[00:10] hi! my Firefox in Ubuntu focal cannot display Unicode character U+2420. Am I missing some font? which one? [00:11] it's the "Symbol for space" in block "Control pictures" [00:17] interesting. xterm also cannot display U+2420, but it can indeed show U+2424 albeit Firefox cannot. [00:27] tomreyn - no, it's a bug in the acpi-call package that has been fixed but not yet backported to 20.04. The bug is reported here: [00:27] https://bugs.launchpad.net/ubuntu/+source/acpi-call/+bug/1953261 [00:27] Launchpad bug 1953261 in acpi-call (Ubuntu) "kernel >= 5.13 BUG: kernel NULL pointer dereference" [Undecided, Fix Released] [00:27] good evening. i'm noticing something bizarre with my HP dy2037nr notebook's keyboard when I log into X (no matter which environment [KDE or Unity). When I first log into X immediately after a reboot/startup, my keyboard is unresponsive for at least 20-30 seconds. It's as if the PS/2 bus isn't fully initialized by the time I log in. I'm using 21.10. [00:28] i have autologin enabled too. [00:28] tomreyn - what I need to know is if I should delete dump-* in both ../efivars and ../vars [00:29] hey hey [00:29] oddly enough, my trackpad is responding without an issue. [00:30] where do you suggest i _start_ troubleshooting a mount error using mount.cifs against a remote samba/cifs share, that renders a "mount error(2): No such file or directory" ? [00:30] sybariten, look in syslog when you try the mount. [00:31] THe command line used is sudo mount.cifs -o user=aUserName,pass=HhsdfXs5dfsFDs4f9d //server.in.foreign.country.com/backup /mnt/lester [00:32] except for the mount error line, i also got "Refer to the mount.cifs(8) manual page (e.g. man mount.cifs) and kernel log messages (dmesg)" - is dmesg related to syslog, dwitt1974 ? [00:33] My problem with USB mic on Linux is solved ( It was either that or going bald stressing me out for nothing ): https://twitter.com/Mitacho_/status/1498088474838192129?s=20&t=tFMzKs3DX9kV_ilbsm2ZkA [00:33] dmesg does the kernel messages [00:34] /var/log/syslog is what i would look at too [00:34] try tailing the syslog when you try the command. [00:34] Bonsoir. [00:34] try sudo mount -t cifs .... instead of mount.cifs directly [00:36] this may seem silly, but verify that cifs-utils is installed. [00:38] dwitt1974: hmmm yeah i think i see it.... [00:40] no worries. been there alot before. :) [00:40] dwitt1974: https://pastebin.com/vQ9jTsim its the things around 00:37 [00:41] its a commercial storage service, but i gotta admit i've never done samba over WAN before [00:42] ah ok. I would try checking the firewall/port forwarding and make sure those are correct. like ports 135, 137, and/or 139 [00:42] dwitt1974: at my place? [00:42] (or whether ssh port forwarding is correctly setup) [00:42] wherever the drive(s) are at. [00:43] Ah, OK but that's over at the server company... i cant change anything there but i could ask them ofcourse. [00:44] yeah. also make sure your port forwarding is correct (if you're using ssh for example). [00:45] i usually use ssh port forwarding when going into my company's server (for java debugging, sql servers, etc). [00:45] and that's if port forwarding is even allowed for my ssh account (on the server side). [00:45] I just realized that maybe this remote mounting of a network share is something that will also require to make some FW changes to my own router.... on the other hand i feel like there's seldom any need for me to make a lot of adjustments these days, most things be it torrent or other stuff, just "works" with my simple router [00:46] sybariten: Does the provider have instructions on how to connect to the share? [00:46] i connect to ssh servers now and then but i've never done port forwarding.... this account though, there's no ssh shell access, only scp [00:47] jhutchins: https://docs.hetzner.com/robot/storage-box/access/access-samba-cifs/ , ctrl-f "instructions" [00:47] sybariten: You might be able to use sshfs, but you'd be bettter off with CIFS. [00:48] jhutchins: hmmm OK yeah i'd prefer if "real" mounting is possible. I've verified that i could do some simple SFTP communication with the account, made some subdirs etc [00:48] but yeah i may need to talk to the support for this one. I'm guessing there are too many variables to give an educated guess here :) [00:48] s/give/make [00:49] sybariten: Which of the