| Emerald23969 | Hi, all Ubuntu based Linux systems I use are FREEZING randomly. Asus, AMD Ryzen CPU, Ubuntu 22.04. | 00:05 |
|---|---|---|
| Bashing-om | Emerald23969: See: https://ubuntuforums.org/showthread.php?t=2477028 Maybe related ? at least gives some hints on where to look. | 00:22 |
| whiskey76 | Greetings. Which ubuntu flavor has the most robust documentation? | 00:52 |
| wez | whiskey76: Under the hood they are all the same. It is just the windows manager that differ. | 01:26 |
| === M4he is now known as mahe | ||
| === Starmina3 is now known as Starmina | ||
| === esv_ is now known as esv | ||
| === ootput0 is now known as ootput | ||
| === keypushe- is now known as keypusher | ||
| SpeedStick | be having war thunder crash on my 22.04 ubuntu install all day | 03:18 |
| SpeedStick | just uninstalled steam/war thunder/nvidia drivers and re-installing everything hopefully that fixes it | 03:18 |
| === diskin is now known as Guest4088 | ||
| === diskin_ is now known as diskin | ||
| lotuspsychje | i got a user on launchpad reporting an apport error regarding SSL/login bug #1981901 anyone experienced this before? | 06:57 |
| ubottu | Bug 1981901 in xserver-xorg-video-intel (Ubuntu) "Weird behavior on display, since kernel versions 5.15.x" [Undecided, New] https://launchpad.net/bugs/1981901 | 06:57 |
| arraybolt3[m] | lotuspsychje: I've not seen that before. I wonder if something like Lynx got set to their default web browser? | 06:59 |
| lotuspsychje | that could be indeed an issue, ill ask him wich browser he uses tnx arraybolt3[m] | 07:00 |
| === javaJake_ is now known as javaJake | ||
| === ootput5 is now known as ootput | ||
| === SteelRose_ is now known as SteelRose | ||
| arthurvk | If I enable HWE on 18.04 and run kernel 5.4 all my php(-fpm) applications become much slower (~30%) than they are when running on the 4.15 kernel. I played around with mitigations=off but that didn't help. Anyone here an idea what is happening here, or what to look at to diagnose? The ubuntu machine is running in a KVM virtual machine btw. | 08:32 |
| lotuspsychje | arthurvk: could you share your dmesg in a paste so the volunteers can take a look for you? | 08:35 |
| irgendwer4711 | hi, anyone using Xen with Ubuntu and able to get cpu temp? | 08:54 |
| arthurvk | lotuspsychje: you mean the kernel boot logs? | 09:09 |
| lotuspsychje | arthurvk: lets start with a dmesg booting kernel 5.4 where your slowing occurs | 09:10 |
| arthurvk | this is kernel log from 18.04 booting 5.4: https://pastebin.com/C55FDew5 | 09:15 |
| arthurvk | and this is 18.04 booting 4.15: https://pastebin.com/yQxfjpbr | 09:17 |
| lotuspsychje | arthurvk: is this on ubuntu server? | 10:31 |
| arthurvk | lotuspsychje: yes it is | 10:33 |
| lotuspsychje | arthurvk: you might wanna ask that in #ubuntu-server togheter with your logs, i dont find a culprit right away | 10:33 |
| arthurvk | lotuspsychje: alright thanks for looking :) | 10:34 |
| Sh3n7u3ru | hi! | 10:36 |
| luna__ | hi | 10:37 |
| McQuestionable | Hi, I accidentally saved to my keychain, a blank password as the password for an ssh key | 11:24 |
| McQuestionable | For some reason it accepted it (I think because I'd already used that ssh key with the correct pass recently) | 11:25 |
| McQuestionable | Does anyone know how I can clear this blank password? I can't find anything in 'passwords and keys' | 11:25 |
| arraybolt3[m] | McQuestionable: I don't know off the top of my head how to do this, however, there might be other people here who do. It might take a bit before someone sees your message. | 11:26 |
| arraybolt3[m] | McQuestionable: If you're not finding anything in "passwords and keys", are you sure it actually saved to the keychain? | 11:27 |
| oerheks | private user keys are stored in ~/.ssh | 11:27 |
| McQuestionable | Looks like it did save, but I'm a dumb dumb and generated the key with the default name (my device name) | 11:27 |
| arraybolt3[m] | I mean, if the key still asks for your password and still works when you give it the right one, it may have just looked like it took it but didn't. | 11:28 |
| McQuestionable | so it was in passwords and keys, just under a dumb name | 11:28 |
