[00:05] soon I had to quit slow chrome and this will take a while. (been with this problem since 2020. [00:05] tomreyn, and I worked on it a lot. [00:07] back then' Maybe cause you are seiing the problem someone else has an idea of why my browsers slow to a creem and will freeze up the system, effectively. Soing something, twidding their thumbs. I have 8G ram [00:09] Appolinaire: depending on how many tabs you have open, that _can_ be an issue [00:16] https://pastebin.com/E2X2RtQF [00:17] I thought so but then someone was showing over 700 tabs using ff and a tree system of accessing open tabs. I was shocked. I am always under 100 usually under 35 [00:17] under 100 tabs? [00:18] imi, yes, ssh is the only way to log in by default [00:18] Appolinaire: literally the first line says deb cdrom:// ;) so just put a # in front of it [00:18] murmoi about 45 now. [00:18] Appolinaire: well with 8gigs, that's almost a feat :) [00:19] I have seen hundreds on a phone and the user didnt know. [00:19] yeah, but the tabs are getting "unloaded" [00:20] ogra: ok what is the username and the password? [00:20] well I have had that many tabs all the time way back with 2 gigs of ram, years ago. [00:20] iir oc [00:20] imi, no username, no password, the ssh key you put into your login.ubuntu.com account is used [00:20] wow [00:21] mumixam, Bashing-om etc. https://pastebin.com/E2X2RtQF [00:21] Appolinaire: well, probably when js wasn't going rampant ;) [00:21] what is "js is going rampant"? [00:21] and how does it figure out to use my key instead of anyone else's [00:22] imi, it has a first-boot setup tool where you configure the network and give it your login.ubuntu.com mail address [00:22] Appolinaire: javascript, makes browsers quite bloat(y). [00:22] imi, it then downloads the key from login.ubuntu.com ... [00:23] tell me about it. so many automatic videos and moving things. I use ublock origin. [00:23] Appolinaire: I disable js, which reduces RAM usage quite a bit. but some websites are basically unusable. did the commenting out of the first line help? [00:24] https://pastebin.com/E2X2RtQF REGARDING murmel> Appolinaire: verify that it's deleted or has a # in the front in the file /etc/apt/sources.list [00:24] Appolinaire: yes? [00:25] I couldnt find the line murmel so you asked for the pastebin. Took me forever because I named it ---.bak and I couldnt find it... for a while. [00:26] shortcut to pastebin sources: cat /etc/apt/sources.list /etc/apt/sources.list.d/* |nc termbin.com 9999 [00:26] it is now after 5 lol "Everything Takes Longer" that's VanGelder's Rule [00:27] Appolinaire: yes, and if you look at your own link, literally the first line has the deb cdrom:// line, which you need to comment out (put a # in front of the line) [00:27] Appolinaire: you could run `sudo apt edit-sources` then push # then control-x [00:28] or if you're good with sed, that should be trivial too [00:29] I should learn sed. I commented it out with nano. [00:31] thanks enigma murmel Bashing-om bsmith0931 and sarnold [00:31] everything fine with apt update now? [00:32] looks good. updated [00:32] good :) [00:32] cheers. onward [00:33] can this configuration stuff be done completely headless? [00:34] I have no monitor to attach to my rpi [00:34] no, you need a serial cable or a monitor/kbd ... [00:34] serial cable? [00:34] you mean rs232? [00:34] yes [00:35] iirc rpi4 does not have an rs232 connector [00:35] it sure does 🙂 [00:35] where is it? [00:36] yu wire up a usb-serial adapter to the GPIO pins [00:36] like this https://learn.adafruit.com/adafruits-raspberry-pi-lesson-5-using-a-console-cable [00:37] hm... that's cute [00:37] (works on all Pis, the pins differ between models though) === jesopo is now known as jess [00:48] well, if I have an rs232 connection, why am I not allowed to just copypaste a pubkey? [00:52] can I still choose my username on rpi? === plugger is now known as hey_buddy [01:38] imi: raspi-config, username === Dingo is now known as Pristine [02:17] hi! I've been having a recurring issue with my laptop, it often goes unresponsive after I'm being absent for a while, or after I closed the lid [02:17] even usually responsive Magic SysRq commands are unresponsive and I need to hard-reset / long press power button [02:18] I'm on