/srv/irclogs.ubuntu.com/2022/01/07/#ubuntu.txt

tomoswoah01:02
=== scoobydoo_ is now known as scoobydoo
=== chrfle_ is now known as chrfle
=== M4he is now known as mahe
Hashhi03:50
Hashanyone else using the next lts?03:50
frosthello,  if anyone have interseting on freenode channel #worldhacker? if no, please ignore03:52
HashI doubt anyone even really connects to freenode anymore03:53
Hasheverything moved from there to somewhere else03:53
frostyeah, i know03:53
Hashirc.mozilla.org has lots03:54
Hashalso irc.oftc.net or something03:54
Hashand check out element.io / matrix network03:54
Bashing-om!next | Hash03:55
ubottuHash: 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.03:55
Hashfty03:59
Hashty03:59
theb-lobbeen reading about firejail and apparmor-has anyone used both at the same time? Have you had any conflicts-could you resolve?04:15
theb-lobubuntu 21.10 or 20.04.304:16
sarnoldyou may need some additional ptrace privileges on the profiles, so firejail can ptrace the processes, but I think it ought to work04:23
sarnoldI've not tried it myself04:23
masayahi04:42
=== genii is now known as genii-core
mybalzitchdoes transgui / transgui-test (snap package) completely crash wayland/plasma for anyone else?05:08
Swift110-mobilesup folks05:40
HashHellop05:53
Hash20.04 ubuntu and I firefox cannot see my mic, but chrome can05:53
Hashideas?05:53
Hashfirefox always says on any site, no camera found05:53
Hashchrome asks for permissions and then works05:53
Hashcamera/mic05:53
HashSeemns like chrome can see my mic but nothing is recorded05:59
Hashstatic05:59
HashI deleted most packages from the system07:08
Hashpurged configs07:08
Hashreinstall 90% of system packages and now my desktop/sound/panels/ etc. is back07:09
Hashwell that was weird.07:09
ggnoredoI'm on 20.04 and want to install a package from jammy repo. How do i do this without adding jammy in to my sources list. The package i want to upgrade is maildir-utils. I don't want to add jammy because it makes 4000 updates on my apt update fetch07:42
trafficjamhow do i remove any versions of a .deb i might've installed (and all associated systemd files)07:44
WickedDekciwtrafficjam, dpkg -P07:47
trafficjamanyone around?08:00
WickedDekciwtrafficjam, dpkg -P08:03
trafficjamWickedDekciw: hmm doesn't seem to work08:04
trafficjami mean the package is there, but I keep getting a connection refused error when i try to curl it08:05
trafficjamsysytemd status is active, so i'm not sure what the problem is08:05
trafficjamis the service running twice?08:08
trafficjami've stopped the service, but `ps aux | grep <name>` still shows it08:09
trafficjam(before it showed two)08:09
trafficjamhow is it possible to have two processes running?08:10
trafficjam*the same two08:10
trafficjami can't kill it it seems08:11
WickedDekciwtrafficjam, sudo apt install --fix-missing08:12
WickedDekciwwait08:12
WickedDekciwdid you dpkg -p packagename08:12
WickedDekciwyou have to include the name08:12
trafficjamWickedDekciw: it ran successfully before, but i still had two of the same processing running simultaneously08:13
trafficjamit's possible that the process is running outside of systemd, how could i check that?08:13
sabas3dghhello guys08:15
sabas3dghI want those squeakily lines under spell checked error words inside hexchat. I installed enchant hunspell and linbspell but no luck. Could you help?08:17
ELQEYNNI have icons in Activities. On right-clicking them, is there supposed to be an option to put it on the desktop?08:33
tuxinatorHi, can somebody please check if https://bugs.documentfoundation.org/show_bug.cgi?id=146633 also happen on his System?09:12
ubottubugs.documentfoundation.org bug 146633 in LibreOffice "LibreOffice Draw does not detect Electronic Signature in PDF with valid skribble.com signature" [Normal, Unconfirmed]09:12
=== shuck is now known as Guest4881
* worldhacker-org come to irc.worldhacker.org channel #0dev , get lost frog from irc.worldhacker.org , By The Most High Lord Jesus Christ with [ 0day (xc) Our ] 09:34
=== via_c7_ is now known as via_c7
=== Guest3387 is now known as krabador
user03kek11:14
=== riderjj is now known as juanjo
mncheckgiven a package how do you go to the vcs repository where eg. its debian control files are located?12:55
* DUPS any spammer here?13:10
lotuspsychjeDUPS: this is the ubuntu support channel13:21
BluesKajHi folks13:32
fricklermncheck: it depends, some are in launchpad, some in debian, some elsewhere. a good first step is to check e.g. https://code.launchpad.net/ubuntu/+source/python3.913:33
fricklersometimes apt-get source gives additional hints, too13:33
mncheckfrickler, how do I know which debian release is relevant for 20.04?13:41
mncheckfrickler, thanks for the info so far13:42
ioriamncheck, cat /etc/debian_version13:43
mncheckioria, thanks13:45
ioriaok13:45
mncheckfrickler, ioria thanks, I've found a debian source repository for my package13:50
mncheckwhy the bullseye version of a package is different than the focal version? I have it from universe13:52
geirhaubuntu releases don't follow debian releases13:54
ioriathere is no direct mapping13:54
mncheckgeirha, ioria: I found the salsa repository from the .dsc of my package. I got the .dsc via apt-get source after enabling the focal universe deb-src in apt/sources.list and I can see the .dsc on archive.ubuntu.com/ubuntu/poool/universe for the package. from where does the .dsc get to archive.ubuntu.com?14:05