| arraybolt3[m] | Oh. Well, hey, I've done sillier things before, so you're not alone! | 11:29 |
| athos | arthurvk: What version of php are you running? is it the one shipped with 18.04 on both cases? | 11:39 |
| arthurvk | athos: no, i'm running php 7.4 from deb.sury.org | 11:42 |
| === leon_ is now known as Guest8779 | ||
| === plvicente|2 is now known as plvicente | ||
| lblume | 'afternoon channel | 12:46 |
| oerheks | :-) | 12:46 |
| lblume | I've set up an Ubuntu 20.04 as an NFS client, configured its /etc/fstab. It works fine if run manually after boot (sudo mount -a), but it's not mounting automatically during boot. I found and tried various suggestions, but none changed anything. Is there documentation for that? | 12:48 |
| oerheks | pretty good nfs howto; https://www.digitalocean.com/community/tutorials/how-to-set-up-an-nfs-mount-on-ubuntu-20-04 | 12:50 |
| oerheks | mount -t nfs -o mountvers=3 <server_ip>:/path/to/shared-folder /path/to/local/mount options ## or mountvers=4.0 | 12:53 |
| === SteelRose_ is now known as SteelRose | ||
| === Bencraft14195 is now known as Bencraft1419 | ||
| lblume | I've found that one, it's using pretty obsolete options, such as "intr" ("intr / nointr This option is provided for backward compatibility. It is ignored after kernel 2.6.25") | 12:54 |
| lblume | Like I just said, I got no issue mounting, it's rather only /etc/fstab not doing it during boot | 12:55 |
| john | hi | 12:56 |
| john | i have a question about lubuntu flavour | 12:56 |
| john | does it use opengl or mesa3d | 12:56 |
| oerheks | after editing fstab, did you run update grub? | 12:56 |
| john | like if i want to play games that requires opengl 3.2 like Minecraft but my gpu doesn't support opengl 3.2 will it work or no? | 12:58 |
| john | oerheks u talking with me? | 12:58 |
| oerheks | Mesa comes preinstalled on Ubuntu with the open source graphics drivers of Radeon, Intel and Nvidia (sometimes0, to check: glxinfo | grep Mesa | 12:58 |
| oerheks | john^ | 12:58 |
| mjt | lblume: probably a network race condition, you could either use autofs, or see the options here: https://wiki.archlinux.org/title/NFS#Mount_using_/etc/fstab_with_systemd | 12:58 |
| john | oerheks so that mean i just have to check if they installed or no and update them then play? | 12:59 |
| lblume | oerheks: I wasn't aware that update-grub was needed after an fstab edit. Anyway, just tried, no change. | 13:00 |
| mjt | it'll most likely be because the network isn't available when it's trying to mount | 13:01 |
| OnkelTem | Hi all | 13:02 |
| OnkelTem | As far as I see, Kubuntu 20.04 cannot do Samba shares out of the box | 13:02 |
| OnkelTem | I tried to share a directory from Dolphin, and it didn't work | 13:03 |
| OnkelTem | I don't know the reason, I try to create a share for a printer | 13:03 |
| OnkelTem | How to check if my share is working? | 13:06 |
| oerheks | OnkelTem, correct, you will need to install samba, and do some trickery https://ubuntu.com/tutorials/install-and-configure-samba#1-overview | 13:11 |
| oerheks | mind the ufw part.. | 13:11 |
| lblume | mjt: Thanks, your link did the trick. I had found similar snippets earlier, but with less explanations: the mount only happens when it's accessed, not on boot, so I didn't see it right after boot. Adding "auto"/"noauto" doesn't change anything, "_netdev" doesn't appear to be useful either, but the x-systemd bits do work. | 13:12 |
| lblume | Works well enough for me: | 13:14 |
| lblume | [fc00:cafe::1]:/export/espace /export/espace nfs4 defaults,x-systemd.automount,x-systemd.mount-timeout=10 0 0 | 13:14 |
| mjt | yup, sorry I didn't know the precise flags, but the arch folk know what they're doing | 13:14 |
| mjt | from past experience, autofs is also generally more reliable with a flakey server | 13:14 |
| athos | arthurvk: in this case, it might be worth checking if ondrej did't backport https://salsa.debian.org/php-team/php/-/merge_requests/12 to the package you are using, which could explain recent performance differences | 13:15 |
| ubottu | Merge 12 in php-team/php "Fix detection of unknown gcc function attributes" [Merged] | 13:15 |
