14:50] crst: "umount /target/etc/resovl.conf; rm /target/etc/resolv.conf" [14:51] crst: then "ln -s ../run/systemd/resolve/stub-resolv.conf /target/etc/resolv.conf" [14:51] crst: errr, typo in my first line [14:51] crst: "umount /target/etc/resolv.conf; rm /target/etc/resolv.conf" [14:51] done [14:52] crst: check it looks correct with "ls -l /target/etc/resolv.con " [14:52] oh typos! [14:52] crst: check it looks correct with "ls -l /target/etc/resolv.conf " [14:52] lrwxrwxrwx 1 root root 39 Sep 20 16:51 /target/etc/resolv.conf -> ../run/systemd/resolve/stub-resolv.conf [14:53] crst: good - reboot [14:54] crst: you'll likely want to watch that KVM [14:55] vm doesn't work anymore, but man: I'M BACK IN!!!! [14:56] Linux fondi 5.11.0-34-generic #36~20.04.1-Ubuntu SMP Fri Aug 27 08:06:32 UTC 2021 x86_64 x86_64 x86_64 GNU/Linux [14:56] crst: phew! [14:56] TJ-: Outstanding job man, thank you so much! [15:00] !cookie | TJ- [15:00] TJ-: Wow! You're such a great helper, you deserve a cookie! [15:03] He's better than Hetzner :D === genii-core is now known as genii [15:16] huh, think thats the first time I've seen a 2. ip [15:44] is there a way to force emacs to 27 under ubuntu 18.04 LTS? [15:47] I'm having some strange update issues. Does anyone know why I'm getting "cannot find EFI directory" (and more) here? https://bpa.st/raw/OCIQ [15:47] I don't have a /sys/firmware/efi dir, so I don't think I even have EFI. It is unclear to me why grub is having issues. [15:48] yates, use the snap [15:48] yates, snap ver is 27.2 [15:48] snap info emacs|grep stable [15:48] latest/stable: 27.2 2021-06-07 (1161) 183MB classic [15:52] ogra: ack - thanks! [16:25] Can someone help with a zfs root issue on 21.04? https://pastebin.ubuntu.com/p/Tm7BKZh4xN/ I rebooted today and was unable to mount root. Recovery grub option didn't work. It can't seem to find any of my mounts. The pastebin link is a view from a USB Ubuntu instance where I imported the pool. Is there a way to get my volumes back? === Guest51 is now known as ppetraki [16:30] ha, again damn touchpad not working and I'm not sure even how I did switch it off with some Fkey [16:31] vimart: you might wanna share your dmesg in a paste with the volunteers so they can take a look for you [16:32] I'm booted from a USB disk now, there's nothing in dmesg of consequence [16:33] Single physical disk. nvme, default install of 21.04 [16:33] worked fine until it didn't [16:36] zpool history is interesting, https://pastebin.ubuntu.com/p/SNf6yBfZ33/ , a lot of that looks familiar, I just need to expose those volumes again. [16:38] lotuspsychje: actually had to hit one of the F-keys that switched off the TouchPad during cleaning then when activated it back the touchpad was a bit wet yet and not responding nice. [16:42] Hi all, is there any easy way to determine why Gnome-shell is starting to use 100% of one of my CPU-cores if I leave my computer one for a few days? [16:44] This is a reoccuring issue, killing gnome-shell and letting it restart works-around the issue temporarily, but it's not really good way to deal with it [16:44] TJ-: Could it be that this particular kernel messes up my microk8s? https://github.com/ubuntu/microk8s/issues/728 [16:45] Issue 728 in ubuntu/microk8s "Microk8s cluster nodes stuck in NotReady state on ARM64" [Closed] [16:46] There's something going on with the storage and I'm not sure what's up [16:47] crst: wich kernel were you on? [16:48] Was on 5.4 now on 5.11 [16:48] crst: we currently notice a lot of 5.11 bugs, so def worth a shot to different kernel test [16:49] lotuspsychje: Which kernel would you recommend? [16:50] crst: well what we usualy try is previous kernel(s) from your current list for testing purposes, or !mainline kernels to go higher [16:53] Hmmm, running some simple zsysctl show rpool/ROOT/ubuntu_tz6r79 appears to have prompted zfs to remember it's a storage array. zfs list is actually returning something useful now. https://pastebin.ubuntu.com/p/cjY5V84fTF/ [16:54] * ppetraki reboots to try [16:54] who can help me https://paste.debian.net/1212533 [16:55] cws6969: cat /etc/os-release | nc termbin.com 9999 [16:55] thought so [16:58] I'm still a bit dissapointed Canoncial decided to Require login for ther pastebin-server nowadays [16:59] hello, I am using ubuntu 20.04 on a vps, and I have the problem that every day or two, the eth0 network interface seems to go down. A reboot fixes it, but I can't find anything in the logs. [16:59] zteam: the pastebininit command doesn't require it [17:00] But then you are just shifting the login req to the reader.. [17:00] zteam: so silly [17:02] leftyfb, no, that's correct, but you can't see the logfiles, without having an account, nowdays, at least that's what it tells me everytime, it kinda defeats the whole purpose, Should we really need to create an account to JaTop Shaving Head look at other logfiles...? [17:04] that's only making less people able to help people out