[00:45] compdoc: it works with gnome just fine, except that you need an active session already logged in... which is sad in case of crash/reboot [00:47] I am thinking of setting up auto login during my trip so I can access remotely since I will have no physical access to my machine during that time :< [00:47] Enissay, I always install a minimal Mate desktop [00:48] no need for an active session [00:48] hmmm, I may do it... that machine is a mini pc actually not my main [00:49] I'll try that tomorrow then... time to hit bed at 3am :x [00:52] who am I kidding... I cant sleep with this in mind... Installing... [00:55] lol [00:56] you installing ubuntu mate from scratch? [00:56] no, over regular ubuntu 20.04 [00:56] thats the best way. works great as a server or desktop [00:57] I mean from scratch [00:59] good luck! [01:02] ofc, that's best way always [01:02] my machine still new, so I will try it as is now and format if needed [01:03] Mate Terminal is ugly a-f though [01:03] But on the bright side it indeed started a session even while none was active === Guest9398 is now known as war === M4he is now known as mahe [03:15] seems systemd and networkmanager are fighting over who controls hostapd, what's the right way to doi it? [03:24] hello [03:38] ezakimak: Out of curiosity; what shows for the render ' cat /etc/netplan/01-network-manager-all.yaml ' ? [03:48] there is no /etc/netplan [03:49] ezakimak: cat /etc/os-release | nc termbin.com 9999 [03:50] 20.04.2 LTS (Focal Fossa) [03:50] ezakimak: that's not the whole output [03:50] ezakimak: if you don't have /etc/netplan then you're not running Ubuntu 20.04. Or have modified it and removed netplan [03:51] https://paste.ubuntu.com/p/Pfr9HZN9Yv/ [03:52] ezakimak: so you removed netplan? [03:52] I didn't [03:52] it's possible the vendor did. [03:52] ezakimak: vendor? [03:52] it's for an arm mx8 device [03:52] they provided the image [03:54] ezakimak: ok, then it's unlikely this is a version of ubuntu that we can support. You'll need to seek support from the vendor. Especially when the support issue is specifically due to what they modified [03:54] i've sent an inquiry as to why they removed it [03:55] so is netplan the new "standard" way of configuring networking in ubuntu moving forward? [03:55] ezakimak: you'll also need to inquire about your networking issues regarding systemd and networkmanager [03:56] ezakimak: yes [03:56] ezakimak: https://netplan.io/ [03:57] i already have my hostapd configured how I need it [03:58] as well as my dnsmasq [03:59] seems there still isn't a built-in way to make firewall rules persistent [03:59] ezakimak: However: in ubuntu netplan is the configuration generator, for either NetworkManager, or systemd-networkd. [04:03] another config generator layer. great. [04:03] one more reason reminding me why I prefer gentoo [04:58] It does sound silly to have a generator just for 2 possible layouts. Just select one of these layouts and create a converter, not a third format... === tomaw_ is now known as tomaw === Avago_Broadqual7 is now known as Avago_Broadqual [06:32] Does anyone know what process/service is responsible for running the scripts inside /etc/update-motd.d ? [06:32] I'm trying to disable motd, and I'd like to stop this at its source, rather than make these files not executable [06:40] I'm running into issues with my system having the video card crash on boot on kernel 5.11, but it works on 5.8. I'm wondering if there is a bug. [06:42] Guest914, what are you trying to achieve by stopping that script from running? [06:51] Trying to disable the motd [06:52] In that case, just set the MOTD to be blank. [06:53] That way it will update it to be blank. I know running a useless script seems unusual, but it makes more sense than ripping out a primary function, and you never know when you might need to add an MOTD back. [06:54] Is there a singular process that runs all those scripts? [06:54] Yes, systemd handles