| lblume | mjt: It's okay, I've been doing NFS on and off for more than a quarter century now, my eyes are used to glaze over when hitting terrible advice, like "rsize=8192,wsize=8192", which by now is an established word of mouth tradition :) | 13:16 |
| lblume | In my case, it's just a VM guest talking to its host for files, so flakiness is not expected. | 13:17 |
| OnkelTem | oerheks: it looked like Dolphin triggered installing of Samba so I have it running. But let me check out the link your shared | 13:18 |
| mort | hey, I have a couple terabytes of data on an SSD formatted with EXT4. I want to move to using LVM. Is this possible at all without first making a back-up of everything on the SSD and wiping it? | 13:18 |
| oerheks | mort, in short; no. you need to wipe the drive before formatting as LVM | 13:23 |
| mort | hm, I see | 13:23 |
| mort | where exactly is our volume groups stored? | 13:31 |
| mort | and logical volumes | 13:31 |
| lotuspsychje | !lvm | mort | 13:35 |
| ubottu | mort: Tips and tricks for RAID and LVM can be found on https://help.ubuntu.com/community/Installation/SoftwareRAID and http://www.tldp.org/HOWTO/LVM-HOWTO - For software RAID, see https://help.ubuntu.com/community/FakeRaidHowto | 13:35 |
| SteelRose | mort: you can change the partition label with parted, provided that you used GPT, and label it as LVM ... if everything goes well, you should be able to use LVM to manage the disk | 13:51 |
| SteelRose | mort: with that said - ALWAYS have a backup ready just in case | 13:52 |
| === zakalwe` is now known as zakalwe | ||
| OnkelTem | oerheks: thanks. What has worked actually, is the old manual editing of smb.conf | 14:11 |
| oerheks | :-) | 14:12 |
| === sinvet is now known as ViktorasCNC | ||
| === ootput5 is now known as ootput | ||
| seecrets | people were talking shit about ubuntu on prism break | 14:18 |
| seecrets | https://prism-break.org/en/categories/gnu-linux/#operating-systems | 14:19 |
| seecrets | though the article was from 2012 do i have to be concerned about anything now? | 14:19 |
| ogra | seecrets, this channel is for support requests only, such a question fits better into #ubuntu-discuss or #ubuntu-offtopic | 14:22 |
| seecrets | oh, sorry | 14:24 |
| octav1a | Is there a tool I can use to make am image map? | 14:33 |
| ogra | octav1a, like https://docs.gimp.org/en/plug-in-imagemap.html ? | 14:38 |
| octav1a | ogra: thank you, that looks good! | 14:44 |
| octav1a | ogra: do you know how to 'complete' the polygon? I've tried like ESC, enter key, can't figure out how to complete the shape. | 14:55 |
| ogra | no, sorry, i have not used imagemaps since the turn of the century i think ๐ | 14:56 |
| leftyfb | octav1a: https://www.youtube.com/watch?v=kcffggKBDrg | 14:58 |
| === Blohsh is now known as Guest755 | ||
| === Guest755 is now known as Blohsh | ||
| BluesKaj | Hi all | 15:31 |
| === bryan is now known as TunlRat | ||
| irgendwer4711 | hi, anyone using Xen with Ubuntu and able to get cpu temp? | 16:01 |
| jhutchins | !lmsensors | 16:03 |
| ubottu | To access CPU temperature sensors and detect fan speeds, install the lm-sensors package. See https://help.ubuntu.com/community/SensorInstallHowto for installation and usage instructions. | 16:03 |
| jhutchins | BluesKaj: o/ | 16:04 |
| BluesKaj | jhutchins, o/ | 16:11 |
| irgendwer4711 | jhutchins: that does not work | 16:47 |
| wbw | psensor | 16:51 |
| jhutchins | irgendwer4711: "does not work" is not a valid problem description. | 16:52 |
| irgendwer4711 | jhutchins: because there is no driver for any cpu sensors found. | 16:54 |
| oerheks | irgendwer4711, zen, zen+, zen , 2 or 3? as it87 and nct6775 modules are included nowadays, kernel 5 or higher.. | 16:55 |
| jhutchins | irgendwer4711: Did you run sensors-detect? | 16:59 |
| irgendwer4711 | oerheks: i7 3770 | 16:59 |
| jhutchins | irgendwer4711: That's a CPU, not a motherboard chipset (which is where the sensors are). | 17:00 |
| oerheks1 | oh xen, not zen | 17:00 |
| irgendwer4711 | lol both wrong | 17:00 |
| irgendwer4711 | https://www.kernel.org/doc/html/latest/hwmon/coretemp.html | 17:02 |