Jammy, but it's been doing it on Hirsute and Focal over the last year [02:18] there's nothing in kern.log (last entry: PM: suspend entry (s2idle)) nor syslog, syslog though sometimes contains ^@^@^@ chars [02:19] I'm using plasma-desktop but it seems a system-wide issue [02:19] afair, it never happened when I entirely disable power saving, or if I manually suspend to ram [02:20] I'm having encrypted rootfs + swap, if it matters [02:22] You mention that there's nothing in the kernel log, but I'm wondering if maybe earlier in the logs that might be helpful. Can you open /var/log/syslog and search for I/O errors? [02:22] sure [02:22] Having the system go into sleep and then panic upon wakeup sounds kinda like maybe the main drive is somehow failing to come back online when the system is woken up. [02:23] (An old laptop did something similar to me once.) [02:23] just before going to sleep I have this: [02:23] kernel: [2572428.552064] Lockdown: systemd-logind: hibernation is restricted; see man kernel_lockdown.7 [02:24] That sounds more like Secure Boot is enabled, though that shouldn't be a problem. [02:24] what's odd though is I have a series of ^@^@^@^@^@^@^@^@^@^@^@^@^@, then I have some more text (I'll paste it somewhere) and only after that, the kernel boot logs appear (at the same second though) [02:25] What determines what kernel updates get added to [release]-updates? [02:25] I'd have expected kernel boot logs to appear first [02:25] kotodama: Yeah, that also sounds kinda like disk problems (corruption?). [02:25] Maybe a RAM error would do that, too. [02:26] kotodama: You might try using this tool to see if your system's RAM is working properly or not: https://www.memtest86.com/ [02:26] ah sure, I'll test it [02:26] one thing that startled me is that it started doing it when I joined Canonical :P [02:26] I created a new profile on my laptop and afair, it never did it when I was on my non-work profile [02:27] but maybe it's a red herring [02:27] Probably :P I've never had a new user cause problems like that. [02:28] (and fwiw, while I don't work for Canonical, I do have an Ubuntu SSO account and various other accounts in lots of Ubuntu- and Canonical-related stuff as part of my volunteer work, and my system's don't behave like that.) [02:28] *systems [02:31] I mean, I dunno, maybe if you're using different software than normal as part of your job, perhaps it's somehow conflicting with something else in just the right way to cause a kernel panic? Like, do you use multiple virtualization software systems on this computer (like VirtualBox and VMWare)? [02:31] I can see that possibly being the cause of a kernel panic if your hardware is unusual. [02:37] arraybolt3: https://paste.ubuntu.com/p/2vQcmjgR2v/ [02:38] this is today's boot logs + logs of when I suspended it yesterday [02:39] kotodama: This doesn't happen to be a modded Chromebook, is it? [02:39] nope, it's an Asus laptop [02:39] (I noticed something about chromeos_pstore, I'm thinking maybe you modified a Chromebook to run Ubuntu.) [02:39] Hmm... ok, still looking... [02:39] thank you :) [02:39] apart from having lxd snap, nothing unusual I think [02:40] and even that might not be unusual :) [02:41] I think I ran a memtest when this first happened ~ a year ago and found nothing [02:41] (over multiple runs during one full night :p) [02:41] I'm seeing some ACPI errors, that might be the problem. [02:42] Lemme see if I can remember the bot command... [02:42] !acpi [02:42] ah ! [02:42] to debug ACPI issues on ubuntu make sure your bios is up to date and follow the procedure here: https://wiki.ubuntu.com/DebuggingACPI [02:42] See if that helps fix things. [02:42] I initially thought it might be encrypted swap, then I moved it from using a random key to a fixed key but didn't help [02:43] le met check that [02:43] (Probably don't do the bit about acpi=off, since I think that may disable sleep entirely...) [02:44] Also, what's the exact model and model number of your laptop? [02:44] I can check to see if it has any known kernel settings that help make it work right. [02:45] it's https://www.asus.com/ca-en/supportonly/ux430uar/HelpDesk_BIOS/ [02:45] I'm one bios version late though [02:47] kotodama: Oh, then I'd definitely give a BIOS upgrade a