The Linux Questions Thread: a bunch of pitfalls, but technically it's possible

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > The Linux Questions Thread: a bunch of pitfalls, but technically it's possible

«‹›1029 »

Rocko Bonaparte: Mar 12, 2002; Every day is Friday!

Pardon my vomit but maybe you see something here that looks wrong. I'm about to reboot after doing all this and will be looking back at it if I still get screwed. I guess I could pastebin but I may edit in place. Let me know if this is too much crap in one post.

code:

# I am sudo -s permaroot in the terminal like a criminal
# My main drive is encrypted
cryptsetup luksOpen /dev/sda3 kubuntu      # Password here

vgscan      # The drive comes up as vgkubuntu

root@kubuntu:/home/kubuntu# vgchange -ay vgkubuntu
#  2 logical volume(s) in volume group "vgkubuntu" now active

mount /dev/vgkubuntu/root /mnt/
mount --bind /dev /mnt/dev 
mount --bind /dev/pts /mnt/dev/pts
mount --bind /proc /mnt/proc
mount --rbind /sys /mnt/sys
mount /dev/sda1 /mnt/boot/efi       # It should be under /mnt/boot/efi, right?
cp /etc/resolv.conf /mnt/etc

chroot /mnt


# My /etc/default/grub has this relevant stuff:
GRUB_DEFAULT=0
GRUB_TIMEOUT_STYLE=hidden
GRUB_TIMEOUT=0
GRUB_DISTRIBUTOR=`lsb_release -i -s 2> /dev/null || echo Debian`
GRUB_DISABLE_OS_PROBER=false
GRUB_CMDLINE_LINUX_DEFAULT="quiet splash"
GRUB_CMDLINE_LINUX=""

# Settings for yelling at nouveau included here but commented out.
#GRUB_CMDLINE_LINUX_DEFAULT="nouveau.blacklist=1 quiet splash"
#GRUB_CMDLINE_LINUX="nouveau.blacklist=1"


# Running for keeps now
# bootloader-id argument here was originally debian which would create a debian folder. This machine
# actually previous had debian on it so I didn't catch this. I also am only doing this now to be very explicit.
grub-install --target=x86_64-efi --efi-directory=/boot/efi --bootloader-id=ubuntu --recheck --debug /dev/sda &> /tmp/grubinstall.txt

In /tmp/grubinstall, it claims there were no errors but it also complains about a missing device.map. However, that looks to be
deprecated. I tried to make one but when I rerun grubinstall, it purges it (and then complains it's missing).

update-grub

Sourcing file `/etc/default/grub'
Sourcing file `/etc/default/grub.d/init-select.cfg'
Generating grub configuration file ...
Found linux image: /boot/vmlinuz-5.15.0-25-generic
Found initrd image: /boot/initrd.img-5.15.0-25-generic
Warning: os-prober will be executed to detect other bootable partitions.
Its output will be used to detect bootable binaries on them and create new boot entries.
Adding boot menu entry for UEFI Firmware Settings ...
done

# Looking at /boot/grub/grub.cfg, I see a lot of entries talking about bab358ba-ce94-4f4c-9b68-52816b47723d:
# search --no-floppy --fs-uuid --set=root --hint='lvmid/ycfLbU-kxqA-uct6-niP6-5psS-o5x5-ZjrTZj/vIVqDF-NRJc-Lxj3-Yx2v-dsfv-MESh-SJdoRy'  bab358ba-ce94-4f4c-9b68-52816b47723d
# And if I run blkid, that id matches vgkubuntu-root (/dev/sda3, the drive I'm chroot'd on that I want to use)


Grub command line to try when this inevitably fails again!:
set prefix=(hd1,1)/boot/grub
set root=(hd1,1)
insmod linux
insmod normal
normal
boot

Might try to explicitly boot a kernel
linux /vmlinuz-5.15.0-25-generic 
initrd /initrd.img-5.15.0-25-generic
(use tab autocomplete)

Edit: aaaaaand nothing. Still stuck on the console. Those commands to try to manually start the menu did not work either. What they hell am I screwing up?

Rocko Bonaparte fucked around with this message at 23:49 on May 31, 2022

# ? May 31, 2022 21:56

Adbot: ADBOT LOVES YOU

# ? May 18, 2024 01:03

Rocko Bonaparte: Mar 12, 2002; Every day is Friday!

Conclusion to the mess (well, so far): I got it working with the boot-repair utility in Ubuntu but only after one retru. Recall I was in a LiveUSB with an encrypted disk that had my stuff on it. The first time I tried it, I had already decrypted the disk. Boot Repair complained that I should decrypt the partition /dev/sda3 and as an example named it "myvolume." Whatever, I just rolled with it. The chroot commands it was telling me to enter wouldn't work as-is but I interpreted them and got through it. Except it didn't boot. I deserved it.