| oerheks1 | i find no way to read temp in Xen.. | 17:05 |
| jhutchins | !xen | 17:07 |
| ubottu | XEN is a virtual machine monitor for x86 that supports execution of multiple guest operating systems with unprecedented levels of performance and resource isolation. Information on installing it for Ubuntu can be found at https://help.ubuntu.com/community/Xen | 17:07 |
| jhutchins | I would be quite surprised to see a virtual machine able to access low-level hardware like sensors. All the hardware a VM sees should be emulated. | 17:08 |
| irgendwer4711 | jhutchins: that work in Ubuntu 16.04 | 17:08 |
| oerheks1 | lmsensors should work. not for xen though | 17:09 |
| irgendwer4711 | and dom0 has access to hardware | 17:09 |
| === Speed2u- is now known as Speed2u | ||
| cbreak | a proper VM like via QEMU? | 17:38 |
| === oerheks is now known as Guest401 | ||
| === oerheks1 is now known as oerheks | ||
| cbreak | or one of those pseudo VMs that just pass through hardware and thinly abstract it? | 17:40 |
| irgendwer4711 | cbreak: proper? So proper, that a linux kernel does not need any hardware drivers | 17:42 |
| oerheks | sure lmsensors detect your intel. just not in Xen. even with dom0 | 17:43 |
| cbreak | if the virtual hardware and actual hardware are correlated, then it's not purely a virtual machine | 17:43 |
| cbreak | in QEMU, I can run 68k software on an amd x64 system, because everything's emulated | 17:43 |
| cbreak | if I'd use kvm, that wouldn't work, that thing can only provide the same CPU architecture, and to some degree, the same CPU | 17:44 |
| cbreak | (because it relies on actual hardware features that allow that hardware to pretend to be exclusively owned while being shared) | 17:45 |
| rfm | cbreak, your interpretation of "virtual machine" is at odds with its use from its invention. what you are calling a virtual machine, everybody else calls an emulated machine (that's the EMU in QEMU.) And these musings are OT here... | 17:47 |
| leftyfb | feel free to discuss further in #ubuntu-discuss or #ubuntu-offtopic | 17:48 |
| jhutchins | rfm: People are slopy about naming them. | 17:49 |
| rfm | I will join #ubuntu-offtopic for a while should anyone wish to continue. | 17:49 |
| jhutchins | rfm: The distinction isn't always important, but I think you're right in this case. | 17:49 |
| irgendwer4711 | cbreak: qemu ist an emulator! | 17:52 |
| irgendwer4711 | oerheks: what do you mean? | 17:52 |
| jhutchins | I don't think irgendwer4711 is here for solutions, he seems to want to complain and find fault instead. | 17:55 |
| oerheks | irgendwer4711, what part do you not understand? | 17:56 |
| cbreak | irgendwer4711: it provides purely virtual machines :) | 17:57 |
| irgendwer4711 | oerheks: last 2 clauses | 17:57 |
| irgendwer4711 | cbreak: you mean emulated machines | 17:57 |
| cbreak | purely virtual, yes. No actual hardware involved. | 17:57 |
| oerheks | So, lmsensors works? just not with Xen | 17:58 |
| irgendwer4711 | really? I tested vmware, vbox, kvm and xen. Xen is the only one which can work without any real hardware | 17:58 |
| arraybolt3[m] | irgendwer4711: The problem you're running into is that virtual machines don't have total access to all the hardware. Xen runs all operating systems on your machine as virtual machines, including the one you use directly (the one that "feels" like the physical machine, Dom0). If Xen doesn't pass through the sensors to the VM, the OS can't access them. And that's what it sounds like is happening. | 17:58 |
| irgendwer4711 | oerheks: I could see cpu temperator for al cores with Ubuntu 16 | 17:59 |
| arraybolt3[m] | (Some hypervisors may pass through the sensor hardware, and maybe Xen has an option to do so.) | 17:59 |
| irgendwer4711 | arraybolt3[m]: then xen peoples changed something | 17:59 |
| arraybolt3[m] | Then you probably need to tweak your Xen configuration. That's what I would guess. | 17:59 |
| arraybolt3[m] | Possibly. | 17:59 |
| irgendwer4711 | would be sad. so there would be no way to monitor your server. that cant be true.... | 18:00 |
| arraybolt3[m] | I believe they did in fact change something, I just found info about it. | 18:01 |