shot. (Please do yourself a favor and back up your data first if you've not already.) [02:48] hehe :) [02:48] hmm indeed acpi=off will disable sleep, would these also disable sleep ? noapic nolapic pnpacpi=off pci=noacpi [02:48] Not sure. I did find that an option "mem_sleep_default=deep" can make sleep more power-effective, according to https://wiki.archlinux.org/title/ASUS_Zenbook_UX430/UX530 [02:49] ah, interesting :) [02:49] That might shake things into working, if it does you can add it into the right line of /etc/default/grub and then run "update-grub" to make it permanent IIRC. [02:50] yup, I'll do more tests and will give you feedback when I have more details ^^ [02:50] thanks ! [02:50] No problem, glad to help! [02:51] :) [03:21] To make a bootable ISO can you just mount the iso and copy the files to the destination flashdrive? [03:21] I figured it was more involved than that [03:21] https://www.onmsft.com/how-to/create-a-bootable-windows-10-usb-on-linux [03:22] explodes: It is a bit more involved than that. [03:22] (Actually, for UEFI systems it might not be, but for best results you should probably flash the ISO directly to the drive.) [03:23] You can flash an Ubuntu ISO to a flash drive using software like balenaEtcher. [03:23] That's what I usually use, it works great. [03:23] (Note that ISO flashing will delete anything else that is currently stored on the flash drive.) [03:23] explodes: ^ [03:29] balenaEtcher-1.7.9-x64.AppImage is some electron app, when I run it I get "FATAL:gpu_data_manager_impl_private.cc(415)] GPU process isn't usable. Goodbye." [03:30] explodes: Oy. I remember there's a workaround, one moment... [03:30] (Also, gotta love the error message. "Goodbye." ?!) [03:31] explodes: Try "./balenaEtcher-1.7.9-x64.AppImage --disable-gpu-sandbox" [03:34] lul https://imgur.com/a/ot7d6uv [03:43] explodes: And that's from trying to flash Ubuntu? [03:43] Oh... no, you're trying to flash Windows on Ubuntu. My bad, Etcher doesn't work for that. [03:48] explodes: Try using this: https://github.com/WoeUSB/WoeUSB-ng [03:50] I have to manually install a bunch of deps for that === takeshii__ is now known as takeshii === popey0 is now known as popey [04:07] Whats the best Ubuntu flavor? Xubuntu, Lubuntu, Etc?? [04:07] Seeking fast and light [04:08] I want to choose a different cipher other than aes-128 for my Linux Install Drive Encryption, how can I set an alternative cipher? === scoobydoo_ is now known as scoobydoo [05:14] cccccchdgruvbghhcnfeddgrtglludtjkcdlhgftbvcj [05:14] Nailed it [05:15] blahdeblah: ...? [05:15] arraybolt3: Bumped my Yubikey [05:15] * arraybolt3 attempts to pass text through an aes128 cypher [05:15] Oh LOL [05:15] Hopefully that wasn't a security risk, right? [05:16] Nah, it's just a semi-random string that uniquely identifies the hardware key [05:16] Ah, OK. [05:17] So, what I actually came here to ask is: is there a working replacement for gnome-shell-extension-system-monitor on 22.04? [05:17] When I upgraded from 20.04 my system monitor in the top panel broke and it seems there is no official replacement. [05:17] blahdeblah: One moment, lemme find out... [05:18] It seems like there are several different forks/redevelopments of the same thing on the gnome extensions site, but I haven't found one that works yet. [05:19] blahdeblah: I don't see any official replacement, sadly. Man, that looks like a really useful tool too. [05:20] (I see there's a system-monitor-next extension, but I have no idea if it's trustworthy, and I'd guess you probably already looked at it.) [05:22] (Actually, the system-monitor-next extension looks like it's made by the same guy. Did you try it already?) [05:23] That one is the one that ended up looking like the best bet, but I couldn't get it to work. ;-( [05:23] Argh, and I see it's actually a fork. :( [05:24] And the problem is compounded by the fact that now that Firefox is a snap, it doesn't seem like it can install GNOME extensions anymore. [05:24] Though it says it's only forked for packaging purposes I think. [05:24] blahdeblah: Can you use the Extensions Manager? [05:24] I think it's gnome-shell-extension-manager, you can install it with Apt. [05:25] I'm able to install extensions on my system by using it. [05:26] So that's different from