them. [06:55] The update motd script runs in almost no time. [06:55] BTW, if you want a custom MOTD, there are some seriously cool things you can do with it. It parses all sorts of variables about the current state of your system, and you can set it to either parse on boot or on display. [06:56] Oh, did I mention you can do UTF8 or ASCII art if your terminal is capable. [06:56] lol I'll keep that in mind the next time I want to echo an emoji [06:57] Just be careful, a few UTF extended characters have minor issues with terminals not set to UTF-8. [06:58] I think Ubuntu needs to move the default advanced VT to use UTF8 now that there are options available and all video cards that are the slightest bit modern are capable of it on a hardware level. [06:59] An 8 bit per a glyph framebuffer actually runs slower on some newer cards which are optimized for 16bit and 32bit access partitioning. [06:59] Guest914, if you just want to stop diplaying the motd at login, you want to comment out the pam_motd lines in //etc/pam.d/{login,sshd} [07:00] rfm, yep. that is an excelent way to do it too. [07:01] I need to look into my video card issue further and get back. I hope it's not the same one I was running into earlier on Ubuntu 20.04 (SMU crashing on DPM initializing.) [07:01] Guest914, if you want to climb back further, there's motd-news.service and update-notifier-motd.service (both on timers() at least, and there might be more... [07:02] There is also a whole infrastructure for setting all sorts of neat stuff on MOTD. For instance, you can check the current system load average, and available memory. [07:03] Note that there are tricky little syntax distinctions between parse at runtime, parse at initial set, and parse at boot. [07:04] Okay I found that /etc/profile.d/update-motd.sh is what's running all those other scripts. So I'll just gut that out. Thanks for the help === shubbu_ is now known as shubbu [07:21] Hello. I have a question regarding how to interpret the man page for `sudo`. It's not Ubuntu specific. Can I ask it here? I don't intend to bind anyone by this question. It's OK if post-question, it is considered out of scope. [07:23] alzgh: man sudo [07:27] alzgh: just ask :P [07:27] ayrus it's about how to interpret it. `man sudo` show the following for the `-i, --login` switch: https://paste.ubuntu.com/p/D3JZqgb3rN/. I use this switch to go into root since root is not enabled on my system (ubuntu focal), but as hard as I try I can't see this in the text of the man page. [07:28] alzgh: have you considered 'sudo su' ? [07:29] Croran this isn't my question, but I have not. `sudo -i` worked fine for me. My question is how to learn to interpret the man page. [07:29] alzgh: what do you mean by you can't see the text of the manpage? [07:29] alzgh: your question is unclear. [07:30] OK, sorry. Let me rephrase. [07:30] alzgh: if you type 'man sudo' then type '/-i' and hit enter twice does it not display it? [07:30] *hit "n" once [07:32] I know that `sudo -i` let's me login as the super user. I found this out through means other then the man page (I used `tldr` and google). Then I went to read the sudo man page and looked at it's description for `-i, --login` and didn't find any reference to the fact that this switch helps you login as the super user. [07:32] alzgh: it's the `vim`. [07:32] I know how to search in the man pages. [07:33] My question is, how should I learn from this https://paste.ubuntu.com/p/D3JZqgb3rN/ that the `-i, --login` switch enables me to login as the super user. [07:33] What am I missing there? [07:33] In the man page. [07:33] Maybe you should be more interested in what Ubuntu means by "disabling the root user" [07:33] +instead [07:34] alzgh: are you mentioning there is no entry for `-i` in the man page of sudo? [07:34] alzgh: sudo gives you the permissions of the super user. why wouldn't you think this extends to logging in as the super user? [07:34] sneep maybe that's the root