| arraybolt3[m] | https://xenbits.xen.org/gitweb/?p=xen.git;a=commitdiff;h=72e038450d3d5de1a39f0cfa2d2b0f9b3d43c6c6 Looks like they decided that PV guests didn't need it. | 18:01 |
| arraybolt3[m] | It apparently messed up Qubes OS users too. https://forum.qubes-os.org/t/cpu-temperature-sensors-in-qubes-4-1/11328 | 18:01 |
| irgendwer4711 | but dom0 is not an "unprivileged PV guest" | 18:02 |
| pycurious | In dmesg I keep getting these errors - any ideas how to fix this? -> [313173.108087] Out of memory: Killed process 81951 (python3) total-vm:40509220kB, anon-rss:31442168kB, file-rss:0kB, shmem-rss:0kB, UID:1021 pgtables:72568kB oom_score_adj:0 | 18:03 |
| irgendwer4711 | btw I have one ryzen xen server, this one can see temperature | 18:03 |
| irgendwer4711 | pycurious: have muich memory has that box? | 18:04 |
| leftyfb | pycurious: use htop to determine what process(es) is eating up CPU/memory | 18:04 |
| === diskin is now known as Guest3590 | ||
| === diskin_ is now known as diskin | ||
| Bradipo | What are all the possible methods for configuring IPv4 address on Ubuntu 20.04? /etc/netplan? /etc/network/interfaces? Any thing else to look for? | 18:10 |
| Bradipo | This is on a server install, so no UI. | 18:10 |
| leftyfb | Bradipo: /etc/network/interfaces is not setup by default | 18:10 |
| leftyfb | Bradipo: use /etc/netplan/*.yaml | 18:10 |
| rfm | Bradipo, if itล a ubuntu desktop install netplan just delegates to NetworkManager and itล easier to use the network manager config, setings>network in the gui | 18:12 |
| Bradipo | I've never found anthing more simple or "easy" than /etc/network/interfaces. Meh. | 18:14 |
| leftyfb | Bradipo: ifupdown has been deprecated in favor of netplan. It's new but it's not difficult at all. https://netplan.io/examples | 18:15 |
| === xet7_ is now known as xet7 | ||
| alkisg | Bradipo: other alternatives are systemd-networkd and network-manager; nmtui allows you to graphically create connections (dialogs etc) even from the terminal. netplan uses any of these two to do the actual work | 18:24 |
| leftyfb | Bradipo: what is your issue exactly? | 18:25 |
| Bradipo | Well, the issue is figuring out which of all he fangled methods Linux uses for configuring the network is the one that is in use and which *should* be used. :-) | 18:34 |
| leftyfb | netplan | 18:34 |
| leftyfb | for a server | 18:34 |
| leftyfb | NetworkManager for the desktop | 18:34 |
| Bradipo | Are you saying that there now is no other method on 20.04 to configure the networking? | 18:35 |
| === eleftheros is now known as sentientmetazoan | ||
| leftyfb | Bradipo: why can't you use netplan? | 18:35 |
| Bradipo | That's not what I'm asking. | 18:35 |
| === sentientmetazoan is now known as eleftheros | ||
| alkisg | network-manager works fine on servers too. And ifupdown still works if you do want to install it (or ifupdown2) | 18:35 |
| oerheks | normally one would set a static ip in the router, and use netplan | 18:35 |
| leftyfb | Bradipo: yes, there are multiple ways you can configure networking on an ubuntu server. | 18:36 |
| leftyfb | Bradipo: you've been given all, or most of them I believe | 18:36 |
| alkisg | So, systemd-networkd, network-manager, ifupdown, ifupdown2 => 4 methods; netplan can use the first two of them as backends | 18:36 |
| Bradipo | If there is only one method to configure then when I stumble upon an Ubuntu 20.04, I need look only to one place to figure out how it is *currently* configured. | 18:37 |
| leftyfb | Bradipo: by default, ubuntu server uses netplan | 18:37 |
| Bradipo | But given that there are multiple methods, I may have to look in other locations because I have no idea how the person that configured it may have done it. He may prefer "ifupdown", or he may prefer NetworkManager, or netplan, or whatever. | 18:37 |
| Bradipo | Thanks for mentioning the various locations I will need to look for network configuration details. | 18:38 |
| alkisg | /etc/netplan, /etc/systemd, /etc/network/interfaces*, /etc/NetworkManager/system-connections | 18:39 |
| Guest54 | Does ubuntu use SYSLINUX or ISOLINUX? | 19:15 |