mncheckI ask because the .dsc lists a Version: field that is lesser than what is available on salsa, but I don't see how the dsc is filled14:06
ogramncheck, ubuntu imports from debian for a certain part of the release cycle ... eventually syncs stop and packages get stabilized and eventually released ... once released they do not get synced anymore (there are stable release updates for single backported bugfixes and such though)14:08
ogramncheck, focal was released in april 2020 ... bullseye end of 2021 ... so indeed their package is newer14:09
ogra(or mid-2021 FWIW)14:09
mncheckogra, my /etc/debian_version says bullseye14:09
ograon your debian system ?14:10
mannequinbullseye is excellent14:10
mncheckogra,  DISTRIB_CODENAME=focal (lsb-release) and bullseye/sid (debian_version)14:10
ogramncheck, sid is the point here 🙂14:10
mncheckogra, ok14:10
ograubuntu syncs from unstable usually14:11
mannequinsounds like it's not updated14:11
mncheckogra, so is there some build server somewhere where there is an artifact where I can see a sync from some sid repo into archive.ubuntu.com ? I've google "release-cycle" but couldn't find this process yet14:11
ograwhatever codename of whatever debian release was in the base-files package does really not matter ... just means it was synced from sid during that development cycle of debian ... but does not point to any other relationship14:12
mncheckogra, now I found the debian control files but my interest shifted into who authors the dsc files and how14:13
ogradebian does ... ubuntu occasionally adds patches on top though14:13
mncheckogra, is the sync done through rsync?14:14
ogra(there are a few packages that are ubuntu only, but the vast majority comes from debian as source package imports)14:14
mncheckogra, I think I go into #ubuntu-discuss with this14:16
ogramncheck, https://wiki.ubuntu.com/DebianImportFreeze14:16
yaci just got a new 1tb ssd drive and setup as ext4. this is what's showing with df   /dev/sdb1       916G   72M  870G   1% /home/ssd214:59
yacwhere's the missing storage space ?15:00
yacshouldn't it be 914G of unused space instead of 870G ?15:00
oerheks1TB is raw, partitioned one loses 8%, including the part reserved. ( reserved for recovery purposes). then the rest needs to be formatted, losing a few more.15:03
=== genii-core is now known as genii
oerheksso your number look healthy15:04
yacoerheks: the usable space is 931g in gparted, then 916T15:04
yac916G15:04
oerheksboth are correct, and the same somehow15:05
oerheks1024 or 1000 thingy15:05
yacgparted showing size 931Gb with unused 916GiB. how come it shows 870G ?15:06
yaci would expect at least 916G unused with df, not 870G15:06
yacoerheks: i understand about the 1000/1024 bits to bytes15:06
oerheksdf -h  vs du -sh /15:08
yacbut showing that it's got 916G available, with 72M used, then 870G unused does not make sense15:08
oerheksinodes take that space15:08
tuxinator:oerheks for du i prefer ncdu ;-) really cool15:08
tuxinator:oerheks inodes take that space? i am pretty sure root reserved blocks take that space and not inodes, else i would be wrong since 20 years15:09
yacdf: invalid option -- 's'15:09
yaci've always thought the unused+used part add up to available space with df -h15:10
oerheksdf -h  'versus' du -sh /15:10
tuxinatoryac:  look at tune2fs -l /dev/sda115:10
tuxinatorwhere /dev/sda1 is your device15:10
=== Guest4881 is now known as shuck
tuxinator:yac look for Reserved block count:15:11
=== shuck is now known as Guest2951
yacdu -sh /home/ssd2/15:11
yac1.2G    /home/ssd2/15:11
yacthat's 1.1G i just copied to it15:11
tuxinator:yac then use tune2fs -m  to adjust it15:12
yacReserved block count:     1220624615:12
yacReserved GDT blocks:      96515:13
tuxinatori forgot, please also tell block size15:13
tuxinatorlike explained here: https://docs.cloudera.com/cloudera-manager/7.2.2/managing-clusters/topics/cm-decrease-reserved-space.html15:13
yacBlock size:               409615:13
yacoh well that makes sense now15:14
tuxinator12206246 * 4096 =15:14
tuxinator49GB15:14
yacBy default, the ext3 and ext4 filesystems reserve 5% space for use by the root user15:14
yacso 916*.95 ~= 87015:14
tuxinatorexactly and that makes sense, so you can still login as root localy in case of full disk15:15
yacerhm i never do15:15
yacis there any much downsize to shrinking that down ?15:15
tuxinatoryac: on /home you don't need that root reserved space as on login root's homedir is /root/ which is not under /home/15:16
oerheksonly when you need a recovery, that 5% is reserved for disasters15:16
tuxinatorjust do tune2fs -m 0 /devicename15:16
oerheksi would not touch it, you gain not much space15:16
tuxinator:oerheks there is actually no drawback on /home15:16
oerheksthis space has nothing to do with home15:17
yacoerheks: my free space on /home is 7gb. gaining 40-50gb would help15:17
yacoerheks: what is that space for ?15:17
tuxinatoroerheks: consult the manual next time, as no "root-process" writes to /home there is no problem, see https://www.redhat.com/sysadmin/disk-space15:18
tuxinatoryac: change it to 1% and check how much its then15:19
yacyeah i see that recc for cloudera15:19
oerhekstuxinator,  it is the 'Reserved block count parameter' set on 6% in your url15:20
oerhekserr 5%15:20
oerheksagain, i would not touch it, one better removes unneeded packages wit clean/autoremove15:21