gnome-extensions-app? That seemed to be the preferred extension manager now. [05:27] Sorry: "gnome-shell-extension-prefs" === blackest_mamba_ is now known as blackest_mamba [05:27] blahdeblah: I dunno, I'll see which one I have. ALso, I just noticed some of the dependencies of the system-monitor-next extension were missing on my system, so I'm installing them to see if I can make it work. [05:28] * arraybolt3 has to log out and log back in real quick, one moment please... [05:29] blahdeblah: Well that appears to have done the trick, the extension is now working for me. [05:29] I had to run "sudo apt install gir1.2-gtop-2.0 gir1.2-nm-1.0 gir1.2-clutter-1.0 gnome-system-monitor" in a terminal, then log out and log back in. [05:32] I was missing the -clutter- package - I'll give it another shot now. Well, I need to reboot anyway... :-) [05:48] arraybolt3: 100 cool points to you! Working now. Thanks for the help. <3 [05:48] \o/ Glad to be able to help! [05:55] I installed opengl now what does that do for me as far as playing games 20.04 i5 dell with 8G RAM and Xeon E3-1200 v3/4th Gen Core Processor Integrated Graphics Controller [05:55] but I cant figur out if it has any memory on that. [05:58] Appolinaire: it all dependson the video card and the games needs [05:59] Appolinaire not sure what you are asking. If you have old hardware, new games are gonna be slow. [05:59] If that isnt a video card I dont know if I have one. would I see with lspci? [05:59] if you're running a xeon I"m guessing it's more likely a server type hardware ? [05:59] few desktops run with xeons [06:01] it is a dell 9020. [06:01] Good morning. I am running Pop!_OS but this bug has also happened on Ubuntu and even Fedora. So I would just like to ask where to report it. My HP ZBook is unable to use a second monitor unless I revert to kernel version 5.10.119-0510119-generic. I found out about that via a forum tutorial. But would like to report it so I maybe, eventually, can [06:01] update to the newest kernel. [06:01] I don't know what that is [06:02] schoning: for ubuntu launchpad.net [06:02] Thanks ikonia [06:11] Is it required to disable all 3rd party repos in /etc/apt/source.list.d/ prior to upgrade or does do-release-upgrade fix this ? [06:14] do-release-upgrade disables them, no need to do it manually [06:33] Appolinaire: The Xeon E3-12231 v3 chip (and others in the v3 line) do NOT have integrated GPU. However, you're also listed "i5 Dell". A Core i5 does have an integrated GPU. [06:34] samy1028c, thank you for that info. [06:36] You'll need to determine which CPU you actually have installed on that system. [06:36] how odd. My /boot/ does not have enough space to "do-release-upgrade" even after changing compression. Only have 2 kernels there [06:36] elwisp: what does "df -h" show for /boot? [06:37] https://pastebin.com/zU6ffGuW [06:37] odd. It should be enough. [06:38] Do you need both kernels? Perhaps run "apt-get autoremove" first? [06:38] samy1028c, is this enough infor? and why do I need this? - - product: Intel(R) Core(TM) i5-4590 CPU @ 3.30GHz [06:39] I forgot how to find out how much free space I have on my hard drive. [06:39] aah, maybe i am wrong here, i think the compression made enough room [06:39] Appolinaire: df -h [06:39] Appolinaire: yes - then you don't have a Xeon CPU. There are some people who have installed Xeon's into a Dell 9020 but that's not supported by Dell. [06:39] ok I did that but cnat read it I guess samy1028c [06:40] Appolinaire: https://www.intel.com/content/www/us/en/products/sku/80815/intel-core-i54590-processor-6m-cache-up-to-3-70-ghz/specifications.html [06:40] samy1028c, I am just reading form lshw [06:40] Look under Processor Graphics [06:40] heh, nope, after compression it needs even more space. https://pastebin.com/H7nScAYk [06:40] samy1028c, looking on that page? [06:41] The i5-4590 comes with Intel HD Graphics 4600 and you can allocate up to 2 GB vRAM to it (usually set via BIOS/UFEI). [06:42] elwisp: odd it needs that much space. But I usually do an "autoremove" before a "dist-upgrade" anyways so it's been a long time since I came across that problem. [06:42] samy1028c: yeah i have no old kernels, i have already done autoremove [06:43] what version are you currently on and what are you