| plujon | I have two machines using Ubuntu 20.04 LTS, and I notice they have 2 different versions of the Linux kernel: Linux 5.15.0-41-generic and Linux 5.4.0-122-generic . Is one of these "normal"? | 19:16 |
| leftyfb | plujon: one is running the HWE kernel | 19:17 |
| leftyfb | !hwe | plujon | 19:17 |
| ubottu | plujon: The Ubuntu LTS enablement stacks provide newer kernel and X support for existing LTS releases, see https://wiki.ubuntu.com/Kernel/LTSEnablementStack | 19:17 |
| plujon | In my digital ocean droplet, I have Linux 5.4.0-122-generic , and that version has a bug that has caused my server to panic, and I don't understand why it is using 5.4.0-122-generic. I have already run 'apt-get install update' | 19:17 |
| plujon | THe bug is https://bugs.launchpad.net/ubuntu/+source/linux-hwe-5.4/+bug/1981658 . | 19:19 |
| ubottu | Launchpad bug 1981658 in linux-hwe-5.4 (Ubuntu Bionic) "BUG: kernel NULL pointer dereference, address: 0000000000000008" [Undecided, Confirmed] | 19:19 |
| plujon | So, HWE uses an older kernel base, but adds support for newer [hardware] features that are useful in a cloud environment? | 19:20 |
| leftyfb | plujon: HWE is the newer kernel | 19:20 |
| leftyfb | plujon: sudo apt install linux-image-generic-hwe-20.04 | 19:21 |
| leftyfb | plujon: though, if this is a droplet, I don't think that's going to work out the way it's meant to. You should contact DO for support | 19:21 |
| plujon | leftyfb: Thanks; I've contacted D.O., and I downgraded my kernel to 5.4.0-121-generic to avoid the kernel panic. | 19:22 |
| plujon | But I don't yet understand what you're saying. It sounds like you're saying my droplet is using the non-HWE kernel (5.4.0-122-generic). Is that correct? | 19:22 |
| plujon | And 5.15.0-41-generic is the HWE kernel? | 19:23 |
| plujon | I think my real question is, if a kernel panic occurs in 5.4.0-122-generic, and `apt-get update` does not move my droplet beyond 5.4.0-122-generic, doesn't that indicate that many people are using 5.4.0-122-generic ..? | 19:25 |
| plujon | I had 2 different droplets in 2 different data centers hit the same kernel panic two days ago. | 19:26 |
| sarnold | please do file bugs about those | 19:27 |
| plujon | sarnold: https://bugs.launchpad.net/ubuntu/+source/linux-hwe-5.4/+bug/1981658 | 19:27 |
| ubottu | Launchpad bug 1981658 in linux-hwe-5.4 (Ubuntu Bionic) "BUG: kernel NULL pointer dereference, address: 0000000000000008" [Undecided, Confirmed] | 19:27 |
| plujon | I haven't added a "me too", but that's mainly because I don't have a launchpad account, and I'm trying to understand if I'm on an unusual version of the kernel, or a fairly normal one. | 19:28 |
| sarnold | plujon: dang man, you've got mellanox hardware?? | 19:28 |
| sarnold | plujon: how much do those cx-5s go for? heh | 19:28 |
| sarnold | plujon: oh I see | 19:28 |
| plujon | I did not file that bug; I use a VPS; the RIP is quite similar though. | 19:28 |
| plujon | http://ix.io/44yf | 19:29 |
| plujon | RIP: 0010:tcp_create_openreq_child+0x2fd/0x410 | 19:29 |
| plujon | Because TCP is involved, I was curious whether this was an exploitable flaw. But in that case, servers all over would probably be going down, unless I'm on an unusual release. | 19:30 |
| plujon | Digital Ocean staff told me to update to the latest, but it was the latest version that caused the panic. | 19:31 |
| === u0_a2048 is now known as jadenlian | ||
| ravage | it should be pretty safe to use the HWE kernel. do a backup before if you can | 19:32 |
| ravage | if i remember correctly DO also has a console. so you can boot the old kernel again in case it does not work right? | 19:33 |
| plujon | I've fixed my droplet. | 19:34 |
| ravage | using an outdated kernel version is a temporary fix | 19:35 |
| plujon | ravage: I don't understand what you're suggesting. What do you mean by the "HWE kernel"? The bug report claims that it is the HWE kernel that is affected. | 19:35 |
| ravage | that bug report is for Ubuntu 18.04 | 19:36 |
| ravage | there the HWE kernel is 5.4 | 19:36 |
| ravage | the HWE kernel for 20.04 is 5.13 | 19:36 |
| plujon | Ah, hmm. | 19:36 |
| plujon | I think these droplets were upgraded from 18.04 to 20.04 LTS. | 19:36 |