here [17:08] mybalzitch, yep, it truly is, requring users to create an account to upload logs, is one thing, but requring an account to look at files, uploaded by others, thats beyond being silly :) [17:08] zteam, tell the people that abused it en-masse to post credit card details and other DMCA relevant bits ... [17:11] canonical simply could not cope with the DMCA teardown requests anymore (you have to check each single one, keep logs with IPs of submitters and a ton of other legal requirements) [17:14] Is there any good reason why I should *not* make a cronjob run every 5 minutes "ip link set eth0 up"? [17:15] why would you need to [17:15] my eth0 drops out every day or two [17:15] get a better eth0 [17:15] ogra, well, first off, I didn't know that was the reason, second of all, I don't really see requring an account to view those files as a good solution either, with that being said, it at least explains why they decided to do what they did :) [17:15] i don't know why, and this seems like an easy solution [17:16] webchat78, if you really want to do that i'd rather write a script that pings the default GW and only if it cant reach it, only then call the ip link up command [17:16] webchat78, you should propably hunt dmesg for clues instead, something like dmesg -wH grep eth0 [17:17] webchat78, but eth0, is the old name, newer versions of Ubuntu renames the network interface, nowadays [17:17] webchat78, and ask your cleaning person to not pull the plug of the switch to put in the vaccum cleaner when cleaning the server room 😉 [17:20] So any easy way to diagnose why my gnome-shell likes consuming 100 of 1 of my cpu cores ? (can't find anything useful in syslog [17:21] ogra haha it's a vps. I'm sure it's something else I did, and so far I haven't seen anything in the logs. Could you elaborate on 'a script that pings the defaultGW' [17:21] zteam, broken extension ? [17:24] ogra, well I remember correctly I did try to disable all of them (I'm not 100 sure I did disable all of them, though I looked into this about 1,5 months ago [17:25] zteam, the other idea might be fractional scaling ... do you have it in use ? [17:25] zteam kernel: virtio_net virtio1 enp1s0: renamed from eth0 [17:25] Sep 18 19:18:38.441619 j2buntu kernel: virtio_net virtio1 eth0: renamed from enp1s0 [17:25] This doesn't seem to be related to the event. [17:29] webchat78, something like: ping -c1 $(ip route show to default | sed 's/^.*via //;s/ dev.*$//') >/dev/null 2>&1 || ip link set eth0 up [17:30] webchat78, that will only run the ip command if it can not ping the default GW [17:31] and put that in crontab to run every 5 min? [17:31] (if you do such hacks, only do them with some checks 😉 ) [17:31] ogra, nope I don't. I did upgrade from Ubuntu 20.10 to 21, 21.04, but unfornetly the upgrade failed, so I let apt and dpkg attempt to fix the issues, and it did, so maybe it's because óf that [17:32] ah, yeah, you might have left some old stuff around [17:33] webchat78, since you already know the kernel decided to rename your network card, you should try digging dmesg for clues about enp1s0 as well, if you haven't already tried that === dongcarl1 is now known as dongcarl [17:34] zteam it's only the same messages [17:35] webchat78, try dmesg -wH without grepping then :-) [17:36] webchat78, do note that dmesg -wH only lists new errors happening after you run the command, it doesn't look for what happenend earlier [17:36] zteam i've looked at all the dmesg, and from several boots back in time. I haven't had anything jump out at me [17:37] you should also check your journal ... dmesg is kernel messages only ... if anything in userspace tinkers with it it will be in the journal [17:38] ogra, yep, I rather avoid reinstalling thought [17:41] ogra journalctl -xe seems to show only events from today/last boot also. [17:42] if you could explain the script you wrote above, i would appreciate it :) [17:42] you can enable presistent logging by creating /var/log/journal (though normally this does exist on ubuntu installs) [17:43] yes, i have it apparently [17:43] the script above calls "ping -c1 " ... the || means if the first command fails, call the second command ... the second command is your ip link up [17:44] ... if the first command does not fail the second will not be run [17:44] ogra ok. And i can crontab that to run every 5 minutes? And why is it better than just ip link up? [17:44] because it will only take action if action is needed [17:45] won't ip link up only do anything if the link is down? [17:45] instead of blindly calling ip link up every 5 min [17:45] it simply adds a check if the command is really necessary [17:46] ok. i guess what i'm asking is if one uses less resources or something. [17:46] though this is really just an awful hack ... like having a timed wall plug to reboot your windows server once a night (i have seen such things in "professional" datacenters !!) [17:47] the real solution