upgrading to? [06:43] thanks samy1028c -- it is too late for me to think, it turns out so till later. good night [06:43] you're welcome Appolinaire [06:45] 20.04.05 and i am going to 22.04.whatever is current [06:48] I just checked. I have a VM that was 18.04 and upgraded to 20.04 and it shows /dev/sda2 974M 205M 702M 23% /boot [06:49] and I have an Ubuntu 20.04 on Azure that has /boot under / but even so is only using 55MB. [06:49] yeah my other box is exactluy the same honestly. Its very odd [06:50] samy1028c: yeah but azure is virtualized, with a specialized kernel, so it makes sense that it's super small (basically no drivers) [06:50] Unfortunately I can't help on the oddities of 20.04 > 22.04 as it's going to be a while before we get there. We've been upgrading 18.04 FIPS to 20.04 FIPS slowly but surely. [06:50] when looking what uses space also https://pastebin.com/9HtNUazj [06:50] uts current kernel and old [06:51] how big is /boot and how big is the complete disk? [06:51] When was the last time you rebooted? Would that help and then you can do an autoremove again to get rid of 5.15.0.46? [06:52] 38 mins ago [06:52] murmel: check my previous pastebin [06:52] murmel: he listed it as /dev/nvme0n1p2 704M 269M 383M 42% /boot [06:52] dev/nvme0n1p2 704M 269M 383M 42% /boot [06:53] its all default stuff [06:54] interesting, as my latest initramfs (and the one before) need 108 megs [06:55] i think i can just delete the old kernel, should work honestly [06:55] elwisp: yeah when you want to upgrade anyway [06:55] and you do know that the latest works [07:00] i wonder if this is a nvidia issue. Might make the image to big with their phat drivers? [07:04] elwisp: over 380 megs oO hm maybe [07:05] upgrade running now however :) [07:06] elwisp: gl ;) === denningsrogue0 is now known as denningsrogue === denningsrogue6 is now known as denningsrogue [07:24] :( it broke my system [07:25] elwisp: what happened? [07:25] Its not even runnign anything custom. I installed it in march i think. So it was realatively fresh. Got some obscure error. It lost network connection also [07:25] elwisp: without logs can't say much ;) [07:25] yeah, i rebooted. now its just spinning wheel. I think this is a reinstall [07:26] elwisp: at least you will probably fix your /boot issue ;) (my fresh install is almost 2 gigs) [07:28] yeah, me neither [07:28] got console now but no network [07:30] yeah this is a very broken system. Sad that this was going to be my first experience running ubuntu in prod [07:30] Came from debian when i bought this machine early this year [07:31] Started networkmanager and still no network [07:37] so i asdasdasdad [07:38] . [07:38] [07:38] . [07:38] oh, shit, sorry about that, terminal hung here [07:39] Anyways, i got the nic going in a manual fashion which allows me to run a apt full-upgrade do-release wont allow me to run it befor eall packages are installed [07:39] I guess this is where zfs shines [07:55] it broke luks :( [07:55] elwisp: I really wonder what happened. especially with only official repos it should work perfectly fine [07:57] yeah, me too, the last thing i can remember installing on this machine was some nvidia stuff [07:57] elwisp: and sadly, I wouldn't use the zfs-on-root ubuntu install anymore, still in a testing phase + zsys was almost not included in 22.04 anymore (which would break installs) [07:57] if you want to go that route, you would need to install it the manual way [07:58] i have another ubuntu machine in the house also. And i upgraded it yesterday only minor upgrade but it had a dependency conflict that i had to resolv [07:58] elwisp: oO [07:58] same no 3rd party repos? [07:58] yeah, not 3rd party [07:59] but i guess this is beyond repair now anyways. Just gotta rebuild :/ [07:59] woke up feeling the urge to mess about with the system today hence the upgrade, i have regrets noe hehe [07:59] now* [07:59] elwisp: it is repairable, question is if you want to go through the motions of fixing it [08:01] murmel: what do you suggest? livecd and chroot in? It wont even boot now [08:01] cant find the luks partition [08:01] elwisp: yeah [08:02] im about to try that now. Just gotta fetch the iso :) [08:02] elwisp: sure :) [08:17] murmel: i think i forgot how to chroot in [08:18] is there an