These are just notes for command line stuff I have learned over the years: shortcuts and so on. Some are commands that I keep forgetting, or get messed up on the order. They are in no real order except the most recent discoveries are often on top. Unless otherwise stated, these are CLI from bash shells on Linux. These might also help someone else.
I sometimes need this for some VM deployments.
# This generates an actual random address
openssl rand -hex 6 | sed 's/\(..\)/\1:/g; s/.$//'
# Ensure the MAC address starts with a unicast, locally administered address (for private use), you can modify the first byte:
mac=$(openssl rand -hex 6 | sed 's/\(..\)/\1:/g; s/.$//'); echo "02${mac:2}"
# I like mine to be 02:02:02 for easy of sorting on the DHCP side
echo "02:02:02:$(openssl rand -hex 3 | sed 's/\(..\)/\1:/g; s/:$//')"
set -euo pipefail
# -e means "Stop the script on an error" instead of just continuing.
# -u means "unset variables." If a variable is not already set, then fail.
# This prevents "rm -rf $HOME/$DIRETCORY" types of failures when you mispelled "directory"
# -o pipefail means that if part of a pipe fails, then the whole statment fails.
set -x
# -x means "output step by step" for debuggung
# Echo "Script exited" when script exits
trap echo "Script exited" EXIT
# Clean up temp files when script exits on signals: SIGHUP, SIGINT, SIGQUIT, or SIGABRT
# trap -l will list number/signals for you
$TRASH=$(mktemp -t tmp.XXXXXXXXXX)
trap cleanup 1 2 3 6
function cleanup()
{
echo "Removing temporary files:"
rm -rf "$TRASH"
exit
}
Try to connect for 5 seconds, otherwise give up.
ssh -o ConnectTimeout=5 <user>@<IP address>
To set up forwarding for a gateway, all you have to do is run two commands on the gateway
sudo iptables -t nat -A POSTROUTING -o <outgoing interface towards internet> -j MASQUERADE
sudo sysctl -w net.ipv4.ip_forward=1
The systems behind the gateway then have to set their gateway IP to the IP interface interface on the gateway facing them.
sudoip route add default via <IP of gateway interface facing them>
You may have to set the DNS on those systems as well, but check with a ping 8.8.8.8 to make sure at least they can get to the internet.
The timeout command will run a command for a specified amount of time and then To run a bash script for one hour, you would use the following command:
timeout 3600 bash ./your_script.sh
This will become more of an issue soon, I think
echo $XDG_SESSION_TYPE
x11
Because I keep having to look this up
EHLO test.example.com
MAIL FROM: <SENDERADDRESS>
RCPT TO: <RECIPIENTADDRESS>
DATA
Subject: Testmessage
(Blank line, press Enter again)
This is a test.
(Blank line, press Enter again)
.
QUIT
Sometimes you accidentally put something in the bash history file you need to not be saved. A quick exit from the shell without saving to history is
kill -9 $$
Which will dump you out and not save to history.
A lot of systems can't have the telnet package for security reasons. But there are other ways to test ports on a remote system:
Use the /dev/tcp port:
$ cat < /dev/tcp/192.168.1.4/22
SSH-2.0-OpenSSH_6.2p2 Debian-6
^C pressed here
Use curl:
$ curl http://192.168.1.4:3389
curl: (7) couldn't connect to host
Use netcat:
nc -zv 10.6.0.1 80 # One port
nc -zv 10.6.0.1 80 443 # Multiple ports
nc -zv 10.6.0.1 20-29 # Port range
date +'%Y-%m-%d_%H-%M-%S'
2023-11-13_16-48-37
You can also just add timestamps by using "ts" (usually in the package more-utils):
echo "Hello" | ts
Dec 29 16:28:15 Hello
Note: "ts -r" can solve timestamps in relative format, for example:
$ sudo tail /var/log/syslog | ts -r
5m10s ago server-dev1 systemd[1]: fwupd-refresh.service: Succeeded.
5m10s ago server-dev1 systemd[1]: Finished Refresh fwupd metadata and update motd.