| plujon | But apparently that did not update the kernel version. | 19:36 |
| plujon | cat /etc/issue reports: 20.04.4 LTS | 19:38 |
| plujon | uname -r reports 5.4.0-121 (because I downgraded from -122) | 19:38 |
| ravage | so as leftyfb suggested "sudo apt install linux-image-generic-hwe-20.04" will install kernel 5.13 and probably fix your issue. and you also get updates again | 19:38 |
| leftyfb | ravage: the HWE kernel for Ubuntu 20.04 is 5.15.0.41.44~20.04.13 | 19:38 |
| ravage | even better | 19:39 |
| plujon | Is it normal when one moves from one LTS to the next for the kernel to stay on the older version? | 19:40 |
| ravage | no | 19:40 |
| ravage | but the DO default installations are not the best | 19:40 |
| ravage | maybe they dont provide the metapackage and only have a specific kernel version | 19:41 |
| plujon | I'm clumsy with dpkg and apt stuff; I'll see if I can discover what the "installed" image is... | 19:42 |
| ravage | apt list --installed|grep linux-image- | 19:43 |
| plujon | ii linux-image-virtual 5.4.0.122.123 amd64 Virtual Linux kernel image | 19:43 |
| ravage | that is actually the latest version for the "virtual" image | 19:44 |
| plujon | And the definition of linux-image-virtual comes from D.O. sources, presumably? | 19:45 |
| ravage | it is an official ubuntu package | 19:45 |
| ravage | if you want to stay with the virtual one there is also linux-image-virtual-hwe-20.04 | 19:45 |
| ravage | thats also 5.15 | 19:45 |
| plujon | Hmm. So, anybody who uses linux-image-virtual and who has fully upgraded is using 5.4.0.122.123 ? I wonder if .123 means it has been updated already... | 19:47 |
| ravage | https://packages.ubuntu.com/focal/linux-image-virtual | 19:48 |
| ravage | yes thats the latest one | 19:48 |
| plujon | Wouldn't that imply that 5.4.0-122 is widely deployed? If so, ... hmm... I'm rather surprised I had *2* servers hit that panic the same day. | 19:50 |
| plujon | Both servers may have been updated for 18.04 LTS to 20.04 LTS. Hmm. Anyway, strange. | 19:51 |
| plujon | FWIW, I had never before seen a kernel panic on my droplets in 7+ years. If I were paranoid, I'd say the bug is exploitable via specially crafted TCP packets and somebody was targeting my servers or randomly targeting servers. But both seems quite doubtful; the former because I'm not that important, and the latter because you would have heard complaints about this already... | 19:53 |
| tomreyn | if you'd like to see this fixed in 20.04's vanilla (non HWE) kernel, it would be good to report this bug, since so far the bug report only lists 18.04 LTS as affected | 19:54 |
| webchat93 | Is it possible to run xsplit vcam on wine | 20:01 |
| webchat93 | Is anyone there | 20:01 |
| sarnold | yes, folks are here, but not everyone is going to know what xsplit is.. | 20:03 |
| sarnold | and not many people run wine | 20:03 |
| sarnold | so you may be the only one who's tried xsplit on wine | 20:03 |
| webchat93 | It's a Windows program which I want to run on wine | 20:03 |
| respawn | sarnold: https://alternativeto.net/software/xsplit-broadcaster/?platform=linux | 20:06 |
| === ootput3 is now known as ootput | ||
| sarnold | aha yeah, loads of people use OBS | 20:07 |
| el | webchat93: winehq doesn't have recent results but says it doesn't work. so if you're trying now and it's not working then it probably did not improve in the past 11 years. https://appdb.winehq.org/objectManager.php?bShowAll=true&bIsQueue=false&bIsRejected=false&sClass=version&sTitle=&sReturnTo=&iId=23272&sAllBugs | 20:07 |
| el | let me guess that's a garbage link and everything | 20:08 |
| respawn | clikc on it and see | 20:08 |
| el | yeah it does seem to work | 20:08 |
| respawn | there are lots of alternatives for Linux google is your friend | 20:09 |
| el | the link i mean | 20:09 |
| plujon | tomreyn: https://bugs.launchpad.net/ubuntu/+source/linux-hwe-5.4/+bug/1981658/comments/13 # added comment here | 20:15 |
| ubottu | Launchpad bug 1981658 in linux-hwe-5.4 (Ubuntu Bionic) "BUG: kernel NULL pointer dereference, address: 0000000000000008" [Undecided, Confirmed] | 20:15 |