of my confusion. If I knew what it meant under the hood, maybe I would how the description of this switch worked. [07:35] Well, if you run sudo bash, do you get a root shell? Yes [07:36] alzgh: easier to 'sudo su' to get a root shell though. [07:36] alzgh: in the sudo manpage there's a part where it says if no user is provided it defaults to executing the command as root, and that -i part pertains to the login part where it will execute the command as if you logged in, changes $HOME etc [07:37] (user is provided via the -u argument) [07:37] Ubuntu doesn't set a password for the root user, thereby disabling authenticating as root in the standard login process [07:38] EriC^^ I think that clears things up. I was missing this part. the `-i` switch is just for loading the selected users preferences. [07:38] alzgh: are you saying you can't see the definition for -1 in the man page? [07:39] yeah exactly, just so you access the shell as if it were a login shell [07:39] I can see the definition, I wasn't able to understand it though. Was not able to make a logical connection between the text in the man page and how the tool worked. [07:40] Thanks guys. Really helpful. Sorry for the confusion. I'm trying to read and use the man pages but it isn't always easy to understand them :D [07:41] no problem [07:41] alzgh: the man pages are meant to be definitive, not readable :) [07:42] yeah, true. Good to have this in mind next time I go through them. Right now, I use `tldr` in many cases. [07:43] And after I find what I'm looking for, I look it up on the man page and try to make sense to it. [07:51] BTW, all that bash does when adding -i is call /bin/bash with special arguments. Specifically, it prefixes argv[0] with a hyphen character when doing the system call to execute a program. That is understood by bash and other shells (by convention) to invoke a login shell. [08:21] Hi, I am using grub on my computer. It's a laptop so I have my linux install on the whole internal disk using zfs. But I also have a windows installed on a usb drive and I'd like to add that option in grub instead of having to mash F12 every time I need windows. What's the right way? I took a look at /boot/grub.cfg and it says DO NOT EDIT [08:22] cxl: you can edit /etc/grub.d/40_custom and add a windows entry there, then 'sudo update-grub' to make a new grub.cfg [08:23] EriC^^: thank you I'll look there [08:23] cxl: no problem [08:24] cxl: you can probably just run `update-grub` while you have the stick attached [08:25] alkisg: that did it, it found windows. much easier, thanks [08:26] 👍️ [08:29] alkisg: I spoke too fast... it found it according to the command's output, I can see the new entry in /boot/grub/grub.cfg, but when grub actually loads then it's not on the list [08:30] cxl: that might mean it's using a different grub.cfg from elsewhere? [08:30] Press "c" to get a console in grub [08:30] Then set root=(hd0,gpt1) or something, to get to where your grub.cfg is, [08:31] then: configfile /boot/grub/grub.cfg [08:31] If THAT one shows the entry, then go back and echo root etc to see where your initial grub.cfg actually is [08:31] ok I'll try that now === denningsrogue1 is now known as denningsrogue [08:38] Evening/Morning/afternoon.... :-) [08:40] A Small Q: about Ubuntu Advantage subscriptions... Does anyone know if you get a list on the website when you've got more than a community/free account of the host names / IPs that have subscribed under your key / account? (Just got the free 3x one at the moment and looking at getting a paid subscription for work) === shubbu_ is now known as shubbu === m_ueberall is now known as ueberall [08:53] hello [08:54] Hello [09:23] how can I run a script automatically whenever apt (install or upgrade) is executed for a specific package? [09:25] https://askubuntu.com/questions/869219/how-can-i-run-a-script-after-a-specific-package-is-upgraded [09:25] :-) [09:34] xy? [09:35] sixwheeledbeast: https://en.wikipedia.org/wiki/XY_problem [09:36] it was a suggestion not a question ;) [09:36] I know, but XY is what I was thinking as well. :) [09:37] I should pay more attention my bad === denningsrogue8 is now known as denningsrogue [09:38] Don't worry, it didn't make sense anyway. [10:07] Hi! [10:08] I'd like to do some tests on ubuntu. Installed it through a vm via parallels desktop for macos. Opened the chrome website and pulled the 64 deb. After installing, opened but its not rendering correctly, shows a red square, etc. [10:08] Swahili: Jambo, habari gani? [10:08] Any idea what might've been wrong? I used to work on ubuntu, but haven't for the past 10 years more or less [10:08] im red yes [10:08] gordonjcp: lol sorry if that's some swahili, I have no idea! [10:09] Swahili: heh, my dad taught me how to say hello, goodbye, and please and thank you but that's about it [10:09] I also get that with band t-shirts that I don't actually listen to, but I like the art :~D [10:10] here's a screenshot: https://i.ibb.co/t25VprK/Screenshot-2021-07-08-at-11-10-11.png [10:12] guess its a parallels desktop issue and not the installation processs for chrome. I'll disable 3d acceleration === Thin_icE is now known as pgp [10:18] disabled 3d acc and worked [10:20] 👍 [10:47] hello everyone! === oerheks1 is now known as oerheks [11:46] hello, my laptop screen turned black since a restart. Do you know how to solve it ? (https://linux-hardware.org/?probe=1ec52430a2) [11:58] frogzilla: do you have a ubuntu install disk that you can boot to and use the try ubuntu option and verify if it is hardware or software [12:00] how ? [12:00] I have a usb. [12:00] and i can boot on it [12:00] how can I check if it is hardware or software ? [12:02] if it works with live cd it will be software so ? [12:02] and what can I do ? [12:47] Hi all [13:25] Hello i have question is lightdm still active project and developed by canonical? [13:28] Yes, used up to version 16.04 LTS. While it has been replaced by GDM in later Ubuntu releases, LightDM is still used by default in the latest release of several Ubuntu flavors [13:33] ok ,thanks for answer i wanted to use https://github.com/ArcticaProject/lightdm-remote-session-freerdp2 but this project seems quite dead, no documentation. To start rdp session on startup [13:36] freerdp2 is in our repos too.. https://packages.ubuntu.com/source/focal-updates/freerdp2 [13:37] no need to switch to lightdm [13:42] i think its built directly to DM, and then allow direct connection to remote with RDP from login screen or something like that [13:42] that github project is dead.. [14:30] After the last ubuntu update the bluetooth of my laptop isn't working properly. It is switched on but appears as off. === holgersson is now known as cb4 [14:31] DMESG gives [ 8.551508] Bluetooth: hci0: Failed to send firmware data (-38) [14:31] as well [ 8.705889] iwlwifi 0000:00:14.3: BIOS contains WGDS but no WRDS [14:31] [14:32] I am using 5.4.0-77-generic #86~18.04.1-Ubuntu [14:45] felipe, try to reload the btusb module; sudo rmmod btusb , wait a couple of cs and sudo modprobe btusb [14:46] *secs [14:49] ioria: That did not work === alpha_Aqu1lae is now known as clone === niko is now known as o === cb4 is now known as holgersson [15:31] Bluetooth error segfault lol [15:31] https://paste.ofcode.org/dHBudKPK7We9eVbEuU6THz [15:33] Sheilong: what version of ubuntu are you running? [15:34] leftyfb: 18.04.1 [15:34] Sheilong: why are you so far out of date? [15:34] I just booted a livecd of Lubuntu 20 and the bluetooth did work. [15:35] Sheilong: ok, we're not going to troubleshoot a live cd. Install it and then we can go from there [15:36] Sheilong: why is your installed version so far out of date? [15:36] leftyfb: Because canonical still gives support to this version. Also I have a lot of stuff installed on that I need to do my work. I will only upgrade it when canonical does not support 18.04 lts anymore. [15:36] leftyfb: This version come when I brought a new laptop in 2019. [15:37] Sheilong: I'm not talking upgrading 18.04 to 20.04. I'm talking 18.04.1 to 18.04.5 [15:37] Sheilong: run the regular package upgrades and then we can continue. [15:37] Sheilong: sudo apt update ; sudo apt full-upgrade [15:38] Sheilong: I also suggest the latest HWE kernel: sudo apt install linux-image-generic-hwe-18.04 [15:40] leftyfb: Okay. I will do it. [15:42] leftyfb: Done. [15:42] Sheilong: reboot [15:44] leftyfb: Done. [15:45] Sheilong: ( cat /etc/os-release ; uname -a ) | nc termbin.com 9999 [15:45] leftyfb: it still shows #86~18.04.1-Ubuntu [15:45] Sheilong: please post the output termbin URL here [15:45] https://termbin.com/flpw [15:46] Sheilong: ok, and what do you get for messages about BT now? [15:48] leftyfb: no. Everything seems to be okay. However, it does not see any bluetooth device on. [15:48] https://termbin.com/geyw [15:50] Sheilong: hcitool dev [15:50] Sheilong: you'll need to install bluez-utils # if the command isn't found [15:51] leftyfb: hci0 DC:FB:48:75:1F:B0 [15:51] Sheilong: ok, so your BT chipset is found and working [15:51] That's right. However, my phone is on and it can't find it. I will switch between on and off the device. === Ricardus_ is now known as Ricardus [15:52] after switch on and off, dmesg gives twice the message: "debugfs: File 'le_min_key_size' in directory 'hci0' already present!" [15:52] Sheilong: ok, so this is a usability issue. Why are you trying to pair your phone and laptop? [15:54] leftyfb: sorry, I meant Earphone [15:54] headphone [15:55] Sheilong: when you said your issue started after an update, did you mean after a kernel update, BT update or no clue wich? [15:56] lotuspsychje: It was a kernel update. [15:56] Sheilong: tnx, what you could try is boot a previous kernel as a test perhaps? [15:56] lotuspsychje: I did, but the problem remained. [15:57] lotuspsychje: I think we are passed it being a kernel issue [15:57] ah you tested [15:57] lotuspsychje: the chipset is detected and loaded just fine [15:57] kk [15:57] lotuspsychje: they weren't running HWE. I had them upgrade to it now [15:57] Sheilong: hcitool scan [15:57] Sheilong: do you see your headphones in the results? [15:58] leftyfb: No. [15:58] Sheilong: do you see any devices in the results? [15:58] Sheilong: are your headphones currently in pairing mode? [15:59] leftyfb: Also to make sure that the headphone is working I just connect it to my phone and it is working fine. [15:59] leftyfb: that command did returning nothing. [15:59] Sheilong: also, have you deleted the existing profile for your headphones from ubuntu so you can repair them? [16:00] leftyfb: Just did. Now I was able to connect. [16:00] leftyfb: Thanks so much for your help. [16:01] cookies for leftyfb [16:22] join #nginx [16:28] So what are the caveats of disabling feature-pinning for apparmor? As an alternative to just uninstalling or disabling apparmor altogether which is the recommended 'fix' by Proxmox Mail Gateway [16:39] Hello - having real strife with an anomalous boot issue - anyone able to invest a bit of time to assist? Long story short, this is a headless home media server, which seems to (re?)boot into read only mode on / for reasons unknown. the uuid I can see in fstab doesn't match anything I can see in blkid or grub.cfg and I cannot see why/how it has [16:39] changed (nor where it is even set to begin with) [16:39] . Any help & support greatly appreciated [16:39] fstab: https://termbin.com/4jrfp mount: https://termbin.com/d68i blkid: https://termbin.com/hplz grub.cfg https://termbin.com/z3zl grub probably is correct according to current blkid (547bf7de-c9c9-44c3-85ed-f02a07cbf4c9), it's fstab that has the wrong uuid: UUID=75755639-921e-4817-a790-353a6ea4d6c5 / ext4 errors=remount-ro 0 1 [16:40] however I cannot know (by nature of the problem) if that is what is causing it or not, it's the only obvious anomaly however [16:54] Hornet: look in "dmesg" for any I/O or related storage errors [16:55] if there were any on the initial boot would they not have been lost? this is read only this boot so dmesg might as well not exist [16:55] aiui anyway [16:58] Hornet, dmesg goes to a ringbuffer in ram it will have all info from the moment the kernel starts [16:59] (we're talking about the dmesg command, not about the /var/log/dmesg file) [17:00] right thanks :) [17:00] will look one sec [17:00] https://termbin.com/2966 ogra` [17:01] the UUID difference is normal, one is a GPT UUID the other a filesystem UUID ... [17:03] Hornet: did this used to work, and then this happened, or is this a new setup and you are trying to get it to work the first time? [17:04] used to work, was off for a while after a broken apt issue (circular dependancy unfulfillable yet somehow also uninstallable) [17:04] I sorted that BUT then this issue appeared behind it [17:04] apt was unable to work at all, so it's not as if it died part way [17:04] it just jammed and couldn't operate [17:05] re UUIDs, if you're sure then I'll take your word for it, thanks [17:09] Hornet, that does not look like there are any errors, and yor fstab seems to be preocessed fine as well, given the mount shows everything mounted properly [17:10] Hornet, what are the exact errors you get ... the system seems to have booted pretty correctly based on the info you give [17:12] ogra`: there are no errors - I'm just stuck with a read only / partition [17:12] and as there are no (save) errors I can't work out why [17:12] *saved [17:13] NetworkManager failed to start, but I can't find any actual log info showing the cause [17:13] ogra this is the fdisk -l output if it helps at all https://termbin.com/2a7u [17:14] "NetworkManager.service: Failed with result 'exit-code'" is all I got [17:14] https://termbin.com/m5ky lsblk -f also [17:14] Hornet, check if you see any errors with "journalctl -b", the dmesg log looks like a clean boot ... fdisk wont help here, the disks are all properly assigned and mounted, you look for mount errors [17:15] looking one sec, thanks [17:15] Jun 13 10:59:21 ArcaneSanctum kernel: Command line: BOOT_IMAGE=/vmlinuz-4.4.0-173-generic root=UUID=547bf7de-c9c9-44c3-85ed-f02a07cbf4c9 ro nomdmonddf nomdmonisw [17:16] last entry in there about booting I think [17:16] ezakimak: you are running a modified version of ubuntu. The modification specifically being networking. I don't think we can support that here. Please contact your vendor for support. [17:16] cause you don't use network manager? [17:16] and you don't use systemd? [17:17] you can't tell me how to get log files out of it for an error message? [17:17] ezakimak: because your vendor removed netplan and who knows what other modifications [17:17] they claim it never had netplan to begin with [17:17] Hornet, well, not sure where the last two options come from, the rest looks like a proper kernel cmdline [17:17] all netplan does is create config files [17:17] so how do you debug when that goes south? [17:17] look at the date though - it's jun 13th. that hasn't been written to in ages [17:18] certainly not today's boot [17:18] yeah, right [17:18] older version of ubuntu did not use netplan, do you not support that? [17:18] ezakimak: journalctl -u NetworkManager [17:19] Hornet, well, try fdisk with the -N option, that will run it in no-op mode so you only see what it would do but it wont apply any changes [17:19] Hornet: did you run "journalctl -b" or just "journalctl"? [17:19] ezakimak: if you install an official release of ubuntu, we can help you [17:19] oh, right [17:19] -b should really come from the ringbuffer [17:19] hornet@ArcaneSanctum:~$ sudo journalctl -b [17:19] the journal is useless. I already told you what it says [17:20] ah, I take that back, there's one more line that -xe didn't show me [17:20] Hornet: what does "uptime" report? [17:21] you said it was a server, so it's well possible that