tuxinatorthey explain in which case it is needed and in my 21 years of Linux Experience i never met somebody who said don't do that for non / partitions or lets say partitions where no root access is actually ever made ;-) you are a premier :D15:21
yacmy /home is on a separate partition from /15:22
tuxinator:oerheks why would you not touch it? bring a plausible reason, everybody does that, thats a default task you usually do on creation or as a cleanup task when forgot15:22
oerheks'everybody does that' is certainly not true.15:23
oerheksthere is a reason why one should not touch it, i keep repeating myself15:23
=== Leoneof|2 is now known as Leoneof
oerheksbut go ahead, tune it and have fun!15:24
tuxinator:oerheks just keep cool, any case where root would need 49 GB of space in case of full disk?15:25
grkblood13Does sparse not work on LVM partitions? I'm trying to omit the zeroized space and my final dded image is the same size of the drive.15:25
tuxinatoryac: also see the documentation: https://www.linux.org/docs/man8/tune2fs.html15:30
tuxinatorsearch for reserved-blocks-percentage and read that part15:30
tuxinatoryac: also see Arch's recommendation of changing it for /home here: https://wiki.archlinux.org/title/ext4#Reserved_blocks15:31
yactuxinator: oerheks: thank you. i will read through everything before making changes15:36
yacat least now i know it's lost space by design vs. something's wrong with my system15:36
yacor the new drive15:36
jkctuxinator: Out of the "everybody" that I know, which is root only knows how many Linux professionals across hundreds of companies, precisely zero change the default reserved space parameter. So be careful using the word "everybody."15:37
tuxinatorjkc: i talk about all everybody i know and whenever we hit the term, even the ext4 developer says it's ok and Arch the oldest distro also15:39
jkcYour logical fallacy is "appeal to authority."15:39
jkcDefaults exist for a reason. IF you have an articulable and defensible reason to diverge from that, then go nuts. But "everybody does it" is not such a reason.15:40
tuxinatorjkc: well not sure what the problem is, there are multiple opinios and thats ok, however i am a person that likes facts and don't likes senseless opinion discussion, and no, the default has just forgot to adjust15:40
tuxinatorlook at https://wiki.archlinux.org/title/ext4#Reserved_blocks and follow the link to the ext4 author15:41
tuxinatorand feel free to start an endless loop, i am out15:42
oerhekswe agree that this is a  senseless opinion discussion.15:42
oerheksyac got answered, there is nothing wrong with his ssd15:42
tuxinator:oerheks oh you are still here ;-) agreed!15:44
tuxinator:oerheks so lets skip and search the next :D15:44
oerhekssure, lets have fun15:45
tuxinatorhopefully there is one more today i am hot for it :D :D :D :D15:45
oerheksalmost beer o'clock here15:45
mannequinthis a an acoholic's apartament https://vlepy.github.io/2022/01/07/setup2.html15:45
mybalzitchMURICA15:46
mannequinnotice the flag is upside down15:46
lotuspsychjenot here mannequin15:46
tuxinator:oerheks here also15:46
tuxinatorlocation?15:47
=== denningsrogue3 is now known as denningsrogue
gneeriiloeepdeerwhat subtitle app can I use to recognize and convert mks or sub to srt? subtitld doesn't recognize ny of the 2 first formats16:26
gneeriiloeepdeerany*16:26
jhutchinsShould be something in the ffmpeg suite.16:43
gneeriiloeepdeerok...16:44
coz_gneeriiloeepdeer, hmm didnt think of ffmpeg ...checking commands16:44
=== beaver is now known as pong
merpnderpLooks like one of my security fixes broke a remote mount. Any idea how I'd try to manually mount an entry in fstab?16:56
EriC^^merpnderp: sudo mount /mountpoint16:59
merpnderpThanks EriC^^16:59
merpnderpermegerd, sudo is not taking like 30-60 seconds to auth17:00
merpnderpWe finally get a chance to move off windows to Ubuntu and the first thing I'm assigned to do is completely junk up ubuntu17:01
EriC^^no problem merpnderp17:03
merpnderpAnyone know which port a nfs mount would be on17:14
merpnderpI think it's an overly restrictive firewall setting17:14
merpnderpWe deny all incoming outgoing and routed. And then have a rule for every bit of traffic.17:14
merpnderpNoticed that on login there was message that the machine could no longer even connect to its repo servers17:15
merpnderpAnyone know how to set ufw to allow apt to do updates?17:19
merpnderpApt just worked17:20
merpnderpBut couldn't hit any of the mirrors. Timed out on them all. So looks like ufw is the culprit. But how to configure it to allow apt to hit mirrors?17:21
lotuspsychje!firewall | merpnderp17:22
ubottumerpnderp: Ubuntu, like any other Linux distribution, has built-in firewall capabilities. The firewall is managed using the 'ufw' command - see https://help.ubuntu.com/community/UFW | GUI frontends such as gufw and ufw-kde also exist. | An alternative to ufw is the 'iptables' command - See https://help.ubuntu.com/community/IptablesHowTo17:22
merpnderpubottu: that doesn't help as I need to know which port apt uses.17:23
chazzoHello. Looking for direction, want my server to email plain text generated from an HTML form entry to a gmail address. Confused on how I'd get Ubuntu Server to send emails and what address those would even be originating from17:26
jkcThe send emails part is typically postfix in most deployments. The from address would assumably be noreply or something @ a domain that you own.17:39