Given all those variables, I just restarted into the LiveUSB, didn't decrypt the disk, and literally named it "myvolume" when Boot Repair complained about it. This time all recommended the commands worked fine. What came out the other end was a bootable grub menu.

So, what the gently caress? I had speculated at the end that something went wrong with decrypting the main drive during boot. The uuid that grub was using everywhere was for the decrypted volume, not the encrypted one. This included the search uuid in /boot/efi/EFI/ubuntu/grub.cfg. The search.fs_uuid command that lead off that configuration was using my decrypted volume's uuid. When I tried to manually load this .cfg file using the configfile command in grub, it complained the volume wasn't found. Well, it didn't complain so much as very quickly flash a message about it before flinging me back into the grub command line, but I deciphered it after trying the command a few times.

The Boot Repair utility managed to figure out what to do, which had ... nothing to do with the encrypted or decrypted uuids at all! The EFI grub.cfg needed to use the uuid for my /boot partition (/dev/sda2). So, how the gently caress did the efi grub.cfg get changed in the first place? I don't know. I wasn't wading around in there. But now I know what to sniff for if it happens again (oh God I hope not).

While getting everything back up, I managed to find that nouveau.blacklist=1 and modprobe.blacklist=nouveau were not enough kernel parameters to block the nouveau driver. I ultimately also needed to add nouveau.modeset=0. I don't know if that's the only actual flag I needed, and I don't really feel like finding out. I can say though that normal module blacklisting just wasn't enough. I'm amazed with how tightly coupled the nouveau driver is in this kernel! My day job has veered into building internal kernel stuff with lots of driver poo poo and none of it is stapled so tight to the kernel itself that you have to throw a slurry of kernel flags at the boot options to disable one of them. And this is with me working with a bunch of electrical engineers who did one semester of C and are stuck in that poo poo show.

Thank you for listening to my TED Talk.

# ? Jun 1, 2022 05:47

The Atomic Man-Boy: Jul 23, 2007

Con someone recommend a webcam that works out-of-the-box on Ubuntu? After fiddling to no avail with some cheap amazon one, I decided to return it and just ask the thread.

# ? Jun 1, 2022 22:56

BlankSystemDaemon: Mar 13, 2009

The Atomic Man-Boy posted:

Con someone recommend a webcam that works out-of-the-box on Ubuntu? After fiddling to no avail with some cheap amazon one, I decided to return it and just ask the thread.

I can't believe you managed to get a webcam that doesn't send a YUV stream that can be accessed via /dev/video0.

That's how basically every single webcam works and is known as the Video4Linux2 API that's being used - that should give you enough to go on for finding out more, although it's very possible that you need to install something to handle the API.
On FreeBSD, webcams work through multimedia/webcamd - although ironically that still supports V4L1 features and codecs that was dropped from Linux.

I have the cheapest 1080p Sandberg webcam I could find, and it works great for the few times I've had occasion to use it for WebRTC video conferencing in Firefox - with the caveat that I usually have to do some user-agent fuckery, but that's not unheard of on Linux either, from what I've been told.

# ? Jun 1, 2022 23:05

Rocko Bonaparte: Mar 12, 2002; Every day is Friday!

I think I have a Logitech C922 and it has given me absolutely no problems in Linux. That is high praise! I got 99 problems but my webcam ain't one of them.

Edit: Confirmed from receipt from Logitech's own store: C922 PRO HD STREAM WEBCAM - Black (960-001087). It's a super trooper.

Worked in Kubuntu 20.04 and 22.04, as well as this mutt intermediate Kubuntu I had for a week when it half-assed upgraded itself without asking me and it was a complete mess.

Rocko Bonaparte fucked around with this message at 00:46 on Jun 2, 2022

# ? Jun 2, 2022 00:42

Chilled Milk: Jun 22, 2003; No one here is alone,
satellites in every home

Yeah, I'm impressed you found one that doesn't just work. Then again, there's no bottom to the junk amazon will sell you.

# ? Jun 4, 2022 20:47

Rocko Bonaparte: Mar 12, 2002; Every day is Friday!

Monday update on my Linux rig: It actually booted fine from the get-go this time. I think I finally reined it in with last week's changes.