5m4s ago server-dev1 CRON[1673984]: (root) CMD (command -v debian-sa1 > /dev/null && debian-sa1 1 1)
1m4s ago server-dev1 CRON[1674253]: (root) CMD (/bin/bash /opt/heatbeat.sh > /dev/null 2>&1)
4s ago server-dev1 CRON[1674279]: (root) CMD ([ -x /etc/init.d/anacron ] && if [ ! -d /run/systemd/system ]; then /usr/sbin/invoke-rc.d anacron start >/dev/null; fi)
4s ago server-dev1 CRON[1674280]: (root) CMD (/bin/bash /opt/heatbeat.sh > /dev/null 2>&1)
There are ways to get a single element from a line using echo and awk:
$ echo "/dev/mapper/volgroup00-lv00" | awk -F'/' '{print $3}'
But another way can be using read to read into an array
PART_INFO=$(df -hT | grep ${PARTITION}) # Read in some parititon info
IFS=' ' read -ra PART_ARRAY <<< "${PART_INFO}" # Put the partition info in an array
PART_DEV=${PART_ARRAY[0]} # The device is the first part
PART_FS=${PART_ARRAY[1]} # The filesystem is the second part
IFS='/' read -ra MAPPER_ARRAY <<< "${PART_DEV}" # You can further divide the device as dev, mapper, volgroup00-lv00
You can also use mapfile
mapfile -d' ' foo <<< $(df -hT | grep /dev/mmcblk0p2); echo ${foo[1]}
I used to create a function call "fLog" that would log everything, but there's an even quicker way:
OUTPUT_FILE="/path/to/output.log"
echo "After this line, all output goes to $OUTPUT_FILE"
exec >> "$OUTPUT_FILE" 2>&1
I'd often just output things like "INFO - sometext" and "ERR - someerror" and then do a grep on $OUTPUT_FILE. The script wouldn't output anything, but if I needed to see what the script was doing, I could comb through the log. If you want to output to the terminal as well"
LOGPATH="./testlog.log"
echo "This will only go to the screen"
exec > >(tee -a ${LOGPATH}) 2> >(tee -a ${LOGPATH} >&2)
echo "This, and further text will got to the screen and log"
You can also run "script" which records a session:
script -a ./log_session.log
A lot of sites tell you about /etc/inittab that don't quite work in modern systems. Ignore those if you have systemd (which most modern systems do). In my case, I had a 9-pin serial connection at /dev/ttyS0. Here's how I found out the bit rate:
:~$ stty -F /dev/ttyS0 speed
9600
:~$ stty -F /dev/ttyS0 115200 # To set at 115200
:~$ stty -F /dev/ttyS0 speed
115200
I made sure my dumb terminal was set at 8 bit per character, no parity, and 1 stop-bit. Then, for systemd, I had to attach the serial to the getty service:
:~$ sudo systemctl enable serial-getty@ttyS0.service --now
Created symlink /etc/systemd/system/getty.target.wants/serial-getty@ttyS0.service → /lib/systemd/system/serial-getty@.service.
I wanted to have some color to my bash prompts:
localhost$ read -p $'\e[31;1mEnter in your codex:\t\e[0m ' FOO; echo -e "\e[32mCodex you entered:\t\e[0;1m $FOO"
Enter in your codex: 45645674 # This will be red except for the codex entered
Codex you entered: 45645674 # This will be green except for the codex
I had a series of boxes I had to log into via a AWS key, and I found I could just add it to a "keychain" via an agent.
$ ssh-agent bash
$ ssh-add ~/aws-setup-ec2-key.pem
$ ssh-add -l # to list what keys are loaded
There's also a way with passwords, if sshpass is installed. This takes a list of hosts from "hosts_file" and runs through them. Note the space in front of the "SECRETPASS," it's supposed to keep it from showing up in your history, but this is still at-your-own-risk:
[space] SECRETPASS="passwd123";\
for HOSTLIST in $(cat hosts_file);\
do sshpass -p $SECRETPASS ssh -q username@$HOSTLIST 'uptime';\
done
Say you have an LVM-system, where the root is LVM. Some of the newer Red Hat amis on AWS have this, but this also applies to a drive rescue. In any case, you have a "foreign LVM structure" mounted onto another system. In this case, we're going to say an AWS rescue: an fstab was improperly edited or corrupted, and the ec2 instance hangs.
First, shut down the bad instance.
Next, set up a rescue instance made with the SAME KEY as the unbootable system, with the bad volume as an attachment (but obviously not the boot/root drive), spin up the rescue instance with LVM service enabled, with the volume (or hard drive from the dead system) enabled.
Note: if the system is working fine, and you just want to move the Volume Group (VG), then you need to mark the VG as offline and export it.