steps there have you tried? They look pretty straightforward to me, and pretty much what I use for a CIFS mount. [00:50] jhutchins: mount.cifs -o user=,pass= //.your-storagebox.de/backup /PATH/FOLDER (but with sudo preceding) [00:50] sybariten: //minerva/apps /minerva/apps cifs credentials=/etc/samba/auth.minerva.root,file_mode=0777,dir_mode=0777,_netdev 0 0 [00:50] sybariten: That's one of my fstab entries. [00:51] hmmm yeah i have to admit i dont know the fstab syntax off hand, and how it relates to a mount command [00:51] do you keep UN and PW in a clear text file and refer to that? [00:52] Credentials file is username = \n password = \n domain = [00:52] sybariten, I can certainly imagine your router could be blocking traffic out on port 445. It might be worth trying "nc that.server.de 445" just to see if it is the net [00:52] From the look of the directions, you can omit domain. [00:53] smbclient is useful in diagnosing connections as well. [00:53] hmmmmm do i always need to mess with a domain for a samba mount ? [00:53] sybariten: Probably not. [00:53] ok [00:53] i think that is what i always set to "Workgroup" for my own samba mounts, to be used from Win :) [00:54] btw the original goal was to use this from Win too, but that did absolutely not work :) so i figured i'd take it one step back first and see if i can hand mount it from *nix [00:54] sybariten, if it can't connect nc will just return silently, if it connects nc will wait for you to type something, ^C out... [00:54] rfm: gotcha [00:55] sybariten, oh hey, nc has a -v switch to actually have it say what happens.. [00:57] rfm: hmmm yeah i got a prompt that is sort of like waiting... btw i did this against the subdomain which is my account, so to speak. nc myaccount.theservercompany.de 445 [00:58] got no extra info using -v as far as i can tell? btw netcat is quite cool [01:03] sybariten: Sorry, my smb client is dropping off the net, gotta go chase that. [01:04] sybariten, so it seems like the net is working. Uh, you did mkdir /mnt/lester? "No such file or directory" is what I get when I try to mount to a nonexistent mountpoint... [01:04] jhutchins: gotcha. Thanks for the help! [01:05] rfm: yeah, sudo mkdir /mnt/lester [01:05] it's there [01:06] gonna see what smbclient -U myUsername -L myUsername.theservercompany.de says... [01:06] What I really need to do with this upgraded legacy system is back it up, do a clean install and check that for stability. I can't tell if it's hardware or software that's failing. [01:09] hmmmm do_connect: Connection to myUsername.theservercompany.de failed (Error NT_STATUS_IO_TIMEOUT) [01:11] sybariten, that sure sounds like something along the way is dropping packets. I have no idea what ports/protocols smbclient talks on... [01:13] rfm: ok! i suppose by now we are reaching areas which are better answered by this companys support, they should have hundreds of users doing this thing i am trying [01:16] sybariten, at least they'd probably know which ports have to be open. I found https://serverfault.com/questions/346196/tcp-ip-ports-necessary-for-cifs-smb-operation which does seem to say tcp/445 is enough if not using SMB name resolution (which you aren't) [01:17] okay, thanks for checking that, and other suggestions! Gonna catch some Z's now...! [01:18] rfm: yep, 445 is mentioned : https://docs.hetzner.com/robot/storage-box/access/access-overview/ === px_ is now known as px [01:23] sybariten, I see they support WebDAV over HTTPS which would seem vastly superior (encrypting the traffic, if nothing else.) It seems to be integrated in Windows https://www.maketecheasier.com/map-webdav-drive-windows10/ [01:24] sybariten, I know there are WebDAV file systems in Linux too but I've never used them [01:26] rfm: you really think that feels as transparently as a real disk? [01:28] !info davfs2 [01:28] davfs2 (1.6.0-1, impish): mount a WebDAV resource as a regular file system. In component universe, is extra. Built by davfs2. Size 148 kB / 388 kB [01:30] sybariten, I would expect the performance characteristics to be a little different, since it probably moves bigger chunks of data than cifs blocksize. mounting a filesystem over a WAN is never likely to be super zippy no matter the protocol.. === aaguhagegaooeg is now known as westor === westor is now known as Guest5492 [01:34] rfm: hehe no, that may be