# ? Jun 6, 2022 13:26

Super-NintendoUser: Jan 16, 2004; COWABUNGERDER COMPADRES; Soiled Meat

Any recommend a USB bootable distro that I can use to experiment with kubernetes? I have a server running an media farm with docker-compose, it's an ancient (but upgraded to pretty current) Ubuntu install. I'd like to be able to boot off a stick, fiddle with k8s, and then switch back to the internal hd and go back to my docker setup until I have k8s working well.

All the media/docker files are on a separate drive from the bootable ubuntu, so I can just remove that drive when I'm done.

# ? Jun 6, 2022 14:35

RFC2324: Jun 7, 2012; http 418

Why not just get k8s working in a lab environment first? Install vbox on your desktop, install a decent OS, and have fun.

# ? Jun 6, 2022 15:22

minato: Jun 7, 2004; cutty cain't hang, say 7-up.; Taco Defender

Yeah, something like MiniKube is tailor made for messing around with k8s, you can use Docker or a VM. https://minikube.sigs.k8s.io/docs/start/

# ? Jun 6, 2022 15:36

Volguus: Mar 3, 2009

I have a question about iperf:

On Fedora I was looking at my local network yesterday and wanted to run an iperf test. I was getting around 740Mbps with the gateway. Worried, I changed cables, bypassed switches, everything I could think of, still 740Mbps. On another box in my room, on the same switch, that box running Debian, I was getting 940Mbps. Burned a live debian image on a flash drive, and from Debian on my machine I got 940Mbps as well.

The normal google speed test, which basically tests my connection to the internet, I was getting 960Mbps.
From what I can gather is not my network card. Not the driver. Not the cables or switches, not anything except iperf itself.

Anyone has any idea what could be going on?

I tried a test with nc itself, still, to the gateway, around 740Mbps. I'm a bit stumped.

# ? Jun 6, 2022 16:31

Super-NintendoUser: Jan 16, 2004; COWABUNGERDER COMPADRES; Soiled Meat

RFC2324 posted:

Why not just get k8s working in a lab environment first? Install vbox on your desktop, install a decent OS, and have fun.

So this is a really valid suggestion, one of the main reasons is that I want to setup a webserver with Traefik/Lets Encrypt with a real domain name from a real public IP through cloudflare. For me to do that I need a valid domain/DNS, a public IP opened on 443. On a Virtualbox lab that'd be a bigger hassle. Probably what I'll do is just spin up a cheap EC2 instance with an elastic IP and point a subdomain that I already own at it.

I have a second dockerfile question. I haven't worked in windows in years, but $NEW_JOB has given me a windows laptop. I have Docker desktop setup, but I can't for the life of me get a dockerfile to work mounting a local windows directly. I using docker in I guess in WSL2, it works from the cli:

code:

docker run -it -v "/mnt/c/Users/jmjf/Development:/mnt/jmjf" --entrypoint /bin/bash test

but in a docker file I can't nail down the syntax:

code:

FROM fedora:latest

RUN echo "date" >> /var/tmp/output
RUN echo "date" >> /var/tmp/output
VOLUME "/c/Users/jmjf/Development":"/mnt/jmjf1"
VOLUME "/mnt/c/Users/jmjf/Development":"/mnt/jmjf2"
VOLUME /c/Users/jmjf/Development:/mnt/jmjf3
VOLUME /mnt/c/Users/jmjf/Development:/mnt/jmjf4
VOLUME "/c/Users/jmjf/Development":/mnt/jmjf5
VOLUME "/mnt/c/Users/jmjf/Development":/mnt/jmjf6
VOLUME /c/Users/jmjf/Development:"/mnt/jmjf7"
VOLUME /mnt/c/Users/jmjf/Development:"/mnt/jmjf8"
VOLUME c:/Users/jmjf/Development:/mnt/jmjf1
VOLUME c:\Users\jmjf\Development:/mnt/jmjf2
VOLUME "c:/Users/jmjf/Development:/mnt/jmjf3"
VOLUME "c:\Users\jmjf\Development:/mnt/jmjf4"
VOLUME "c:/Users/jmjf/Development":/mnt/jmjf5
VOLUME "c:\Users\jmjf\Development":/mnt/jmjf6
VOLUME c:/Users/jmjf/Development:"/mnt/jmjf7"
VOLUME c:\Users\jmjf\Development:"/mnt/jmjf8"

None of these work, I get this weird looking nonsense:

code:

/dev/sdc on /c:/Users/jmjf/Development:/mnt/jmjf

# ? Jun 6, 2022 18:59

RFC2324: Jun 7, 2012; http 418

Jerk McJerkface posted:

So this is a really valid suggestion, one of the main reasons is that I want to setup a webserver with Traefik/Lets Encrypt with a real domain name from a real public IP through cloudflare. For me to do that I need a valid domain/DNS, a public IP opened on 443. On a Virtualbox lab that'd be a bigger hassle. Probably what I'll do is just spin up a cheap EC2 instance with an elastic IP and point a subdomain that I already own at it.