mbeierlmerpnderp, apt uses HTTP as a transport, so it's either 80 or 443, depending on the mirror17:49
mbeierlfor the origin port, that would be a randomly assigned ephemeral port17:50
jkcmerpnderp: Deny all outgoing traffic can be an absolute bear to config and admin.17:50
jkcBut, if you're absolutely sure you want to do that, `sudo ufw allow out 80,443/tcp` should do the trick.17:51
marc_Anyone know how to get ubuntu to use GEN3 cards bios is set to GEN3 but its running as GEN2 in ubuntu18:08
marc_Bus Type:PCI Express x16 Gen218:09
marc_PCIe Generation:Gen218:09
merpnderpAwesome thanks for the help18:09
marc_Maximum PCIe Link Speed 5.0 GT/s18:09
jkcmarc_: Is the card itself a gen 3 card?18:09
marc_yea its gen318:09
jkcWhat card is it, exactly?18:09
marc_GTX78018:10
marc_pci slots are GEN3 as well18:10
jkcYeah, that's a gen 3 card.18:10
jkcWhat motherboard?18:10
marc_my NvME are also not on GEN3 as well by the looks of things18:10
marc_SUPERMICRO-X9DRH-iF18:11
marc_x7918:11
marc_E518:11
marc_dual processor18:12
jkcWhat CPUs?18:13
jkcE5, but which model?18:13
jkcAnd version.18:13
marc_currenlty 2 x E5-2643 v2 i think but i got E5-2687W v2 here and im about to buy the second E5-2687W v218:14
jkcEverything involved supports PCIe 3.0.18:15
marc_yep18:15
jkcOther than confirming that BIOS has EVERYTHING configured to 3.0, I'm at a loss.18:15
jkcI'd completely reset BIOS if it still wants to act up.18:15
marc_the way i see it it has to be enabled in ubuntu same as windows 1018:16
jkcNo.18:16
jkcMaybe like... years ago, but I'd be stunned if that was the case now.18:17
jkcWhat version of Ubuntu?18:17
marc_if i got to windows 10 which i dont have installed and change the pcilink speeds in the registry this system works on GEN318:17
marc_5.13.0-23-generic #23-Ubuntu SMP Fri Nov 26 11:41:15 UTC 2021 x86_64 x86_64 x86_64 GNU/Linux18:17
jkcUbuntu version, not kernel version.18:18
marc_how do i get its version18:18
jkccat /etc/os-release -> VERSION18:18
marc_i tried that the cursor is just blinking in the terminal18:19
ogragrep VERSION /etc/os-release18:20
ogratry tht oe18:20
geniihit ctrl-c and doit again without the -> VERVION npart18:20
ogra*hat one18:20
geniimeh typos18:21
marc_PRETTY_NAME="Ubuntu 21.10"18:21
marc_VERSION_CODENAME=impish18:21
marc_VERSION="21.10 (Impish Indri)"18:21
jkcLast time I remember having to force a card to a specific PCIe revision was over six years ago.18:21
jhutchinsmarc_: I wonder if you might find anything about PCI handling at kernel.org, since this is unlikely to be an Ubuntu specific issue.18:21
ogradid you check for a BIOS/UEFI update for the board ? might simply be a shrtcoming in the BIOS18:22
marc_ubuntu update just came through seems to be a lot of updates over last two days18:22
marc_there is only one update for this board and thats for the exploits that slow down the machine18:23
marc_i tried this it didnt work modprobe nvidia ENABLE_PCI3=118:26
marc_thats what they were going on about on one site18:26
jkcmarc_: That's a very important update, and "slow down the machine" is an awfully universal statement for an effect that is far, far from universal.18:28
marc_I used to disable it in windows 10 it slows the processors too much18:29
marc_spectr18:29
marc_thats the one i disabled18:29
jkcYeah, don't disable that.18:29
marc_i have18:29
jkcThen re-enable it.18:29
jkcThose are fairly nasty vulnerabilities with real-world exploits.18:30
marc_i think its disabled in ubuntu by default18:30
jkcEverything I run is in VMs, and the hypervisors have mitigation enabled.18:30
marc_i was looking through code in ubuntu yesterday and noticed its disabled18:31
jkcAnd no, they're enabled by default.18:31
jkcAs they ought to be. If you're disabling them because "they slow the CPU down," you're making a pretty significant mistake.18:34
* DUPS anybody have leads @hotmail.it?18:34
marc_just turn off all the mitigations after checking18:40
marc_thats whats been hitting my game performance18:40
* DUPS any spammer here?18:57
ograDUPS, none beyond you .... (please stop that)18:59
marc_I made a file in /etc/modprobe.d called nvidia.conf then added the line NVreg_EnablePCIeGen3=1 saved it then did update-initramfs -u its throwing an error anyone know why19:01
marc_i alos tried nvidia NVreg_EnablePCIeGen3=1 still threw an error19:02
marc_nvidia              35311616  453 nvidia_uvm,nvidia_modeset19:03
jkcWould help to know what the error is.19:06
marc_https://askubuntu.com/questions/1073006/how-to-enable-pcie-gen3-on-ubuntu-16-04-04 this is what i am trying to do19:06
jkcmarc_: If you're disabling security mitigations, you are making a mistake. That wasn't a question.19:06
jkcmarc_: And don't use instructions for a five year old, no longer supported release.19:06
marc_libkmod: ERROR ../libkmod/libkmod-config.c:657 kmod_config_parse: /etc/modprobe.d/nvidia.conf line 1: ignoring bad line starting with 'NVreg_EnablePCIeGen3=1'19:06
TheCodeTherapy@marc_, try using 'options nvidia_453 NVreg_EnablePCIeGen3=1' (without the single quotes) instead of the line you mentioned in your /etc/modprobe.d/nvidia.conf and then sudo update-initramfs -u19:06
TheCodeTherapy@marc_, here I'm assuming 453 is the version of your driver. Edit it accordingly.19:07
geniimarc_: Did you write in the file "options NVreg_EnablePCIeGen3=1" or just "NVreg_EnablePCIeGen3=1" ....19:07
marc_just NVreg19:08