# vgchange -an vg_opt
vgchange -- volume group "vg_opt" successfully deactivated
# vgexport vg_opt
vgexport -- volume group "vg_opt" successfully exported
Then log on to the rescue instance using the ec2-user account with the pem key
ssh -i old_instance.pem ec2-user@<ip address>
Show that the device is attached
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
xvda 202:0 0 8G 0 disk
└─xvda1 202:1 0 8G 0 part /
xvdf 202:80 0 100G 0 disk
├─xvdf1 202:81 0 500M 0 part
├─xvdf2 202:82 0 49.5G 0 part # Attached old LVM2 drive <<<<<<<<<<
│ ├─vg_root-var 253:2 0 15G 0 lvm
│ ├─vg_root-home 253:3 0 10G 0 lvm
│ ├─vg_root-tmp 253:4 0 5G 0 lvm
│ ├─vg_root-root 253:5 0 18.8G 0 lvm # This is the LVM2 partition we want to mount
│ └─vg_root-swap 253:6 0 2G 0 lvm
└─xvdf3 202:83 0 50G 0 part
├─vg_root-opt 253:0 0 4G 0 lvm
└─vg_root-app 253:2 0 15G 0 lvm
Run pvscan to see the LVM structure
sudo pvscan
PV /dev/xvdf4 VG vg_opt lvm2 [<1000.00 GiB / 0 free]
PV /dev/xvdf2 VG vg_root lvm2 [99.00 GiB / <26.67 GiB free]
Total: 2 [1.07 TiB] / in use: 2 [1.07 TiB] / in no VG: 0 [0 ]
Use lvdisplay to find the correct LV to mount for the root filesystem
[...]
--- Logical volume ---
LV Path /dev/vg_root/root # <<<<<<<<<< This is the path we want to use below
LV Name root
VG Name vg_root
LV UUID QEdQYT-6QG9-AauM-7foI-jXtQ-3lGX-l6m64c
LV Write Access read/write
LV Creation host, time localhost, 2017-09-18 14:31:14 +0000
LV Status available
# open 0
LV Size <18.79 GiB
Current LE 4810
Segments 2
Allocation inherit
Read ahead sectors auto
- currently set to 256
Block device 253:5
[...]
Note: in some cases, you might need to do a "vgimport" which is slightly different on each LVM type. For LVM1, you have to add ALL PV's if there is more than one in the VG, separated by spaces
vgimport vg_root # for LVM2
vgimport vg_root /dev/xvdf2 # for LVM1 as per above
vgimport vg_root /dev/xvdf2 /dev/xvdf4 # for LMV1 if there were 2 PVs
vgchange -ay vg_root # Optional: only do this if the volume was disabled and exported
You should be able to mount it now, in this case /mnt
[root@ip-10-12-34-56 ~]# mount /dev/vg_root/root /mnt
[root@ip-10-12-34-56 ~]# df -h /mnt
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/vg_root-root 119G 7.0G 112G 38% /mnt
Then copy the old bad fstab to study later:
[root@ip-10-12-34-56 ~]# cp /mnt/etc/fstab /mnt/etc/fstab.bad.2022-12-30
Edit the fstab, and make the corrections. It's best to make the system minimally bootable but commenting out everything EXCEPT what you need: keep root, var, or whatever to make the systems bootable, and comment out an CR-ROM drives, non-root USB drives, nfs, smb, and so on.
Shut down the rescue instance, and boot the old instance with the volume mounted as before.
Then, if it boots successfully, you can uncomment the line item in fstab one by one after after boot and do a "mount -av" to remount the filesystems on the live system.
chronyd -q 'server SERVER-ADDRESS iburst'
Note that chronyd must not be already running, just as ntpd must not be running when setting the date with ntpdate. If that doesn't work, after that you may need an additional:
sudo systemctl start chronyd; sudo chronyc -a 'burst 4/4'
sed '/^u/d' file # Remove all lines that start with 'u'
sed '/x$/d' file # Remove all lines that end with 'x'
sed '/^[A-Z]*$/d' file # Remove all lines that have capital letters
sed '/debian/d' file # Remove all lines that have the word 'debian'
sed 'M,Nd' file # Remove lines M to N (don't forget the small d)
sed 'Nd' file # Remove line N from file
rsync -auvz -e "ssh -i ~/.ssh/somekey" codebase/foo user@hostname:/var/www/app
It can also pull the IdentityKey, Host, and User from .ssh/config if you have one.
If you have an encrypted ansible vault, with all your passwords and so on, and you have THAT locked by a password (you should), sometimes it makes the command line laborious. I have created two aliases:
alias apvault='ansible-playbook --vault-password-file=~/.ssh/vault_pass'
alias avault='ansible --vault-password-file=~/.ssh/vault_pass -b'
The ~/.ssh/vault_pass is the key and the -b is the "become" needed for root access when just running ansible commands.