true..... alrite, peace! === mkv is now known as m4v === M4he is now known as mahe === Guest13 is now known as JasJas [03:00] hello [03:01] hi foe [03:17] got ourselves a little net split tonight. lol === JasJas is now known as Chorizo === _miruoy is now known as miruoy [04:17] while booting, when it shows the steps ans [Ok] and [FAILED] I saw a few FAILED [04:18] But it went too fast. Where do I see that log after I've fully booted? [04:19] ednash: ' journalctl -b -0 ' :D [04:19] yes, -0 is from now, -1 from previous boot [04:21] wow that's cool [04:25] looks like the failed messages was because i turned off swap [04:25] i am not sure why it is still trying to find a swapfile [04:26] "Failed to activate swap /swapfile." [04:27] i don't need one if with 32 GB of physical ram right? [04:28] how did you turn off swap?"just removing the swapfile is not enough, disable and mask swap service [04:28] you asked that before, 32 gb could do without swap [04:28] i also used swapoff. i think i did a swapoff -a if memory serves [04:29] yes i did... :) [04:29] ednash: Did you also remove the swap entry in the /etc/fstab file ? [04:30] ahh Bashing-om no it was there! commented it out now, thanks [04:30] even though i didn't create a swap partition, ubuntu installation added a 2GB /swapfile earlier [04:31] the `journalctl -b -0` trick is very neat, thanks for that [04:31] Hey everyone [04:32] Anyone else here tried 22.04 yet? [04:32] ednash: Yeah - fresh install now makes up a swapfile rather than a dedicated partition. [04:32] prestocaso[m]: is it out already? [04:32] prestocaso[m], some do, but the next version support has its own channel [04:33] !next [04:33] Jammy Jellyfish is the codename for Ubuntu 22.04. For technical support, see #ubuntu-next. For testing and QA feedback and help, see #ubuntu-quality. [04:33] !22.04 prestocaso[m] [04:33] Bashing-om: i see! i manually created the partitions, so i thought that would be the end of it [04:33] good to know [04:33] ednash, swappartition <> swapfile [04:33] what did you disable exactly? [04:33] Sorry for the confusion I am referring to the daily build [04:34] I am not looking for support though, just anyones thoughts [04:34] yeah... i mean i thought if i explicitly remove the partition, it wouldn't create a file for me. but the installation has some magic logic perhaps which detects that there's no partition, and then creates the /swapfile instead? who knows [04:34] 22.04 is LTS, so not that much new features, stable is the goal [04:34] oerheks: earlier i did swapoff and deleted the file. i missed the /etc/fstab, removed it now [04:35] ednash, good spot [04:37] learning from the best === William is now known as Conna === not_phunyguy is now known as phunyguy [06:07] hey guys whats a good app for non techies to convert mkv to avi or divx, i'm running a plasma DE [06:09] hey guys whats a good app for non techies to convert mkv to avi or divx, i'm running a plasma DE === not_phunyguy is now known as phunyguy [06:14] There are lots of good apps for stuff like that. I'm about to hop off to bed though. Good luck, Mystified1234 [06:14] thanks [06:14] https://www.makeuseof.com/tag/10-applications-to-convert-audio-and-video-files-in-linux/ [06:14] ... Try that [06:15] If you're still around tomorrow (or tonight, depending on your time zone) with the same issue, I will attempt to assist. === not_phunyguy is now known as phunyguy === Jubes1 is now known as Jubes === jgee636 is now known as jgee63 === diskin_ is now known as diskin [08:11] I want to correct the color on some of my videos. to change the white balance. is ther a good linux-native software that will do that? [08:26] Hello :) question regarding package management. Client has ubuntu 20.04, we have our own dev env. they want to update a wide array of packages on their system and sent us the suggested upgrade log if it's okay. Is there a better way of tracking/comparing packages between our and their system without connecting them? maybe a tool too export/import [08:26] these kind of things and automate further [08:37] fruity_tomato: you mean that in general they don't update their systems, and now you want to do a selective update? Why don't you fully update them so that you won't need to worry about security issues? [08:40] alkisg : debian allows custom upgrades, for example, i upgrade from buster to bullseye, i can