I have a second dockerfile question. I haven't worked in windows in years, but $NEW_JOB has given me a windows laptop. I have Docker desktop setup, but I can't for the life of me get a dockerfile to work mounting a local windows directly. I using docker in I guess in WSL2, it works from the cli:
code:
docker run -it -v "/mnt/c/Users/jmjf/Development:/mnt/jmjf" --entrypoint /bin/bash test
but in a docker file I can't nail down the syntax:
code:
FROM fedora:latest

RUN echo "date" >> /var/tmp/output
RUN echo "date" >> /var/tmp/output
VOLUME "/c/Users/jmjf/Development":"/mnt/jmjf1"
VOLUME "/mnt/c/Users/jmjf/Development":"/mnt/jmjf2"
VOLUME /c/Users/jmjf/Development:/mnt/jmjf3
VOLUME /mnt/c/Users/jmjf/Development:/mnt/jmjf4
VOLUME "/c/Users/jmjf/Development":/mnt/jmjf5
VOLUME "/mnt/c/Users/jmjf/Development":/mnt/jmjf6
VOLUME /c/Users/jmjf/Development:"/mnt/jmjf7"
VOLUME /mnt/c/Users/jmjf/Development:"/mnt/jmjf8"
VOLUME c:/Users/jmjf/Development:/mnt/jmjf1
VOLUME c:\Users\jmjf\Development:/mnt/jmjf2
VOLUME "c:/Users/jmjf/Development:/mnt/jmjf3"
VOLUME "c:\Users\jmjf\Development:/mnt/jmjf4"
VOLUME "c:/Users/jmjf/Development":/mnt/jmjf5
VOLUME "c:\Users\jmjf\Development":/mnt/jmjf6
VOLUME c:/Users/jmjf/Development:"/mnt/jmjf7"
VOLUME c:\Users\jmjf\Development:"/mnt/jmjf8"
None of these work, I get this weird looking nonsense:
code:
/dev/sdc on /c:/Users/jmjf/Development:/mnt/jmjf

is c:\Users\jmjf\Development the path to that directory from inside WSL? Remember, WSL2 is running inside a VM, and even WSL1 required translating paths because C: doesn't exist in a linux context

# ? Jun 6, 2022 21:00

Super-NintendoUser: Jan 16, 2004; COWABUNGERDER COMPADRES; Soiled Meat

RFC2324 posted:

is c:\Users\jmjf\Development the path to that directory from inside WSL? Remember, WSL2 is running inside a VM, and even WSL1 required translating paths because C: doesn't exist in a linux context

I'm running this command from inside WSL and it works fine:

code:

docker run -it -v "/mnt/c/Users/jmjf/Development:/mnt/jmjf" --entrypoint /bin/bash test

I did some more digging and turns out you can't define bind mounts in a Dockerfile, since the file really only defines how a container is built, not how it's used.

# ? Jun 6, 2022 22:09

RFC2324: Jun 7, 2012; http 418

Jerk McJerkface posted:

I'm running this command from inside WSL and it works fine:
code:
docker run -it -v "/mnt/c/Users/jmjf/Development:/mnt/jmjf" --entrypoint /bin/bash test
I did some more digging and turns out you can't define bind mounts in a Dockerfile, since the file really only defines how a container is built, not how it's used.

That makes sense. I was thinking about the compose file, that does define how its used.

and /mnt/c/Users/jmjf/Development isn't the poo poo that you posted as coming up in the final result, which is why I didn't look at it. /mnt/c is the only path you should have needed/used (somewhere other than the docker file, apparently

)

# ? Jun 7, 2022 01:02

BlankSystemDaemon: Mar 13, 2009

Can someone verify that the NFSv4.2 client works on Linux kernel v5.17.4 or newer?

A system I'm trying to help someone with is refusing to accept minorversion=2 as an argument, despite it being supported and nominally enabled.

# ? Jun 9, 2022 16:42

Cool Dad: Jun 15, 2007; It is always Friday night, motherfuckers

I haven't used linux in years, but I want to set up a computer to use as a sort of multipurpose server. I'm primarily going to use it to run Foundry Virtual Tabletop but I'd also like to use it as a print server and maybe a file server. I figure it doesn't really matter, but is there a distro that would be particularly well suited to this kind of thing?