marc_libkmod: ERROR ../libkmod/libkmod-config.c:657 kmod_config_parse: /etc/modprobe.d/nvidia.conf line 1: ignoring bad line starting with 'nvidia_453'19:08
geniimarc_: Yes, you need add the options at the beginning19:09
ograisn't th module name just "nvidia" not "nvidia_453" ?19:09
marc_options nvidia NVreg_EnablePCIeGen3=119:09
marc_ this worked19:09
marc_thanks19:09
marc_Ok people here told me that it was BIOS that was at fault for gen3 its not its Ubuntu thats at fault i got the system running on Gen3 with a screwup19:21
marc_after i typed sudo update-initramfs -u19:22
marc_the system wouldnt boot with 5.13.0-23-generic19:22
marc_its now booted with 5.13.0-22-generic19:23
marc_so how do i fix that19:23
marc_Bus Type: PCI Express x16 Gen319:23
marc_PCIe Generation:19:23
marc_Gen319:23
leftyfbmarc_: wait, so you got Gen3 working? What was the issue?19:23
marc_Ubuntu was the issue19:24
leftyfbmarc_: can you be more specific?19:24
marc_i added a file to /ect/modprobe.d called nvidia.conf19:25
marc_then added to nvidia.conf with help from this forum19:26
marc_options nvidia NVreg_EnablePCIeGen3=119:26
marc_then did sudo update-initramfs -u19:26
marc_rebooted19:27
marc_it wouldnt boot on 5.13.0-23-generic19:27
leftyfbmarc_: ok, so you needed to enable gen3. How is that a "screwup" of ubuntu? Do you think that driver should be loaded and enabled for every install on the planet even if they don't have this feature or an nvidia card?19:27
marc_so i selected 5.13.0-22-generic and its now on with gen3 working19:27
marc_because people here were telling me that it was my bios at fault  and i knew it had to be enabled19:28
ograwell, thats a bit unfair ... the nvidia driver is proprietary ... that it fails to detect GEN3 is not ubuntus fault19:28
marc_same goes in windows 10 got to be enabled19:28
leftyfbmarc_: ok, so it was neither the fault or your BIOS nor an "ubuntu screwup". You just didn't enable the required feature for your driver19:29
ograno ubuntu developer has access to the driver code ...19:29
marc_i alsp got nvme drive that are running like shit and i bet they are not enabled on gen 3 as well19:29
leftyfbmarc_: Windows also requires very non-standard customizations in order to get it working. Sounds like this is the fault of Nvidia not documenting their products very well19:29
marc_nvidia supplied a file for windows 10 there is no mention for linux19:30
leftyfbmarc_: please watch the language in this channel19:30
=== Syntax-1 is now known as Syntax-
TheCodeTherapy@marc_, people here are just trying to help you with your problem (and it worked). I fail to see how this makes your issue "Ubuntu"19:30
leftyfbmarc_: again, that is Nvidia's fault, not ubuntu19:30
marc_so how do i get this back to 5.13.0-23-generic it wont boot on that19:30
leftyfbmarc_: what error do you get when you boot to -23?19:31
marc_intrid not found or something to that effect19:31
oerheksand what fancy GPU need such override 'options nvidia NVreg_EnablePCIeGen3=1' ??19:32
marc_last command i used was sudo update-initramfs -u19:32
marc_Palit Super Jetstream GTX780 3GB19:33
leftyfbmarc_: sudo apt install --reinstall linux-image-5.13.0-23-generic linux-modules-5.13.0-23-generic linux-modules-extra-5.13.0-23-generic19:33
ogramarc_, try: sudo update-initramfs -c -k all19:34
ogra(or what leftyfb said ... should have the same effect)19:34
leftyfbogra: reinstalling the kernel should run all that properly19:34
ograyeah19:34
marc_ok thanks i will try that19:35
marc_would there be a way to enable gen3 on my Nvme drives19:36
marc_using same procedure19:36
marc_the nvme drives are plugged into 8x Gen3 slots19:36
oerheksHow do you tell they are not running on the pci v3.0 speed?19:41
marc_first how would i check that19:42
oerheksyes, tell us19:42
marc_they are slotted into pcie slot and in the bios they are set to GEN3 but i need to check i could see the graphics card wasnt set to Gen3 in the nvidia panel19:43
marc_is there a command i can use to see the link speeds19:44
marc_my drives write speeds are only running not even half speed19:44
oerheksi wonder what hardware gives 8Tbit speed to test19:45
oerheksyou might want to try a testsuite package, like phoronics https://phoronix-test-suite.com/releases/repo/pts.debian/files/phoronix-test-suite_10.8.0_all.deb19:47
marc_the nvme is running at 348.8 MB/s and it should be somewhere near 2000 i think19:47
marc_my average read spped is 3.1 GB/s19:48
orange1thats fast19:53
marc_348.8 MB/s write speed is slow20:01
orange1how do you test the speed?20:04
orange1i want to test mine20:05
marc_you can look at it with the disk utility or i have used dd20:05
merpnderpAnyone know what subnet I would use to grant ufw access to everything in 10.67.5?20:05
orange1ok20:05
marc_you have to tick the box for write speed under disk utility20:05
WaVmerpnderp: something like "sudo ufw allow proto tcp from 10.67.5.0/24 to any port 1234"20:07
merpnderpWaV: will sudo ufw allow 10.67.0/24 work?20:07
merpnderpThen narrow it down later?20:08
merpnderpI'll just try it and find out20:08
merpnderp:)20:08
merpnderpNo that does bad port.20:08
WaVmerpnderp: I believe the syntax is "sudo ufw allow from 10.67.5.0/24 to any"20:08
merpnderpoh, thanks :)20:08
merpnderpWaV: what about allowing outbound to that subnet?20:09
mannequinhttps://vlepy.github.io/2022/01/07/ubuntu.html20:09
WaVmerpnderp: I believe ufw should allow that by default unless you've changed it.20:10