is indeed to find the cause and fix that 😉 [17:47] ogra. haha yes, i would prefer to solve the root issue, but it seems like it will take some time. in the meantime, i need this server connected [17:48] Here is a syslog, of my gnome-shell issue with gnome eating 100 cpu if anyone feels borted https://pastebin.com/w8tfeVtR [17:48] bored * [17:49] lots of gjs errors you have there [17:50] ogra, actually one of former employes used to do just that, with some Webserver they had running (I had no part of it) :D [17:50] zteam: and a lot of nvidia connects/disconnects [17:52] ogra, what's gjs short for? [17:52] zteam: a full dmesg could be usefull too [17:52] gnome javascript ? not sure [17:52] lotuspsychje, coming right up :-) [17:53] zteam: got a lot of gnome extensions? [17:53] zteam, apt show gjs [17:56] lotuspsychje, dmesg is herer https://pastebin.com/GUhb8Ls9 [17:57] ogra: based on `apt-cache search`... [17:57] gjs - Mozilla-based javascript bindings for the GNOME platform [17:57] right [17:57] pretty much the core of gnome-shell (given the whole desktop is javascript) [17:58] yup [17:58] lotuspsychje, about 7 gnome extenesion installed, I believe I did try to disable all of them (although I might have leaven one or two on) [17:59] zteam: the first things the devs will ask to debug is remove the gnome extensions [18:01] zteam: did you try a fresh user yet, to see if gnome acts up there? [18:02] ogra, apt show gjs https://pastebin.com/z0KmCkN1 [18:02] yep [18:05] lotuspsychje, yes, obvisiosly, as I mentioned I believe I did disable most of them (Ubuntu ships with a few extras enabled out of the box, those I think I leaved on. [18:05] zteam: other ideas crossing my mind, is testing other kernel versions and a lower nvidia version, or a 470 reinstall [18:06] a lot of tests you can try [18:07] lotuspsychje, yes, I was running an older version of the Nvidia driver, but I upgraded it to see if that would fix the issue [18:08] zteam: since when did you get higher cpu? [18:10] lotuspsychje, as of trying gnome-shell as a different user, thats a very good idea, I should definetly try that :-) [18:16] lotuspsychje, I was upgrading from a healthy install of Ubuntu 20-10 to 21-04 but the upgraded crashed at the end of the upgrade, so I ran sudo dkpg --configure -a and sudo apt install -f to fix broken dependencies, and after running a couple sudo apt full-upgrade I got myself a working system, note that gnome-shell behaves normally for a day or so, only if I leave it on for a few days, it's starting to use 100% of one core not all [18:16] four) [18:17] i see, a bit of non-lts fun :p [18:19] lotuspsychje, well yes, but I needed a more recent version of Kodi :P (and I had my reasons to avoid there ppa, in this case) === MrA is now known as Ders === genii is now known as genii-away === diskin_ is now known as diskin [18:55] anyone know why ubuntu is saying I have a manually installed driver in jockey? [18:58] jockey is an old application, removed from 20.10 IIRC [18:58] ubuntu-drivers-common is current [18:59] *if* you have nvidia/wireless drivers installed by the 'drivers' tab, so the message is true [19:00] jockey isn't available in any supported version of ubuntu from what I can tell [19:00] indeed. [19:00] err i made a mistake; removed from 12.10* [19:01] lol [19:11] didn't they just merge jockey into software sources [19:12] because that is what i am talking about [19:12] looks identical if they didn't [19:12] yes, it is. [19:13] only when you select 3rd party stuff during install, it gets automatic installed [19:13] ahh [19:13] and counts as manual [19:14] thats that then [19:14] so if an update rolls around I will still get it? === genii-away is now known as genii [19:14] since nvidia just released a new driver [19:14] yes, but the new 470 might not be valid for older cards. [19:15] okay [19:24] Hi, I'm trying to ipxe boot ubuntu cloud images to servers with no disks. I have this so far: https://pastebin.com/SNesz515 [19:24] But I get errors. Does anyone know how to do this? [19:27] Hello, anyone here got experience with resizing partition while a server is running? [19:27] Found an old StackOverflow post detailing how its done. But my newer fdisk spit out some warnings that I'm not particularly stoked about [19:28] I have some experience [19:28] Are you using LVMs? [19:30] superdad: what errors do you get? do you need to add in some cloud-init datasource to your kernel command line? [19:30] superdad, nope, it's labeled as dos [19:31] It's MBR [19:31] @sarnold. When I ipxe boot, I get "can't open /dev/sr0: No medium found" [19:31] Hmm, that might be an issue [19:32] superdad: hrm. I wonder if the first thing the cloud images do is try to resize the first drive.. [19:32] @PMunch. Ya, it kind of depends on what type of filesystem I think [19:33] PMunch: is the filesystem ext4? Do you want to resize and then reboot for it to take effect, or to take effect immediately? [19:34]