# ? Jun 9, 2022 17:38

RFC2324: Jun 7, 2012; http 418

Cool Dad posted:

I haven't used linux in years, but I want to set up a computer to use as a sort of multipurpose server. I'm primarily going to use it to run Foundry Virtual Tabletop but I'd also like to use it as a print server and maybe a file server. I figure it doesn't really matter, but is there a distro that would be particularly well suited to this kind of thing?

Just pick what you like. I suggest opensuse because yast2 makes admin easy, but it can be a little bit of a pain to and and configure some software for it.

I just want a handy, self contained, easy to manage ldap server in a tiny container, is that so much to ask?

# ? Jun 9, 2022 18:28

Twerk from Home: Jan 17, 2009; This avatar brought to you by the 'save our dead gay forums' foundation.

Cool Dad posted:

I haven't used linux in years, but I want to set up a computer to use as a sort of multipurpose server. I'm primarily going to use it to run Foundry Virtual Tabletop but I'd also like to use it as a print server and maybe a file server. I figure it doesn't really matter, but is there a distro that would be particularly well suited to this kind of thing?

You'd be well served to stick with Ubuntu LTS releases, in my opinion. The least fussy user experience, and consistently easy upgrades. I'd install Ubuntu 22.04, which is supported and gets updates until 2027.

I'd use the server image if you want to keep resource usage really low and don't plan to connect a monitor to it directly. If you want to use this as a desktop and server, just install a desktop image and you'll get a nice GUI out of the box too. Set up automatic upgrades with automatic reboots, which you can do graphically on desktop edition, or like this for server with the unattended-upgrades package: https://help.ubuntu.com/community/AutomaticSecurityUpdates

# ? Jun 9, 2022 20:48

BlankSystemDaemon: Mar 13, 2009

Cool Dad posted:

I haven't used linux in years, but I want to set up a computer to use as a sort of multipurpose server. I'm primarily going to use it to run Foundry Virtual Tabletop but I'd also like to use it as a print server and maybe a file server. I figure it doesn't really matter, but is there a distro that would be particularly well suited to this kind of thing?

I'd be careful of anything that claims to be an appliance; making them do anything else can be an absolute pain in the rear end.

# ? Jun 10, 2022 00:24

LRADIKAL: Jun 10, 2001; Fun Shoe

Had trouble typing the right magic words into Google. I have a Ubuntu desktop install I use as a Plex server with an Nvidia graphics card for transcoding. Can my wife log into the machine with another user account and play games on the TV simultaneously, directly on the system? Will the graphics card be good to go on both sessions?

# ? Jun 11, 2022 01:07

Hed: Mar 31, 2004; Fun Shoe

I just swapped to a bigger hard drive and am very rusty on LVM2. Do I want to lvextend /dev/pve/root? I just want to give my VMs more room since they were thin provisioned and if I run a big update on one I've had it fill up the disk and cause nonsense.

I moved from 512GB to 2TB. I've done fdisk to resize nvme0n1p3, and done pvresize /dev/nvme0n1p3, but now I don't know if I lvextend /dev/pve/root or what.

code:

NAME                         MAJ:MIN RM   SIZE RO TYPE MOUNTPOINT
nvme0n1                      259:0    0   1.8T  0 disk 
&#9500;&#9472;nvme0n1p1                  259:1    0     1M  0 part 
&#9500;&#9472;nvme0n1p2                  259:2    0   256M  0 part /boot/efi
&#9492;&#9472;nvme0n1p3                  259:3    0   1.8T  0 part 
  &#9500;&#9472;pve-swap                 253:0    0     8G  0 lvm  [SWAP]
  &#9500;&#9472;pve-root                 253:1    0    96G  0 lvm  /
  &#9500;&#9472;pve-data_tmeta           253:2    0   3.6G  0 lvm  
  &#9474; &#9492;&#9472;pve-data-tpool         253:4    0 349.6G  0 lvm  
  &#9474;   &#9500;&#9472;pve-data             253:5    0 349.6G  0 lvm  
  &#9474;   &#9500;&#9472;pve-vm--100--disk--1 253:6    0    10G  0 lvm  
  &#9474;   &#9500;&#9472;pve-vm--100--disk--2 253:7    0     4M  0 lvm  
  &#9474;   &#9500;&#9472;pve-vm--101--disk--1 253:8    0   150G  0 lvm  
  &#9474;   &#9500;&#9472;pve-vm--101--disk--2 253:9    0     4M  0 lvm  
  &#9474;   &#9500;&#9472;pve-vm--103--disk--1 253:10   0    80G  0 lvm  
  &#9474;   &#9500;&#9472;pve-vm--111--disk--0 253:11   0    40G  0 lvm  
  &#9474;   &#9500;&#9472;pve-vm--111--disk--1 253:12   0     4M  0 lvm  
  &#9474;   &#9500;&#9472;pve-vm--106--disk--0 253:13   0    32G  0 lvm  
  &#9474;   &#9500;&#9472;pve-vm--106--disk--1 253:14   0     4M  0 lvm  
  &#9474;   &#9500;&#9472;pve-vm--104--disk--0 253:15   0    10G  0 lvm  
  &#9474;   &#9500;&#9472;pve-vm--104--disk--1 253:16   0     4M  0 lvm  
  &#9474;   &#9500;&#9472;pve-vm--105--disk--0 253:17   0     4M  0 lvm  
  &#9474;   &#9492;&#9472;pve-vm--105--disk--1 253:18   0    32G  0 lvm  
  &#9492;&#9472;pve-data_tdata           253:3    0 349.6G  0 lvm  
    &#9492;&#9472;pve-data-tpool         253:4    0 349.6G  0 lvm  
      &#9500;&#9472;pve-data             253:5    0 349.6G  0 lvm  
      &#9500;&#9472;pve-vm--100--disk--1 253:6    0    10G  0 lvm  
      &#9500;&#9472;pve-vm--100--disk--2 253:7    0     4M  0 lvm  
      &#9500;&#9472;pve-vm--101--disk--1 253:8    0   150G  0 lvm  
      &#9500;&#9472;pve-vm--101--disk--2 253:9    0     4M  0 lvm  
      &#9500;&#9472;pve-vm--103--disk--1 253:10   0    80G  0 lvm  
      &#9500;&#9472;pve-vm--111--disk--0 253:11   0    40G  0 lvm  
      &#9500;&#9472;pve-vm--111--disk--1 253:12   0     4M  0 lvm  
      &#9500;&#9472;pve-vm--106--disk--0 253:13   0    32G  0 lvm  
      &#9500;&#9472;pve-vm--106--disk--1 253:14   0     4M  0 lvm  
      &#9500;&#9472;pve-vm--104--disk--0 253:15   0    10G  0 lvm  
      &#9500;&#9472;pve-vm--104--disk--1 253:16   0     4M  0 lvm  
      &#9500;&#9472;pve-vm--105--disk--0 253:17   0     4M  0 lvm  
      &#9492;&#9472;pve-vm--105--disk--1 253:18   0    32G  0 lvm

code:

root@proxmox:~# vgdisplay
  --- Volume group ---
  VG Name               pve
  System ID             
  Format                lvm2
  Metadata Areas        1
  Metadata Sequence No  221
  VG Access             read/write
  VG Status             resizable
  MAX LV                0
  Cur LV                18
  Open LV               8
  Max PV                0
  Cur PV                1
  Act PV                1
  VG Size               <1.82 TiB
  PE Size               4.00 MiB
  Total PE              476867
  Alloc PE / Size       117937 / 460.69 GiB
  Free  PE / Size       358930 / <1.37 TiB

code:

root@proxmox:~# lvdisplay | grep "LV Path"
  LV Path                /dev/pve/swap
  LV Path                /dev/pve/root
  LV Path                /dev/pve/vm-100-disk-1
  LV Path                /dev/pve/vm-100-disk-2
  LV Path                /dev/pve/vm-101-disk-1
  LV Path                /dev/pve/vm-101-disk-2
  LV Path                /dev/pve/vm-103-disk-1
  LV Path                /dev/pve/snap_vm-103-disk-1_plex_working
  LV Path                /dev/pve/snap_vm-103-disk-1_working
  LV Path                /dev/pve/vm-111-disk-0
  LV Path                /dev/pve/vm-111-disk-1
  LV Path                /dev/pve/vm-106-disk-0
  LV Path                /dev/pve/vm-106-disk-1
  LV Path                /dev/pve/vm-104-disk-0
  LV Path                /dev/pve/vm-104-disk-1
  LV Path                /dev/pve/vm-105-disk-0
  LV Path                /dev/pve/vm-105-disk-1

# ? Jun 11, 2022 15:34

Hed: Mar 31, 2004; Fun Shoe

I got it, I didn't realize data was thin provisioned (lvm-thin), it was root that was not, but it didn't really need more space.

So my steps were

fdisk to replace the last partition, pvresize to extend physical volume, then double the size of the metadata (lvresize --poolmetadatasize -r -L +3.6G pve/data in my case) and finally double the data (lvresize -L +350G pve/data).

I've still got a bunch of slack there, but that's good for now.

# ? Jun 12, 2022 14:01

Bob Morales: Aug 18, 2006; ~~Just wear the fucking mask, Bob~~

I don't care how many people I probably infected with COVID-19 while refusing to wear a mask, my comfort is far more important than the health and safety of everyone around me!