merpnderpWaV: yeah, to meet requirements from a security scanner my first three config lines are deny incoming, outgoing, routed20:11
merpnderpSo I have to specifically poke holes20:11
merpnderpBut this didn't work: sudo ufw allow to 10.67.5.0/2420:11
TheCodeTherapy@marc_, First use lspci to find the ID of your nvme. Let's suppose your nvme is at 00:01.0. Then use 'sudo lspci -s "00:01.0" -vvv | grep GT' (without the single quotes). If you have 8GT/s (giga transfers per sec per channel) that means your nvme is running gen 3.20:12
marc_Ok i finally found out how to check the Nvme link speed sudo lspci -vv shows it, it says my nvme are running at 8GT/s which must be gen3 so its not that thats slowing them down20:13
WaVmerpnderp: did it take the rule or did it error?20:13
merpnderpWaV: it took the rule20:13
WaVmerpnderp: you may have to insert it before the deny rule20:14
marc_LnkSta:Speed 8GT/s (ok), Width x4 TrErr- Train- SlotClk+ DLActive- BWMgmt- ABWMgmt-20:14
marc_so its running at x4 not x820:14
WaVmerpnderp: for example, sudo ufw insert 1 allow out from any to 10.67.5.0/2420:15
WaVmerpnderp: then sudo ufw status numbered and delete the other rule that is at the bottom of the list.20:16
merpnderpDidn't work20:17
WaVhrm, are you able to reach that subnet with ufw disabled?20:17
merpnderpWaV: yep20:17
merpnderpTrying to mount a windows share.20:17
merpnderpWorks with ufw disabled20:18
WaVIf what I suggested doesn't work, you may want to wait for advice from someone else.20:18
WaVHoepfully you get the answer in here because I'm curious to what the solution is.20:19
marc_ok after checking out some stuff my drives specs are SSD Max Sequential Read CDM. Up to 3,480MB/s. and SSD Max Sequential Write CDM. Up to 2,000MB/s they are running at GEN3 x4 which is correct but the write speeds are only 338MB/s any idea would could be slowing it down20:27
marc_maybe driver i was thinking20:28
jhutchinsNVME in wrong slots/order.20:28
marc_do you think i could have them on the wrong CPU20:29
marc_half the slots are only available when the second cpu is installed20:29
marc_so i could have them on cpu1 lanes or cpu2 lanes20:30
marc_PCI-Express 20:31
marc_    1 PCI-E 3.0 x1620:31
marc_    6 PCI-E 3.0 x820:31
marc_    (Both CPUs need to be installed for full access to PCI-E slots and onboard controllers. See manual block diagram for details.)20:31
marc_opps20:31
marc_https://www.supermicro.com/products/motherboard/Xeon/C600/X9DRH-iF.cfm20:31
marc_the graphics card is on cpu2 and the two nvme drives are on cpu1 currently20:33
rfmmarc_, how are you measuring the write speed?   If it's doing many small block writes, you may be limited by the CPU context switching speed rather than the drive20:36
orange1he said he used gnome-disks and also dd i think20:39
marc_are you talking about the link speed between the two cpu's its 8GT/s at moment20:39
marc_I ws just about to move one nvme to the other cpu20:39
TheCodeTherapy@marc_, Have you tried measuring your nvme speed with kdiskmark? (sudo snap install kdiskmark)20:41
TheCodeTherapy@marc_, that will give you more reliable results (specially in case you select a file size that is at least 1/1000th of your nvme total size, otherwise the results can be affected by cache)20:42
marc_tried that i just moved an nvme dont know if i moved the right one yet20:42
marc_havent tried than sorry20:42
TheCodeTherapy@marc_, Here's some results of a 4th Gen 2TB nvme for comparison: https://i.imgur.com/pC5BinL.png20:43
jhutchinsmarc_: So what do you do with this system besides run benchmarks?20:44
marc_well i was using for playing DayZ20:46
marc_I was running E5-2687W v2 one failed so i have to get another one i also had a ASU mb fail and a gtx780 over this period with the graphics card shortage20:46
marc_i might have to restart nvme had disapeard off the device list20:48
JonathanDAnd people /act20:48
JonathanDoops, sorry.20:48
marc_brb20:48
marc_ok back just re check my drive benchmarks and now i moved them one has gotten worse its at about 188MB/s and the other one has moved up to 462.2MB/s20:53
marc_might be someone like the order they were placed like has been said here20:53
marc_something like sorry20:54
orange1make the drive ext4 instead of ntfs i read on the first page on google. did you do that20:56
orange1idk if you have it that way already20:57
marc_I just checked they are both Ext4 drives20:57
orange1ah20:57
marc_the Nvme drives are my Steam drives with games files on them i get pauses in games becasue of the write speeds i think20:58
orange1yeah your speeds are slow. write speed20:59
marc_gonna move them back now but in different order21:00
orange1there also was something about adding big_writes in fstab but i never did that stuff before so i dont know if its safe. maybe somebody else knows21:02
marc_yea im gonna reboot that might be something to look at brb21:03
oerheksi have no clue what devices he is testing, i hope he understands the difference between M2 SSD and M2 NVME SSD21:06
marc_ok got a tiny bit more speed out of them so how do i do this fstab big writes21:09
rfmI never heard of big_writes so I looked it up, it is a ntfs mount option, so not for your ext4 file systems21:11
marc_ok so that wont help21:11
=== diskin_ is now known as diskin
oerhekswhat m2 device is this exactly, marc_ ? there is a difference between M2 SATA SSD and M2 NVME SSD21:12
marc_MP510 corsair21:13
marc_one is 240GB the other is 480GB21:13