ubuntu way to do this or is it just the good ol mount it and run the chroot cmd? [08:18] elwisp: the old way ;) [08:20] elwisp: dont forget to mount --bind the /sys /proc /dev /dev/pts [08:21] starting out with gnome crapping out in the live cd env. Not my day today [08:22] elwisp: did you nomodeset? (you got an nvidia ;) ) [08:22] no [08:22] elwisp: would be helpful [08:23] i really hope the blob will be distributable :/ [08:24] pertty sure this was just some wizard app that bugged out because i clicked around a little [08:24] trying again [08:25] murmel: how do i do that? [08:26] elwisp: what? [08:26] nomodest [08:26] nomodeset [08:26] elwisp: during boot in grub edit the entry (select the entry, press e, then at the end of the linux line, add nomodeset) [08:27] ok [08:39] arraybolt3: hey ^^ I haven't upgraded bios yet, but then as I rebooted after a memtest, tried some options [08:39] nolapic or noapic immediately freezes computer (sysrq KO, power button KO) on ctrl+x [08:39] acpi=off option freezes on asking for rootfs passphrase screen [08:39] I'll do more tries (removing the quiet / splash screen this time heh) [08:40] alright, i am chrooted in. murmel got any idae whats the next step? :) [08:41] elwisp: verify that your "crypt" stack is using the correct UUIDS, so first start out with /etc/crypttab (which should have the UUID of your /dev/sdXY [08:44] its the right one [08:44] elwisp: what does your /etc/fstab say about the lvm UUID? [08:45] it shows the lvm path for the root [08:45] all looks right to me [08:45] oh right, I forgot [08:46] elwisp: "sudo update-initramfs -uk all" [08:46] btw, the upgrade went through without issues? [09:05] elwisp: still there? xD [09:06] yeah, its not installed [09:06] wierd [09:07] what is not installed? [09:07] update-initramfs [09:07] which is wierd, because i ran it on that system before upgrading [09:07] oO [09:08] Should i just install it? [09:08] yeah [09:08] its so odd however that it would remove that package [09:09] elwisp: I rather wonder if you don't have /usr/sbin/ in your path [09:10] nah its there [09:10] anyways, updated now, just reboot to test it? [09:11] elwisp: yeah [09:12] interessting [09:12] hm? [09:12] i found something googling, dracut install removes initramfs [09:13] elwisp: yeah, but ubuntu doesn't use dracut [09:13] oh, will try a reboot then [09:17] volume group "vgubunut" not found cannot porcess colume group [09:18] vgubunut or vgubuntu? === nomad1 is now known as nomad_fr [09:18] vgubuntu [09:22] elwisp: I remember that 22.04 had an issue with encrypted installs, but I was sure it was fixed, trying to find it on lp [09:26] well i am out of ideas [09:29] I right now can't find the bugreport :( [09:32] https://bugs.launchpad.net/ubuntu/+source/cryptsetup/+bug/1979159 looks somewhat similar [09:32] Launchpad bug 1979159 in cryptsetup (Ubuntu Kinetic) "Cannot unlock encrypted root after upgrading to 22.04 due to use of non-standard ciphers" [Critical, Fix Released] [09:41] elwisp: ugh, yes. and it seems there are multiple issues where it can lead to not finding the vg [09:50] elwisp: yeah, sounds like it's easier to reinstall, rather than poking around what's going on oO. sad day [09:52] Hello, does anyone have or ever had any performance issues running Ubuntu 20.04 on a build with Threadripper 3975WX and Asus Pro WS WRX80E-SAGE SE WIFI motherboard? Running PassMark CPU benchmark yields half of the expected results... [09:54] Guest54: on which kernel are you on? [09:54] murmel it was the default one, let me check again... [09:56] Guest54: well depends on if you run server or desktop ;) [09:57] desktop has the hwe stack enabled by default [09:57] yeah, it is desktop [09:58] 5.15.0-46-generic [09:59] Guest54: I wonder if 20.04 is already "uptodate" enough for zen(2?) [09:59] was 2 or? [10:00] I think it is Zen 2 [10:00] What do you mean though? [10:01] I've had another problem with this particular board when the SATA disks wouldn't appear until SR-IOV was disabled or the BMC patched to the latest version that ASUS had [10:01] Guest54: your cpu on zen2 came out after 20.04, and additionally, I can remember that amd wasn't that great back in the day in upstreaming before release of the product (it got definitely better since then) [10:02] on