keep kernel 4.19 [08:41] frost-core: that's called "frankendebian" and it's not supported [08:41] yeah [08:41] i did franken debian [08:41] one tile [08:41] time* [08:41] suprisingly it didnt break [08:42] But I haven't understood what you're asking now. Are you asking for something like frankenubuntu, mixed versions? [08:42] Sure, there are many unsupported things that are not really broken; just unsupported [08:42] E.g. you may keep security issues for decades and yet not have any intruder [08:42] you can franken ubuntu by adding old repos [08:42] for example, bionic and focal [08:42] What is your question? [08:42] That's what I didn't get [08:43] nothing [08:43] fruity_tomato: maybe have them send you their /var/log/apt/history.log [08:43] could be a good starting point [08:43] i just read what you said [08:43] Ah sorry you're not the one that asked [08:43] frost-core: eh, you're just confusing things :) [08:43] sorry alksig [08:43] alkisg* [08:43] When they don't know what they want, adding more confusion won't help! :D [08:44] sudo apt list --installed | less [08:44] could also be good === javaJake_ is now known as javaJake [09:22] ahoi, is anybody here familiar with the live cd installier ubiquity? i am trying to automate it (ie. click the icon and use a predefined preseed), things are working fine so far and it installs without any interaction, only things I can't get to work is to make it install a minimal system (and by minimal i mean without gnome and webbrowsers, which is what it does when using ubiquity ubiquity/minimal_install boolean true) [09:23] i tried several options that I found when looking at kubuntu/ubuntu/xubuntu ISOs like "casper tasksel/first multiselect minimal" (for kubuntu it uses tasksel kubuntu-desktp) but none of them showed any effect [09:23] any pointers are appreciated :) [09:45] alkisg, Croran not selective update, just wondering what are some good practices around this. It's the first time I have this "problem", where the client sends us their terminal output of their apt-upgrade command (before upgrading) and asking us to confirm if it's okay to upgrade, which includes 100+ packages. My first approach is to take their [09:45] output, parse it, upgrade our environment to match the upgraded versions they have and give them the green light [09:49] well, the thing is also their env is physically different, so for example they might have some packages related to hardware drivers, we dont have. so some kind of cross compare would be needed. [09:49] Seems the "The following packages will be upgraded:" output is kinda parsable, too bad there's no json type of output built in [10:08] Bochi: we hear good things about the unofficial tool cubic to create your own ubuntu images [10:10] lotuspsychje: thanks, that sounds promising indeed === stevenm_ is now known as stevenm [10:10] so do I get it correctly that the installed packages depend on what is in the sqashfs? [10:11] looks like it because under pool there are only kernels, drivers etc [10:12] one needs internet during install [10:14] Bochi, unless the installer changed a lot since i touched it lasst (which in fact has been years), it unsquashes the squashfs to the target disk, chroots into that and then removes all unwanted packages .. this is a lost faster than installing the deb by deb way [10:14] s/lost/lot/ [10:17] ogra: gotcha, that's what I figured (and thats what filesystem.manifeat-remove is for). well great so I will try cubic to create a minimal sqashfs and should be good [10:17] thanks guys [11:28] Hi guys, I can't think or find a better place to ask. Basically I have an ipsec tunnel that will frequently ' [11:28] hang if I don't send dpd packets through it. networkmanager-ipsec does not support dpd out of the box [11:28] can you think of any hack I could use that pings the tunnel gateway when the connection is supposed to be "up" according to networkmanager? [11:29] Or any other type of solution [11:29] I'm currently "forced" to use ipsec.conf to define my tunnel because of this, but this is suboptimal as it interferes with networkmanager's other config such as dns servers etc. [11:33] alive: your question regards ubuntu server? [11:35] desktop [11:36] well, technically networkmanager. [11:38] Hello folks. Currently I'm running into an issue with openvpn, where systemd-resolved doesn't adopt the