marc_i used to have issued with windows 10 with them there was an AMD drivers i would put that in and the drives would run at full speed21:14
marc_AMD raid21:14
marc_even though i werent using raid21:14
marc_could be a driver issue21:15
marc_nvme_core             126976  3 nvme21:16
marc_nvme                   45056  221:16
oerhekshmm oke, it should be a racemonster https://www.phoronix.com/scan.php?page=article&item=corsair-force-mp510&num=221:17
marc_the 480GB version is faster than the 240GB one21:19
oerhekspoint of concern, there is firmware updates available, but not through ubuntu/ Linux Vendor Firmware Service (LVFS)    https://help.ubuntu.com/community/BIOSUpdate21:19
marc_one says Force MP510 (ECFM22.7) the other says Force MP510 (ECFM13.3)21:20
marc_looks like im not the only one with a problem with these MP510 drives https://linustechtips.com/topic/1210088-corsair-mp510-slow-write-speeds/21:23
oerheks"Corsair SSD Toolbox Software: Enables advanced drive controls from your desktop including secure erase and firmware updates."21:25
oerhekswindows21:25
oerheksif that does not fix, i have no clue21:25
marc_yea i dont have windows21:25
marc_hope i did this right21:27
marc_sudo dd if=/dev/nvme0n1p1 of=tempfile bs=1M count=4096 conv=fdatasync,notrunc status=progress oflag=direct21:27
marc_4294967296 bytes (4.3 GB, 4.0 GiB) copied, 14.313 s, 300 MB/s21:27
marc_slow21:28
oerheksbs=4M could do better21:28
sarnoldthat's just one set of read and write syscalls outstanding at any given moment, too; most of these devices hit their peak throughputs when they've got a handful of operations in flight at once21:30
oerheksjups random read/writes are give a more reliable number too.21:30
marc_17179869184 bytes (17 GB, 16 GiB) copied, 51.0115 s, 337 MB/s21:30
marc_still slow21:30
oerheksbut oke, this gives an indication that it is below specs21:31
mybalzitch554696704 bytes (555 MB, 529 MiB) copied, 0.594914 s, 932 MB/s21:31
mybalzitchon a fancy gen 4 1tb nvme21:31
rfmis that dd reading from and writing to the same device (where is "tempfile") ?21:32
marc_well its got to be one of three things driver, bios, firmware thats casuing it21:32
mybalzitchalso like rfm said, write to tmpfs21:33
marc_Ok i think i have figure out whats wrong with the nvme drives21:48
marc_its to do with the bios bitrification on the pcie21:48
marc_i need to mess around with it i set some to gen2 and some to gen321:49
marc_i got one drive up to 823 MB/s and the other up to 711MB/s21:49
marc_read speeds have dropped though21:50
gneeriiloeepdeerhow can I use geany to search only for capitalized text?21:53
=== tscott_ is now known as tscott
oerheksgneeriiloeepdeer, not, https://www.geany.org/manual/current/#search-replace-and-go-to21:57
oerheksonly case sensitive is an option21:58
oerhekswhoe manual; https://www.geany.org/manual/current/21:58
sarnoldooh nice regex, [A-Z]+ might do it21:59
=== Festivus-Maximus is now known as Kilroy
Maikgneeriiloeepdeer: do you actually have a Ubuntu related support question? This isn't geany support after all.22:01
gneeriiloeepdeeris that a command or something I have to type in geany itself?22:02
Maikgneeriiloeepdeer: i guess you ahve to search for geany support elsewhere22:02
Maikahve/have22:02
Maikthis channel is focussed on support with Ubuntu22:03
sarnoldis geany not in ubuntu?22:03
Maikor, just like with your previous question, ask in #linux22:04
Maiksarnold: is this a geany support channel?22:04
sarnoldMaik: we help people with bash, we help people with libreoffice, we help people with nvidia drivers.. why should we not help people with geany?22:05
tebruno99The topic police have been in full force since the switch from Freenode. Starting to seem a little bit user hostile.22:05
sarnoldyeah22:06
sarnoldthere's space to say "you may get better support in a channel dedicated to your thing", sure22:06
sarnoldbut ubuntu ships geany, and as far as I can tell, gneeriiloeepdeer got an answer that will probably work22:06
Maiknot by defaul but it's in the repo's, that's true22:07
gneeriiloeepdeersecconding what sarnold and tebruno99 say, why is there a topic police now? this channel in freenode was always user friendly and welcoming22:08
gneeriiloeepdeeram I the only one who cannot access https://www.geany.org/manual/current/?22:09
jkcgneeriiloeepdeer: Up for me.22:09
Maikgetting support with ubuntu is one thing, getting support with an app is another imho but go on and continue22:09
Maikgneeriiloeepdeer: no problem accessing that site22:11
sarnoldgneeriiloeepdeer: how about the http version? http://www.geany.org/manual/current/#search-replace-and-go-to22:11
gneeriiloeepdeeroh, you menat than22:12
gneeriiloeepdeerthat22:12
gneeriiloeepdeergot it22:12
gneeriiloeepdeersarnold, how would I search for groups of 2 or more capitalized letters in geany?22:15
sarnoldgneeriiloeepdeer: try the regex search, [A-Z]{2,}22:17
sarnolddifferent tools mean different things by "regex" but this is common..22:17
gneeriiloeepdeersarnold, much appreciated22:18
shepherd_Hi, is there anyway to change the name of a touch interface? I have two touchscreens with the same name, and upon every other reboot the ID change around. I cannot map by ID because of this, but with the same name I cannot map by name either. Thank you guys for any suggestions! this is my xinput output https://pasteboard.co/XdG9kMumk87j.apng22:29
sarnoldoh man :/ that's frustrating..22:30