the other hand, I do remember that zen2 was already out quite a bit itself :/ [10:03] Guest54: can you boot up a 22.04 live ISO and try passmark? [10:03] yeah, the interesting thing is that we can get around 60k in passmark for a 3970X build and 5950X [10:04] sorry, for 5950X it's the same as for this 4975WX ~44k [10:04] sorry again, 3975wx, sausage fingers :)) [10:04] :) [10:05] huh, that's weird then. AFAIR those are also on the same zen2 (besides being more hampered than the pro threadripper) [10:06] do you think switching to 22.04 could help? [10:06] currently stuck with 3 of the same build and their performance is less than of the old 3970X we have [10:07] IIRC 22.04.1 also uses the same version of the kernel [10:07] tbh, I don't even know where to start digging for the potential source of the problem [10:07] Guest54: sure but the surrounding compilers are newer [10:08] and it's a quick test if something is going on with 20.04 or the mb ;) [10:08] I have seen boards devestating cpu performances [10:08] but it has been a while [10:08] the board we had before wouldn't post at all, had to return it [10:09] (same model) [10:09] ugh, wow. sounds like a "Pro" mainboard [10:10] and last time i checked it was the only option for this particular socket back then :/ [10:11] to be honest, i can't really do anything with it right now, it is in another office and i have to wait for them to put it back on next week [10:11] yeah, HEDT is quite dead [10:12] i mean at this point i am even considering to ask them to install windows or another distro and repeat the benchmark [10:12] but this behaviour is consistent on two different machines [10:13] consistent in 50% performance? [10:13] yes, it stays around 40k in passmark on both PCs, with low of 38k and high of 46k on a good day [10:13] as opposed to 62k i could hit on 3970x [10:14] now it's not just passmark, it seems to be affecting the software we use as well [10:14] did the latest intel microcode fix retbleet? [10:15] oerheks: afaik yes [10:15] that could explain the drop for zen/zen2/zen3, 50%++ [10:15] we run a video capture and process each frame, with 3970X it would be consistent ~50fps but with 3975wx it is around 25fps at best [10:15] but why this particular model :S [10:16] Guest54: I assume you checked that cpu temps are okay? [10:17] they stay around 80-85c [10:17] which I think is considered ok [10:18] yeah max is 90, just can't remember when it starts to throttle [10:18] we also checked if memory speed was a major factor, turns out DOCP (XMP) wasn't enabled, but this didn't help at all [10:19] i would honestly try 22.04 (as it's a fast test) and then win10/11, but afaik windows doesn't have a great scheduler for ccx configurations [10:19] if 22.04 I would probably load the live ISO into ram, so there is definitely no bottleneck [10:20] but sounds like there is definitely something wrong, (I mean outside of sw) [10:20] yes, that is the plan for when they are able to bring this system back online [10:20] but just my intuition [10:20] How to check luks encyprted disk via fsck? its ext4 [10:20] i was also considering using 18.04 [10:20] Intelo: unlock luks and then run fsck [10:21] thank you [10:21] Guest54: eh, I wouldn't, but it's also on the same kernel ;) if we are talking about that [10:21] I'm on 18.04. I just ran "sudo update-ca-certificates" and it did something (removed a bunch of old certs). Isn't that supposed to run automatically, when new certificates are installed or removed? [10:22] My actual issue is that curl doesn't recognize letsencrypt certs, while firefox does. curl uses /etc/ssl/certs/ca-certificates.crt, which is created by update-ca-certificates. [10:23] if I download the letsencrypt R3 cert and use it with --cacert, it works, but isn't letsencrypt supposed to be in the cert store by default? [10:25] TomyWork: maybe? https://letsencrypt.org/docs/dst-root-ca-x3-expiration-september-2021/ [10:25] murmel whats the path to disk ? [10:25] which maybe implies that the new one is not available for 18.04? [10:25] the /dev/.sss? [10:25] my original one is /dev/sda [10:25] but after unlock. i do not know what it is [10:25] murmel, that was a year ago though and certs from other hosts work [10:25] Intelo: /dev/sdX or /dev/nvmeXXX [10:26]