Nameservers from NetworkManager. Can someone please hint me to a mitigation or at least what the actual problem might be? [11:39] topoi: I had the same problem i think and ended up specifying the dns servers manually under ipv4 settings. [11:40] topoi: Also an other solution is to edit /etc/NetworkManager/NetworkManager.conf and comment out "dns=dnsmasq" [11:41] topoi: did you try the solutions posted here? https://serverfault.com/questions/528773/networkmanager-is-not-changing-etc-resolv-conf-after-openvpn-dns-push [11:58] I have a PC with an old but still fairly high-end graphics card, an AMD HD7970 [11:59] it's kind of half-supported in Ubuntu, it seems [11:59] is there a way to get it to work properly with OpenCL? === JoeBk__ is now known as JoeBk [12:05] alive: Thanks, I really appreciate. I need to clarify, because I mixed some things up: I'm using Ubuntu's GUI for configuring VPN and I can reproduce the problem when using openconnect. Addressing your mitigations: setting it manually shows no effect ("resolvectl dns" lack any DNS settings like those "mcli" shows). In /etc/NetworkManager/NetworkManager.conf there is no "dns=dnsmasq" to comment out and as [12:05] there is no client.conf I cannot change any client.conf. Also the configuration in /etc/NetworkManager/system-connections/ seems fine. Though I can set those manually with "resolvectl dns vpn0 $nameservers" each time I need to get this without manual intervention. === Jubes5 is now known as Jubes === five643333491 is now known as five64333349 === Jubes9 is now known as Jubes === five643333496 is now known as five64333349 [13:47] topoi: https://i.imgur.com/OjdxCNM.png <- does it work to add your dns server here? [13:48] "Additional DNS servers" enter the ip address(es) of your DNS servers [14:25] Hi folks === diskin is now known as Guest3350 === diskin_ is now known as diskin [14:28] hi kaj [14:29] hi frost-core [14:49] frost-core: Are things working well for you now? [14:50] jhutchins : wdym? everything is fine [14:50] also our old pc broke [14:50] and someone had to fix it [14:50] now its fixed [14:50] sooo bad monday for him [14:50] frost-core: Good that it's fine. Good that PC is fixed. What did it need. [14:51] reinstalling windows [14:51] windows kinda broke [14:51] it had xp or something [14:51] frost-core: What did it end up with? [14:51] 7 [14:51] like literally [14:52] i wish that it ended up with linux [14:53] anyway [14:53] this is a support channel, i got confused with offtopic [14:53] sorry :( [15:03] so [15:03] how do i remove a ppa? [15:04] oh [15:04] nvm [15:05] frost-core: delete it from /etc/apt/sources.list.d/ or install and use ppa-purge which will remove the aforementioned file but also remove any packages installed from the ppa, which is recommended [15:05] leftyfb : i used the gui [15:05] instead [15:08] leftyfb : also, how can i install gala wm? [15:08] i wanna replace my openbox in lxqt [15:09] i get a error when adding repositories [15:13] 404 Not Found [IP: 91.189.95.85 80] [15:13] thats what it says [15:13] along with an error about release file not existing [15:15] ooh [15:15] i releasied that its not for impish indri [15:15] realized* [15:16] but its for jammy [15:16] and focal [15:16] so do i have to modify the sources.list file to add focal? [15:17] frost-core: That's the problem with PPAs, there is no quality control, and they're often unmaintained. [15:18] frost-core: You'd be far better off to stick with mainstream sources and projects that are supported by Canonnical. [15:18] i just want a fancy window manager [15:19] frost-core: The minute you step away from the official projects you are left with just the few people who are on those third-party projects, and they aren't likely to be on any irc channel. [15:19] frost-core: Pick one of the mainstream desktops and learn how to use and customize that. [15:21] i tried and it didnt work [15:21] so, yeah [15:21] never trust indris [15:25] frost-core: You would have to contact whatever project is maintaining that repo, and find out what to do about more recent packages. Like === Starmina- is now known as Starmina [16:21] I create a pair ot keys using ssh-keygen. How can I see key's properties? [16:25] RMOTAO: what sort of properties are you looking for? [16:25] ssh-keygen -l -v -f ~/.ssh/id_rsa [16:26] i dont think you will get more info than that from key [16:40] I keep running into the issue after a few days of use "Maximum number of clients reached" when trying to open programs. Is there somewhere to figure out where all of them are being used? === frost1 is now known as frost-core [16:50] https://unix.stackexchange.com/questions/498652/does-x-window-have-a-maximum-number-limit-on-clients [16:50] @octav1a maybe this can help you [16:52] dul: yeah, I've found that while searching, but the thing is that I don't have that many actual windows open. It's like something is 'leaking' them in the background. [16:52] I'm looking for something like "ps" that can show clients connected to a display, so that I can find what is making the issue. [16:53] maybe some app open more socket [16:54] yes [16:54] How can I find out which one? [16:55] try lsof? [16:57] or in stupid way, close one, then test, then another, :) [16:59] xlsclients can list all xclients [17:02] dul: That sounds useful. [17:03] octav1a: What kind of things do you have running? Is this a default gnome desktop? Have you added widgets or extensions to the desktop? [17:03] top will show the most active programs (ps is pretty noisy these days). [17:05] it seems to be an issue with chrome [17:05] xlsclients is saturated with a bunch of WebKitWebProcess [17:06] (many, many more than I have tabs open) [17:07] it's extensions, i guess [17:16] i'm having some serious problems setting up an lvm writecache [17:16] " lvconvert --type writecache --cachevol fast_vol vg-cache/slow_vol" [17:16] fails with Invalid argument for --type: writecache [17:16] i've been digging in to it for hours and the best i can come up with is this [17:16] https://bugs.launchpad.net/ubuntu/+source/lvm2/+bug/1868129 [17:16] Launchpad bug 1868129 in lvm2 (Ubuntu) "I can't attach a writecache to LV" [Undecided, Confirmed] [17:16] It's deprecated afaiḱ [17:17] is there something else i should be using to accomplish a writecache goal? from what I can tell the only option is here flat out don't use ubuntu lol don't know what else i'd use [17:17] really don't want to put gentoo on a vps [17:19] i tried bcache but ubuntu flat out doesn't support that either, the only way to pull it off is during installation with a nasty workaround [17:19] ultimitely that fails trying to write echo 1 > attach, permission denied, and the end result of that rabbit hole was "use lvmcache" rofl === ChunkzZ is now known as jailbreak [17:25] it's nuts reading these bug reports. how can a distro that's sitting on half of the servers on the world just say "welp tough luck no lvmcache" "no bcache either" lol [17:25] these issues are 6 months old, given i'm not running 1000 boxes i'm not exactly sure how enterprise does it, but clearly not with ubuntu rofl, last thing I want to do is start using Arch as a server OS but i have no choice now [17:26] octav1a: Have you said what browser you're using? [17:27] Just "google-chrome-stable" === Jubes6 is now known as Jubes [17:30] octav1a: I would suspect extensions. I've run it for years and don't see that kind of back-up. [17:30] lvmissues: ( cat /etc/os-release ; apt-cache policy lvm2 ) | nc termbin.com 9999 [17:31] jhutchins: even when I close the chrome program completely I still see a whole bunch of "webkitprocess" [17:32] but now at least I can use killall to resolve the issue [17:32] temporarialy [17:34] octav1a: chrome->setting->find keyword "background" [17:34] you can find option "Continue running background apps when Chome closed" [17:34] Yeah, that sounds reasonable, but you need to figure out what's launching them. You might be able to track PIDs. [17:35] Does that mean it could be some electron thing? [17:35] There's also the "tree" function of ps, haven't used that in a while. [17:36] octav1a: Are you running electron apps? [17:36] I will try out that chrome option for a while and see if it helps [17:36] I have slack sometimes, and also authy [17:36] I suspect both are using some kinda browser backend [17:37] (though both of these were also closed at the time of runing commands before [17:37] ) [17:42] this an interesting rabit hole to go down, before i saw that I was investigating putting the ext4 journal on the ssd with data going to hdd, which looked promising but hacky [17:43] lvmissues: ( cat /etc/os-release ; apt-cache policy lvm2 ) | nc termbin.com 9999 [17:44] already installed the version; there