shepherd_It is indeed, that's why I was hoping, that maybe somewhere I could modify the name of one of the touch interfaces22:33
sarnoldshepherd_: if it were my problem to solve, I think I'd start with trying to write specific udev rules for the two devices and hope they have serialnumbers that are different, or something like that -- this page looks like a decent starting point for udev rules and the device properties that udev knows http://granjow.net/udev-rules.html22:34
sarnoldshepherd_: but I don't know for certain if anything udev sets will influence what libinput sees :/22:34
sarnoldshepherd_: this suggests something else that *might* work, Xorg InputClass sections, https://unix.stackexchange.com/questions/58117/determine-xinput-device-manufacturer-and-model#answer-220082  -- https://www.x.org/archive/current/doc/man/man5/xorg.conf.5.xhtml#heading9 the 'MatchTag' entry might be useful .. dunno :(22:36
shepherd_Thanks so much, I'll take a look at these links and see if I can make something of them hehe22:37
sarnoldgood luck :)22:39
sarnoldit seems like an obvious need once you see it raised..22:40
sarnoldI'm surprised it's not easier to find something about it22:40
sarnoldbut not all usb devices have serial numbers, so it might not be 'easy' to do..22:40
shepherd_I was quite surprised too, I'm not a super super linux user, but hopefully I'll be able to figure something out hehe, still reading through those links now22:41
marc_how do i check my current cpu speed22:43
marc_its ok i remember cat /proc/cpuinfo22:45
marc_so my next question is, is there a way to limit the cpu speed so it doesnt drop below a certain value22:46
sarnoldmarc_: that's gotten increasingly complicated over the years .. there's a bunch of documentation in https://www.kernel.org/doc/Documentation/cpu-freq/ that is perhaps worth a skim22:48
sarnoldthere's multiple userland tools that also want to own your processor speed; gamemode and thermald are the first two that come to mind, there's probably a dozen more22:49
cbreakwhy would you want to limit the cpu speed?22:50
ograand you can directly modify bits and pieces in /sys/devices/system/cpu/cpufreq/policy[0-9]*22:50
cbreakfrequency scaling is essential for power / thermal management22:50
marc_when im doing disk benchmarks sometimes i get a higher speed sometimes i get a low speed and i noticed my cpu speed is scalling down to 1.2ghz22:51
ogracbreak, well, if you use the system as game rig you probably want to limit the min_freq values22:51
ograand keep everything a bit higher up22:51
ograwasnt there some gaming app that did all that ?22:52
marc_frequency scalling is in play and its a pain22:52
ograhttps://www.omgubuntu.co.uk/2019/08/feral-gamemode-ubuntu22:52
ograthere is is22:52
ogra*it is22:52
marc_isnt gamemode already installed in ubuntu i saw a file for it yesterday an .ini file22:53
ogranot that i know of22:54
ograi definitely dont have it on any of my systems22:54
marc_its in etc22:55
marc_gammode.ini22:55
marc_gamemode.ini22:55
marc_but what i am after is a way to linit the minumum cpu frequency in ubuntu22:57
marc_limit22:57
marc_cat /proc/cpuinfo currently says im running at 1200 mhz22:58
ograwell, see above ... we gave you some suggestions22:58
marc_another thing i noticed in power its got two settings and no other power plans ???22:59
marc_balanced and power save where is performance23:00
ograogra@acheron:~$ cat /sys/devices/system/cpu/cpufreq/policy0/scaling_available_governors23:01
ograperformance powersave23:01
ograi dont have "balanced" ...23:02
marc_well i dont have performance23:03
ograask your CPU vendor ? 🙂23:03
marc_i went into the bios and told it to be performance but ubuntu wants to be a slug :D23:04
marc_This is getting frustrating I'm seriously thinking of ditching ubuntu for something like Gentoo because it wont give me performance settings for the power management.23:11
sarnoldmaybe pastebin what commands you ran and the output you got?23:12
marc_nothing cos i dont have performance setting in the power menu23:12
marc_and im on a desktop PC23:12
marc_maybe there is a way to disable frequency scalling23:14
sarnoldmarc_: that's the 'performance' governor23:15
sarnoldwhat settings are available depends upon your cpus23:15
sarnoldhere's a recent-ish laptop https://termbin.com/dh1w23:15
marc_E5-264323:15
marc_intel23:15
marc_xeon23:15
sarnoldhere's an e5-...v3 xeon system https://termbin.com/2drg23:16
marc_sudo apt-get install cpufrequtils23:17
marc_nano /etc/init.d/cpufrequtils23:26
marc_that will sort it Govenor = disabled23:26
rfmmarc_, I think you need to disable the systemd ondemand service too so it doesn't fight wth cpufrequtils, at least according to this https://askubuntu.com/questions/1322492/how-to-set-as-default-performance-mode-on-ubuntu-20-04-instead-of-powersave23:29
marc_i was just thinking about that instead i thought i would triy min = 3000 max = 350023:30
marc_so in  moment i will reboot see if it works so its always on startup23:31
marc_brb23:31
marc_well that didnt work im still at 1200mhz23:36
marc_looks like cpufrequtils doesnt work on twin processors23:44
sarnoldhonestly I think you'd be better suited to *remove* all the power management packages and just rely on the kernel interfaces23:45
marc_i was thinking of going into bios and setting the multipler fixed and disable power mangment from the bios23:46

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!