cbsch.no

Linux tips

linux ubuntu

User management

Add user to group

usermod -a -G groupName userName

Disks

Read some info about file systems and disks

df -h
sudo fdisk -l /dev/sda
sudo lvdisplay
sudo pvdisplay

Rescan disk (if needed)

sudo bash -c "echo 1 > /sys/class/block/sda/device/rescan"

sudo parted
print all
fix

Resize disk

Check where your LVM partition is on the disk. This will prompt you to fix the disk, as certain things that should be on the last sectors of the disk no longer are (as the underlying disk has been extended). Letting it do the fixes is OK.

sudo parted -l

If for example the LVM partition is on /dev/sda3, run this command to extend it to use 100% of the free space on the disk

sudo parted /dev/sda resizepart 3 100%

Check the name of the physical volume for lvm

sudo pvdisplay

Then we need to resize the physical volume for LVM

sudo pvresize /dev/sda3

Then we extend the logical volume (here we found the VG by typing df -h)

sudo lvresize -r -l +100%FREE /dev/mapper/ubuntu--vg-ubuntu--lv
# sudo lvresize -r -L +1G /dev/mapper/ubuntu--vg-ubuntu--lv

The -r switch from the lvresize command will automatically resize the file system. If you omit this you can manually resize the filesystem with this:

sudo resize2fs /dev/mapper/ubuntu--vg-ubuntu--lv

Kernels filling up /boot

https://askubuntu.com/questions/620266/how-does-apt-decide-how-many-old-kernels-to-keep

There is a file that is auto-generated that tells apt-get what kernels to autoremove and which ones to keep. The file that tells apt-get which kernels they are is /etc/apt/apt.conf.d/01autoremove-kernels which is generated from /etc/kernel/postinst.d/apt-auto-removal.

Usually what happens is that when you are receiving kernel updates, when the kernel version changes, say from 3.13 to 3.16, /etc/apt/apt.conf.d/01autoremove-kernels is then updated to keep the 3.16* kernels and is then set to remove all of the 3.13 kernels unless specified by the generating script to not be removed.

Forcibly cleaning up /boot

List kernels installed

sudo dpkg --list 'linux-image*' | awk '{ if ($1=="ii") print $2}' | grep -v `uname -r`

Delete any kernels with a version between 104 and 150, modify this to your needs. It could be wise to keep at least the 1 or 2 newest kernels.

sudo rm -rf /boot/*-4.4.0-{104..150}-*

Remove all but the current kernel. This assumes the kernel type is generic (all files have generic as suffix)

cd /boot
ls -1 | grep "\-generic" | grep -v $(uname -r) | xargs rm

Add new /boot partition

fdisk /dev/sdb
m # Print help
p # Print partition table

g # Create new GPT partition table (this will remove everything on the disk)
n # New partition table
p # Primary partition
+512M # Start at default, then extend 512MB
t # Change type (look at fdisk -l /dev/sda for the types on the current /boot and /boot/efi)

Format the two new partitions, mount and copy data from the old /boot partitions

# Check the file systems on the current partitions
df -Th

sudo mkfs.fat /dev/sdb1
sudo mkfs.ext2 /dev/sdb2 -L boot

sudo mkdir /mnt/boot
sudo mount /dev/sdb2 /mnt/boot
sudo cp -av /boot/* /mnt/boot/
sudo mount /dev/sdb1 /mnt/boot/efi
sudo cp -av /boot/efi/* /mnt/boot/efi/

Mount the new /boot partitions on the correct place and run update-grub

sudo umount /mnt/boot/efi
sudo umount /mnt/boot
sudo umount /boot/efi
sudo umount /boot
sudo mount /dev/sdb2 /boot
sudo mount /dev/sdb1 /boot/efi
sudo update-grub
sudo grub-install --recheck /dev/sdb

# blkid /dev/sdb1

Change /etc/fstab uuids to match new partitions

sudo blkid /dev/sdb1
sudo blkid /dev/sdb2

Networking

Test network speed

On the listening computer, listen to port 2222 and pipe everything to null

nc -v -l 2222 > /dev/null

On the connecting computer, read 512 blocks of 1024K from /dev/zero and send it with netcat to the listening computer

dd if=/dev/zero bs=1024K count=512 | nc -v $HOSTNAME 2222

Misc

Restart networking

netplan apply

Check listening ports

sudo lsof -i @127.0.0.53:53

Iptables

Sources:

Drop incoming requests from ip

iptables -I INPUT -s 1.2.3.4 -j DROP

Drop only from specified interface

iptables -I INPUT -i eth0 -s 1.2.3.4 -j DROP

Show rules (-v for verbose, -n for supressing reverse DNS lookups)

sudo iptables -L -v -n

Tcpdump

Basic usage:

# tcpdump -i <interface> <filter>
sudo tcpdump -i eth0 port 443

Some useful switches:

Updates

This will set set –force-confold (force old configuration) and –force-confdef (force default configuration), hopefully preventing prompts

Unattended upgrade

apt-get -q -y -o DPkg::Options::=--force-confold -o DPkg::Options::=--force-confdef upgrade

Crontab update

0 1 * * * bash -c 'for i in update {,dist-}upgrade auto{remove,clean}; do apt $i -y; done'
0 2 * * * bash -c 'if [ -f /var/run/reboot-required ]; then /sbin/reboot; fi'

SSL

Converting from pfx to pem

openssl pkcs12 -in cert.pfx -nodes -nokeys -chain -out domain.cer
openssl pkcs12 -in cert.pfx -nodes -nocerts -out domain.key

Hot add CPU

This script was created for adding CPUs in VMWare. However it should work in all cases (Tested on Hyper-V)

#!/bin/bash
# William Lam
# http://engineering.ucsb.edu/~duonglt/vmware/
# hot-add cpu to LINUX system using vSphere ESX(i) 4.0
# 08/09/2009

for CPU in $(ls /sys/devices/system/cpu/ | grep cpu | grep -v idle)
do
    CPU_DIR="/sys/devices/system/cpu/${CPU}"
    echo "Found cpu: \"${CPU_DIR}\" ..."
    CPU_STATE_FILE="${CPU_DIR}/online"
    if [ -f "${CPU_STATE_FILE}" ]; then
        STATE=$(cat "${CPU_STATE_FILE}" | grep 1)
        if [ "${STATE}" == "1" ]; then
            echo -e "\t${CPU} already online"
        else
             echo -e "\t${CPU} is new cpu, onlining cpu ..."
             echo 1 > "${CPU_STATE_FILE}"
        fi
    else
        echo -e "\t${CPU} already configured prior to hot-add"
    fi
done

Copyright (c) 2023 Christopher Berg Schwanstrøm