was a new version of lvm2 available in backports, installed it and looks like it has writecache support [17:46] lvmissues: ok, so there was no issue. You were running a version of lvm2 which was released several months before the writecache feature even exited right? [17:46] just did a swap out of my htpc. it was nvidia shield -> HDMI -> samsung TV -> HDMI ARC -> vizio sound bar. This was working with 5.1 passthru. Now I have an amd pc, ryzen 5600, renoir graphics ubuntu 21.10 wayland, gnome 40.4. I go to sound settings, select hdmo 5.1 and test, only get sound out of front 2 channels, not center, surround or lfe [17:56] howudodat: You describe your graphics hardware, what about audio hardware? [17:57] howudodat: It sounds like the issue may be with the driver for the audio chipset. [18:00] looks like the audio is AMD renoir hd audio [18:00] pci id: 1002:1637 === genii-core is now known as genii [18:13] lvm2 is still broken even with the upgraded packages, now it's a kernel issue [18:13] howudodat: it seems Renoir AMD audio device have more issue on google [18:14] all the way out to the last part, cache LV and data LV all created and good, go to activate: [18:14] lvchange -a y /dev/vg/data [18:14] device-mapper: reload ioctl on (253:3) failed: Cannot allocate memory [18:14] dmesg: [18:14] device-mapper: table: 253:3: writecache: Unable to initialize device [18:15] howudodat: my G keyword "amd audio 1637 device snd_hda_intel" [18:16] howudodat: Looks like it's supported by snd-hda-intel, which might not be full support. === JoeBk_ is now known as JoeBk === polve2 is now known as polve [18:28] lvmissues: meta volume? [18:28] i'm just trying to setup a writecache volume, 2 drives [18:29] --cachevol cache should but both the meta and the cache on the same cache drive, i wasn't trying to setup a cache pool of multiple drives [18:29] *should put both* [18:29] i found a post, it said lvm cache need 3 parts data+cache+meta [18:30] lvcreate -n data, lvcreate -n cache, lvcreate -n meta [18:30] I'm on LTS at the moment i've got to check to see if this works on mainline [18:31] then use "lvconvert --type cache-pool --poolmetadata vg/meta vg/cache [18:32] i think there's multiple options here, you can split the cache and meta data into multiple devices too [18:32] like what your suggesting [18:32] the cache and meta can be on the same device though [18:32] then use "lvconvert --type cache --cachepool vg/cache --cachemode writeback vg/data" [18:33] i force upgraded lvm2 to get past that command i think that's where the issue is now [18:34] --type cache would work, --type writecache does not, i force upgraded lvm2 to get around that; it worked, but now the kernel won't detect it [18:34] or it detects it but it's all jacked up lol won't activate it [18:34] maybe the kernel module doesn't match now but i'm in uncharted territory bug searching wise when i start force upgrading things [18:38] lvmissues: if you would like someone here to assist you, you should provide pastebins of commands you are using and/or error messages in full [18:38] So, we only have a couple of ubuntu servers at work, so I'm not superfamiliar with "accounts-daemon" - but why the heck is it using ¾s of a gig of ram? This seems excessive. [18:39] can I kill it? or at least restart it? [18:39] nemo: Is this causing issues? Generally, linux looks at ram as being there to use. Unused ram is just wasting electricity. [18:39] jhutchins: it is actually [18:39] jhutchins: it's forcing extra swapping because it's interfering with disc caching [18:40] nemo: a good place to start https://unix.stackexchange.com/a/491551 [18:40] jhutchins: not a huge amount, but the stuff we actually care about needs the rest [18:40] leftyfb: oh. yeah. it's also using an unusual amount of CPU [18:40] nemo: How much system ram do you have? [18:40] thanks; i would but i don't think they'd be useful, not with lvm2 and half a dozen dependencies with non-standard versions installed now [18:40] i think the next step is trying mainline not LTS [18:40] nemo: start with the link I gave you [18:40] leftyfb: kk [18:41] lvmissues: non-standard versions of what? [18:41] jhutchins: 12 gigs, but again, it's being used for other things. and we started examining where our ram was going when we noticed swap usage creeping up === morpheuz5 is now known as morpheuz [18:42] jhutchins: most everything else is on devuan, and I guess we're used to the only things using the ram, for the most part, being the processes we started. But, it's about familiarity I suppose [18:42] leftyfb: anyway. thanks. will read through that [18:42] lvm2, which required dmeventd dmsetup and liblvm2cmd2 to be force upgraded too [18:43] lvmissues: non-standard as you installed from source or a 3rd party PPA or repo? [18:43] backports [18:43] lvmissues: backports is fine [18:43] no custom repo's i just installed by specific version [18:44] that worked, but now i think, just a guess, the upgraded lvm2 doesn't like the kernel module [18:44] lvmissues: installed from official repo's is standard and supported. If you'd like help, please post commands and errors. Otherwise, good luck troubleshooting till you run into an issue you need help with [18:46] i'm gonna try on 21.10 and if that messes up i'll post some debug logs for sure. [18:46] nemo: It seems that the accounts daemon as implemented in Ubuntu is rather questionable: https://utcc.utoronto.ca/~cks/space/blog/linux/UbuntuAccountsServiceProblems [18:47] nemo: You could consider disabling it if you're not using it for cross-system management or something. [18:47] nope [18:48] It looks like one of those projects that somebody decided to implement universally whether it was needed or not, because the old CLI utilities weren't cool enough. [18:49] jhutchins: not sure I would recommend disabling AccountService based on a biased article from 10 years ago [18:52] leftyfb: well. we're definitely not using gnome [18:53] leftyfb: we have a slim+xfce4 session up just so that the used-to-Windows VM admins can get into a desktop if they need it [18:53] leftyfb: we switched to those just to save RAM [18:54] nemo: AccountService is not a Gnome-specific system [18:54] leftyfb: ok. that article claimed gnome was using it for user management [18:54] would xfce use it? [18:55] nemo: the article from 10 years ago that jhutchins linked to doesn't even mention the word "gnome" [18:55] oh. oups [18:56] I was reading one coworker sent me [18:56] https://unix.stackexchange.com/questions/383138/should-i-disable-accounts-daemon [18:57] nemo: if I'm understanding this properly, disabling it would disallow you to use any GUI tools to manage user and group accounts, a one example [18:57] leftyfb: yeah. we don't care about that at all [18:57] so if it's sucking up a significant amount of CPU and RAM, let's get rid of it [18:58] and. seriously WTH at using ¾s of a gig for a pretty gui for users ☺ [18:58] nemo: did you look into the post I have you earlier? [18:58] maybe if it was linked to AD and caching the whole thing [18:58] leftyfb: yes... it was long, and fixing it seems not that critical if I can just safely get rid of it [18:58] Hello everyone. Does anyone know how to set gl_sync_to_VBLAN=0 permanently. I can't seem to get vsync off for my nvidia card [18:58] one less bit of bloat [18:59] gonna start by turning it off in the test system. no worries [19:00] nemo: seems GDM use, please confirm [19:01] dul: yeah we removed GDM 'cause it was also sucking up an insane amount of ram [19:02] dul: switched the 4 remaining ubuntu servers to slim+xfce - saved an astounding amount [19:02] this is just one more improvement I guess [19:02] I might have history of that discussion with coworkers lying around. but it was quite a lot. almost as much as what accounts daemon is vacuuming up [19:03] sounds like more efforts need to be put into adding more memory to the system and optimizing code to not use as much memory as opposed to basically designing your own distro [19:04] leftyfb: heh. the 4 remaining ubuntu machines are running Atlassian [19:04] leftyfb: I have very very very little control over how atlassian wastes its memory [19:04] leftyfb: but there's no reason to have like 1½ gigs of it go to running a barely ever used desktop that's just a fallback for windows admin [19:04] that's just silly [19:05] nemo: why do you have a GUI on a server to begin with? [19:05] but. thanks for links, leftyfb and jhutchins [19:05] 13:53 < nemo> leftyfb: we have a slim+xfce4 session up just so that the used-to-Windows VM admins can get into a desktop if they need it [19:05] hire better admins [19:06] heh. I'm just glad I'm able to run any linux VMs at all [19:06] when I came here there were 0 [19:06]