you last booted it on jnue 13? [17:21] 18:21:16 up 1:46, 1 user, load average: 0.00, 0.00, 0.00 [17:21] it was powered off so that would be a feat [17:22] the journal is indeed older than it should have been, though. [17:22] Hornet: what does this report? cat /proc/version [17:22] I promise you it was off. if it can run while unplugged I should have had it mining bitcoin [17:23] hornet@ArcaneSanctum:~$ cat /proc/version [17:23] Linux version 4.4.0-173-generic (buildd@lgw01-amd64-037) (gcc version 5.4.0 20160609 (Ubuntu 5.4.0-6ubuntu1~16.04.12) ) #203-Ubuntu SMP Wed Jan 15 02:55:01 UTC 2020 [17:23] that's a really old kernel. i assume your apt problems are not really solved [17:23] it's not been turned on for a long time, I fixed the apt circular issue, but then this appeared [17:24] apparently this kernel was built on (and probably for) ubuntu 16.04, too [17:24] did you post output of "mount" above? [17:24] DISTRIB_DESCRIPTION="Ubuntu 16.04.6 LTS" [17:25] which is EOL [17:25] I'm trying to update it! [17:25] fstab: https://termbin.com/4jrfp mount: https://termbin.com/d68i blkid: https://termbin.com/hplz grub.cfg https://termbin.com/z3zl fdisk -l: https://termbin.com/2a7u lsblk -f: https://termbin.com/m5ky [17:25] mount reports "/dev/sde2 on / type ext4 (ro,relatime,stripe=32730,data=ordered)" [17:25] note "ro" [17:25] I do indeed note it [17:25] that's read-only [17:25] that is the root of the issue (hurr no pun intended) [17:25] thats what we were discussiong the last 20min 🙂 [17:25] cat /proc/cmdline [17:26] probably has "ro" as well? [17:26] yar [17:26] it hopefully does [17:26] thats a default [17:26] BOOT_IMAGE=/vmlinuz-4.4.0-173-generic root=UUID=547bf7de-c9c9-44c3-85ed-f02a07cbf4c9 ro nomdmonddf nomdmonisw [17:26] so you asked to boot in read-only mode, and you got what you asked for., congratulations. [17:26] I didn't ask it to though [17:27] it's rebooting into it itself [17:27] well that'S what you or some software put into your grub.cfg, it would seem [17:27] tomreyn, thats a default ... the initrd remonts it rw usually [17:27] check your own machine 😉 [17:27] ogra`: oh my, right [17:27] UUID=75755639-921e-4817-a790-353a6ea4d6c5 / ext4 errors=remount-ro 0 1 [17:27] sorry [17:28] is it better to use dnsmasq via networkmanager or directly with systemd ? [17:28] there must have been errors somewhere ... but they are not logged anywhere [17:28] that's fstab's line about this (note UUID is different for reasons unclear) [17:29] is there any way I can debug what is actually happening AT boot time? [17:30] well, usually via dmesg and journalctl -b [17:30] but the latter didnt give you a recent log [17:30] https://lists.ubuntu.com/archives/ubuntu-devel/2014-February/038095.html discusses nomdmonddf nomdmonisw [17:31] https://termbin.com/s5ui that's dmesg [17:31] i would try "fsck -N /dev/sde2" ... thats a no-op and should tell you about any errors [17:31] journalctl -eb would move to the end of the log [17:31] can I run that fsck from a mounted (even read only) system? iirc fscks on mounts act weirdly [17:32] as long as it is readonly it should be fine [17:32] and as long as you use -N 🙂 [17:34] tomreyn, that shows the last 20 minutes! [17:34] should I reboot it and try to see if anything useful ends up there? [17:34] yay [17:36] hmm, still weird that it started logging this session so long ago. journalctl -p4 -b | nc termbin.com 9999 should give us more clues [17:36] oh wait this is 16.04, it doesn't usually have persistent journald [17:37] sudo journalctl --since=yesterday --until=now >>> https://termbin.com/wasj [17:38] thanks greatly for helping with this btw. I seem to end up with all the exotic issues [17:38] well, thats already readonly ... [17:39] Jul 08 16:35:03 ArcaneSanctum ntpdate[987]: [..] offset 2180128.236919 sec might explain some weirdness about logs. [17:39] yeah [17:40] smells like a wrong BIOS clock or broken battery ... but that still doesnt explain the readonly