Anyone using pcp (performance co-pilot) to log performance metrics on their servers?

If so, what are you using to compile/view the data? We have Xdmod but I'm looking for the most straightforward way to show everything by host. Basically cpu/mem/disk/net usage. Could be as simple as cacti but I'd probably want to do things like heatmaps and stuff as well. These are cluster nodes so being able to group things by chassis/row would be ideal

# ? Jun 13, 2022 15:48

Klyith: Aug 3, 2007; GBS Pledge Week

LRADIKAL posted:

Had trouble typing the right magic words into Google. I have a Ubuntu desktop install I use as a Plex server with an Nvidia graphics card for transcoding. Can my wife log into the machine with another user account and play games on the TV simultaneously, directly on the system? Will the graphics card be good to go on both sessions?

Should work fine in theory -- the video encoder and the 3d acceleration are mostly separate bits, and multiple users can use the same GPU.

Question is how the plex server is running on the other user: as a systemd service, or as a desktop app? Because the 2nd one is likely to be trouble. Running multiple local x sessions is fiddly. If it's running as an at-boot service, and so nobody even needs to be logged on for it to run, that would be the nest set-up.

# ? Jun 13, 2022 21:03

LRADIKAL: Jun 10, 2001; Fun Shoe

It's running as a systemd service, and I am usually logged in remotely to manage and find torrents. Thanks for the tip.

# ? Jun 13, 2022 21:12

Vavrek: Mar 2, 2013; I like your style hombre, but this is no laughing matter. Assault on a police officer. Theft of police property. Illegal possession of a firearm. FIVE counts of attempted murder. That comes to... 29 dollars and 40 cents. Cash, cheque, or credit card?

I've lost the menu bar in Kate. Tips I've seen online are "Press CTRL-M" and "Check katerc to set showmenu=true".

Ctrl-M does nothing, and everything that mentions a Menu in katerc is already set to true.

I'm baffled and annoyed. Any ideas?

Edit: Okay, it's also gone in LibreOffice, so I'm gathering I just turned some global "Have Menu Bar" setting off. Where do I find that in KDE?

Vavrek fucked around with this message at 19:35 on Jun 16, 2022

# ? Jun 16, 2022 19:26

CaptainSarcastic: Jul 6, 2013

Vavrek posted:

I've lost the menu bar in Kate. Tips I've seen online are "Press CTRL-M" and "Check katerc to set showmenu=true".

Ctrl-M does nothing, and everything that mentions a Menu in katerc is already set to true.

I'm baffled and annoyed. Any ideas?

Edit: Okay, it's also gone in LibreOffice, so I'm gathering I just turned some global "Have Menu Bar" setting off. Where do I find that in KDE?

Maybe in System Settings under Application Styles?

# ? Jun 16, 2022 21:40

Klyith: Aug 3, 2007; GBS Pledge Week

Vavrek posted:

Edit: Okay, it's also gone in LibreOffice, so I'm gathering I just turned some global "Have Menu Bar" setting off. Where do I find that in KDE?

Some google suggests

quote:

You seem to have the Global Menu feature enabled, have a look at the System Settings:
In System Settings -> Workspace -> Startup and Shutdown -> Background Services, uncheck the "Application menus daemon" then log out and back into Plasma again

OTOH I have that deamon checked and have no problems?

(I guess this daemon can extract the main menu to either put it in a ≡ hamburger in the titlebar, or into a global menu bar ala macos. Cool feature. I wouldn't want to do that everywhere, but I wonder if you can do that for just some apps via special rules?)

# ? Jun 16, 2022 22:49

Vavrek: Mar 2, 2013; I like your style hombre, but this is no laughing matter. Assault on a police officer. Theft of police property. Illegal possession of a firearm. FIVE counts of attempted murder. That comes to... 29 dollars and 40 cents. Cash, cheque, or credit card?

CaptainSarcastic posted:

Maybe in System Settings under Application Styles?

Didn't find anything there about the menu bar.

Klyith posted:

Some google suggests

OTOH I have that deamon checked and have no problems?

(I guess this daemon can extract the main menu to either put it in a ≡ hamburger in the titlebar, or into a global menu bar ala macos. Cool feature. I wouldn't want to do that everywhere, but I wonder if you can do that for just some apps via special rules?)

Found that same suggestion, unticked the box, ticked the box ...

I don't know. The menu bar is back now. Problem is ... solved? Problem is gone. :shrug:

Thanks for the help.