| oerheks | plujon, that should be linux-image-generic-hwe-20.04, right? | 20:18 |
| plujon | oerheks: I don't know what you mean. | 20:19 |
| oerheks | your server uses the virtual kernel package | 20:19 |
| plujon | Doh, I made a mistake in that report. I copied the microseconds wrong in a couple lines... Oh well. | 20:19 |
| oerheks | i am not sure linux-hwe-5.4 even exists.. | 20:19 |
| plujon | oerheks: Sorry, I don't understand what you mean. dgpk-query -l reports this: | 20:20 |
| plujon | ii linux-image-virtual 5.4.0.122.123 amd64 Virtual Linux kernel image | 20:20 |
| oerheks | that would be correct linux-image-virtual | 20:21 |
| plujon | I don't know what you mean by "should be linux-image-generic-hwe-20.04" ? | 20:22 |
| plujon | I probably should have added in my comment that the droplets were long ago upgraded from 18.04 LTS. | 20:24 |
| plujon | (to 20.04 LTS) | 20:24 |
| plujon | I'm quite confused about how that relates, though. It sounds like perhaps 20.04 LTS normally doesn't use "linux-image-virtual" . | 20:25 |
| plujon | oerheks: If you're saying my server **should** be using linux-image-generic-hwe-20.04 and **should not** be using linux-image-virtual , I can go ahead and update it to the former. Is that what you're saying? | 20:30 |
| oerheks | no, there seems to be no hwe with virtual, my mistake | 20:30 |
| oerheks | just your bugreport should mention virtual kernel, not the regular one | 20:31 |
| plujon | To answer my own questions: it seems that somewhere around 5% of Ubuntu 20.04 desktop users are on 5.4.0-122-generic . About half are using 5.13.0-52-generic or 5.15.0-41-generic. As for servers, I don't know if the percentage would be lower or higher. | 20:39 |
| plujon | In contrast, most of those using Linux Mint 2.0[123] are using 5.4.0-*. | 20:41 |
| oerheks | on linux mint server? | 20:44 |
| plujon | Heh, no; I doubt there are many of those. I meant Linux Mint Desktop. | 20:44 |
| plujon | Anyway, thanks for the help; I'll wait on the bug and see what happens with it. | 20:46 |
| === diego is now known as Guest2886 | ||
| diego | Why use noobunutu | 20:50 |
| diego | ? | 20:50 |
| === diego is now known as Guest8009 | ||
| oerheks | never said we do// | 20:52 |
| sarnold | can't even troll properly smh | 20:57 |
| EternalSunshine | i use arch tbh | 21:01 |
| oerheks | good for you | 21:02 |
| sarnold | there we go. *that*'s trolling | 21:02 |
| arraybolt3[m] | EternalSunshine, diego: You probably should be aware that sarnold is an operator - if this gets much further out of line, you are probably going to get muted or banned for a while. | 21:03 |
| EternalSunshine | I retract my statement | 21:04 |
| EternalSunshine | I have ubuntu installed on my laptop. | 21:04 |
| EternalSunshine | but arch on my server/desktop | 21:04 |
| EternalSunshine | arraybolt3[m]: my apologies, I was merely trying to illustrate trolling properly | 21:04 |
| EternalSunshine | but ironically | 21:04 |
| EternalSunshine | I can't speak for diego | 21:05 |
| ogra | you did well | 21:05 |
| ogra | ๐ | 21:05 |
| arraybolt3[m] | np, just don't want to see any sparks fly. | 21:05 |
| sarnold | arraybolt3[m]: do note that diego there couldn't even remain connected; he kept getting renamed by services, kept quitting, etc. just a big fail all around. | 21:05 |
| === ajfriesen1 is now known as ajfriesen | ||
| === JD_ is now known as JD | ||
| === JD is now known as Guest2481 | ||
| forgotmynick | to allow a user1 to sftp but not open a shell, i set the shell to /usr/sbin/nologin. as expected this prevents me from doing `su user1` with the message This account is currently not available. is there anyway to allow the root account to switch to user1 and allow user1 to access sftp but not a shell? | 22:43 |
| forgotmynick | oh snap i can do su -s /bin/bash user1 | 22:45 |
| oerheks | root account not to access a shell.. | 22:45 |
| Drew_Neilson | Can someone tell me if in Ubuntu permissions can be set on a per-app basis, and if so, how to do it? | 23:30 |
| Drew_Neilson | setting permissions for "system internal users" or something like that. | 23:31 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!