state [17:40] it doesn't. but we can assuem that jorunalctl -b does indeed output the current boot log. [17:41] the fact that npdate runs successfully before the network is up is also interesting [17:41] the battery is old, I can change it, fairly sure I have a spare [17:41] I have no idea how it's doing that bit of voodoo though? [17:42] so paste the full log of journalctl -b, the timestamps might be moot here [17:43] uptime shows correct current time, but will do, one sec [17:43] https://termbin.com/be1p [17:45] Jun 13 10:59:21 ArcaneSanctum kernel: EXT4-fs (sde2😞 mounted filesystem with ordered data mode. Opts: (null) [17:45] wait how are there errors in there about read only filesystesms [17:45] looks all good [17:45] Jun 13 10:59:21 ArcaneSanctum systemd-remount-fs[279]: /bin/mount for / exited with exit status 1. [17:45] there you go ! [17:46] Jun 13 10:59:21 ArcaneSanctum systemd-remount-fs[279]: mount: can't find UUID=75755639-921e-4817-a790-353a6ea4d6c5 [17:46] so you are right ... the UUID is wrong in fstab it seems [17:46] if this is saved on the / drive how was this log even written [17:46] that was my gut reaction but I thought uuids couldn't change ever [17:47] and I certainly didn't put that in there [17:47] its not written ... journald logs to ram by default and only flushes to disk later [17:47] oh I see so that 'old' log is actually today still [17:47] most likely, given your clock is off [17:48] I thought the last time I worked on it was the 13th though (my birthday so I remembered) [17:48] so for a test i'd try to comment the original / entry in fstab and change to the correct one for sde2 [17:48] that's one hell of a coincidence if so [17:49] I'll have to edit that with a live boot usb then, I'll dig one up and wire the box into something visual [17:49] 547bf7de-c9c9-44c3-85ed-f02a07cbf4c9 [17:49] that seems ot be the one you want [17:49] I know! /dev/sde2: LABEL="slash" UUID="547bf7de-c9c9-44c3-85ed-f02a07cbf4c9" from blkid [17:49] but that should -always- have been the one in fstab [17:49] you can just "sudo mount -o remount,rw /" [17:50] and then edit fstab [17:50] something has altered that, or, the sde2 uuid has changed since I wrote it in fstab [17:50] oho, black magic, I'll try 2 seconds [17:50] well, UUIDs can be changed with tune2fs ... [17:50] perhaps you ran something that called it or so [17:51] * ogra` needs to go afk ... but i'll return later ... [17:51] mount: can't find UUID=75755639-921e-4817-a790-353a6ea4d6c5 [17:51] lol. [17:51] thanks greatly for your help though :) [17:51] same fstab error is causing it to begin with [17:52] maybe we should have just done the remount command as debug stage 1! [17:52] that's not safe to do if the file system is corrupt [17:52] it isnt corrupt ... the UUID changed somehow [17:53] which causes: Jun 13 10:59:21 ArcaneSanctum systemd-remount-fs[279]: mount: can't find UUID=75755639-921e-4817-a790-353a6ea4d6c5 [17:53] so it never gets remounted RW [17:54] anyway, out ... [17:54] Why none my SysReq Magic sequences work except hard reboot? [17:54] does this rule out that the file system is corrupt? i don't assume it is since we know that the initial mount failed. [17:54] but we only learnt this in the past 5 miuntes, i think [17:56] Helmholtz1: i think some others should fwork, too, but certainly not all of them: cat /etc/sysctl.d/10-magic-sysrq.conf [18:00] Hornet: so you know how to proceed? [18:00] tomreyn: going to live boot something to edit that fstab manually, then reboot and see what blows up [18:00] wiring up kvm atm [18:01] why live boot, did you not remount.rw already? [18:01] no, it can't remount as fstab is errant! [18:02] that's the root of all this, as I suspected an hour ago [18:02] tomreyn, huh...so enabling them is a security issue [18:02] hornet@ArcaneSanctum:~$ sudo mount -o remount,rw / [18:02] [sudo] password for hornet: [18:02] mount: can't find UUID=75755639-921e-4817-a790-353a6ea4d6c5 [18:02]