# ? Jun 16, 2022 22:57

Lifroc: May 8, 2020

Is there a thread for Linux gaming in general? Proton workarounds, appeasing anticheat software, Lutris, Bottles, etc.?

If not, I could create one but I don't know which forum is more suitable for it, since it's both about gaming and writing incantations in the shell.

# ? Jun 20, 2022 17:06

Twerk from Home: Jan 17, 2009; This avatar brought to you by the 'save our dead gay forums' foundation.

1player posted:

Is there a thread for Linux gaming in general? Proton workarounds, appeasing anticheat software, Lutris, Bottles, etc.?

If not, I could create one but I don't know which forum is more suitable for it, since it's both about gaming and writing incantations in the shell.

I think that the de-facto Linux Gaming thread is the Steam Deck thread: https://forums.somethingawful.com/showthread.php?threadid=3973713, but I'd appreciate a general purpose one in Games also. There's probably going to be a lot of overlap.

# ? Jun 20, 2022 17:11

Lifroc: May 8, 2020

Twerk from Home posted:

I think that the de-facto Linux Gaming thread is the Steam Deck thread: https://forums.somethingawful.com/showthread.php?threadid=3973713, but I'd appreciate a general purpose one in Games also. There's probably going to be a lot of overlap.

Yeah, but there's also going to be a lot of posts about the Steam Deck itself or its hardware, which I don't really care about. I'll see about making a post.

# ? Jun 20, 2022 17:26

Lifroc: May 8, 2020

I've created a Linux Gaming thread: https://forums.somethingawful.com/showthread.php?threadid=4004994

Hope this is alright, I'm quite the newbie around these parts.

# ? Jun 20, 2022 18:17

Kivi: Aug 1, 2006; I care

I have an issue where Linux kills my process when it tries to use over 60% of memory. How do I stop that? I have 128 GB of RAM and the process in question is pg_upgrade sub-process pg_dump.

# ? Jun 21, 2022 11:50

VictualSquid: Feb 29, 2012; Gently enveloping the target with indiscriminate love.

Kivi posted:

I have an issue where Linux kills my process when it tries to use over 60% of memory. How do I stop that? I have 128 GB of RAM and the process in question is pg_upgrade sub-process pg_dump.

It sounds like it triggers the oom killer, somehow. It should say so in the syslog if that is the case. You can either change the general settings or protect the process with a specific setting. Don't remember how to do so exactly, just google oom killer.

# ? Jun 21, 2022 12:01

Lifroc: May 8, 2020

Kivi posted:

I have an issue where Linux kills my process when it tries to use over 60% of memory. How do I stop that? I have 128 GB of RAM and the process in question is pg_upgrade sub-process pg_dump.

I have read the latest Ubuntu has the OOM daemon a little too conservative, killing stuff even when there's enough memory available. Might that be the case?

They seem to use systemd-oomd so I'd start there.

# ? Jun 21, 2022 12:18

Kivi: Aug 1, 2006; I care

VictualSquid posted:

It sounds like it triggers the oom killer, somehow. It should say so in the syslog if that is the case. You can either change the general settings or protect the process with a specific setting. Don't remember how to do so exactly, just google oom killer.

Yeah, thanks! I tried checking the logs but alas, nothing. Thanks for the handy Google words!

Lifroc posted:

I have read the latest Ubuntu has the OOM daemon a little too conservative, killing stuff even when there's enough memory available. Might that be the case?

They seem to use systemd-oomd so I'd start there.

I'm on Ubuntu 18.04.6 LTS but I'll check the systemd-oomd anyway.

# ? Jun 21, 2022 12:21

a dingus: Mar 22, 2008; Rhetorical questions only; Fun Shoe

I upgraded my GPU from a 5700xt to a 6900xt and I'm having some weird issues that seem driver related. My displays will randomly freeze, no mouse movement etc, but I can hear things continue in the background. My system will eventually respond after doing ctrl-alt-f2 and I can use the non-gui environment. I know AMD drivers are in the kernel so I'm not sure how to reinstall or what. I updated my system after I installed the card. Any ideas? I'm on Fedora 35 w/ Wayland & sway.

# ? Jun 21, 2022 12:59

Adbot: ADBOT LOVES YOU

# ? May 18, 2024 01:03

BlankSystemDaemon: Mar 13, 2009

VictualSquid posted:

It sounds like it triggers the oom killer, somehow. It should say so in the syslog if that is the case. You can either change the general settings or protect the process with a specific setting. Don't remember how to do so exactly, just google oom killer.

Doesn't Linux have something like protect(1)?

# ? Jun 21, 2022 14:48

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > The Linux Questions Thread: a bunch of pitfalls, but technically it's possible

«‹›1029 »