Something near and dear to my heart is the openness of sharing ideas, discoveries, and code. One of my pet peeves is, every once in a while, a sysadmin or coder comes along that doesn't want to share. They use basic ways to make their code unreadable on purpose, maybe to secure their jobs, hide the fact what they don't know for better coders, or just to be difficult. This is something I ran into years ago. Say you have a script, ten.sh:
#!/usr/bin/env bash
echo "This is a bunch of numbers"
for foo in $(seq 1 10); do echo -n "$foo - "; done
If you don't want people to see what that script is doing, you can do this:
echo '#!/usr/bin/env bash' > secret_ten.sh && echo "echo '$(base64 ten.sh)' | base64 -d | sh" >> secret_ten.sh
If you cat secret_ten.sh you will get this:
#!/usr/bin/env bash
echo 'IyEvdXNyL2Jpbi9lbnYgYmFzaAplY2hvICJUaGlzIGlzIGEgYnVuY2ggb2YgbnVtYmVycyIKZm9y
IGZvbyBpbiAkKHNlcSAxIDEwKTsgZG8gZWNobyAtbiAiJGZvbyAtICI7IGRvbmUK' | base64 -d | sh
And you can still run it. If you want to see what that hash is, just edit secret_ten.sh and remove the | sh at the end. There are "more secure" ways to make this even harder to decode, but I don't want to encourage this behavior. I have some advice for people who do this, and things like it, to "secure their jobs." I have seen so many sysadmins and coders get let go anyway, either because they are not as valuable as they thought they were, or get let go by management who doesn't even know their value, no matter how obvious or stubborn you are. All you're doing is making it hard for other coders down the line. And you might think, "well, who cares about them, I'm out of a job," and you're part of the problem, I'm afraid.
sort -uo filename.list filename.list
awk '!x[$0]++' /etc/postfix/transport > /tmp/transport_dupes_removed
curl ifconfig.me
curl -4/-6 icanhazip.com
curl ipinfo.io/ip
curl api.ipify.org
curl checkip.dyndns.org
hostname -i
Sometimes I am just testing an new concept or distro, and I install in directly to RAM. RAM is cheap these days, and why clutter my hard drive and wear down my SSD with files I am gonna delete right away, anyway?
sudo mkdir /tmp/ramdisk
sudo mount -t tmpfs -o size=12G vmramdisk /tmp/ramdisk
mount | tail -n 1
Then I set the install directory to /tmp/ramdisk, play with it, then destroy it. If I need the RAM back, I can just unmount it.
sudo umount /tmp/ramdisk/
Or make the RAM disk permenant and persistent (athough data will be wiped between reboots).
vmramdisk /tmp/ramdisk tmpfs defaults,size=12G,x-gvfs-show 0 0
sudo mount -t cifs //diskstation.int.local/Fortran /media/diskstation/fortran -o user=cforrester
passwd: ****************
I had an issue where I had to keep an eye out for runaway php processes, and have a more visual way to eyeball it on another screen. I whipped up this little one-liner to show me the number of php proccesses every 10 seconds:
while true; do date | tr -d '\n'; foo=$(nice ps aux | grep php | wc -l);\
echo -n " :: $foo"; printf '%*s' $foo | tr ' ' '*'; echo; sleep 10; done
This prints out a crude bar graph:
Fri Apr 16 12:41:51 UTC 2021 :: 10**********
Fri Apr 16 12:42:01 UTC 2021 :: 11***********
Fri Apr 16 12:42:11 UTC 2021 :: 26**************************
Fri Apr 16 12:42:21 UTC 2021 :: 27***************************
Fri Apr 16 12:42:31 UTC 2021 :: 27***************************
Fri Apr 16 12:42:41 UTC 2021 :: 26**************************
Fri Apr 16 12:42:51 UTC 2021 :: 15***************
Fri Apr 16 12:43:01 UTC 2021 :: 12************
Fri Apr 16 12:43:11 UTC 2021 :: 10**********
Fri Apr 16 12:43:21 UTC 2021 :: 10**********
adduser --shell /bin/bash --disabled-password jenkins
This assumes that the first disk is /dev/sda, and you're adding a third [3] partition and using LVM
- Power off the virtual machine
- Extend the virtual disk size
- Power on machine
sudo fdisk /dev/sda- p,n,p,3[probably],[enter][enter],t,3[probably],8e[LVM type],w
- Restart machine
sudo pvcreate /dev/sda3sudo vgextend vgubuntu /dev/sda3sudo vgdisplay vgubuntu | grep "Free"sudo lvextend -L+##G /dev/vgubuntu/root[where ## is the GB size from previous step]sudo resize2fs /dev/vgubuntu/root[use ext2online if resize2fs gives an error about live filesystems]- You can also use "lvextend -r" instead of resize2fs or xfs_growfs
- Pray
- Expend the EBS in the console or
aws ec2 modify-volume --size 150 --volume-id vol-1234567890abcdef0where 150 is the size in GB - Then ssh into the instance, check with
sudo lsblkanddf -h sudo growpart /dev/xvda 1where xvda 1 [note the space] is the disk volume, parition 1sudo resize2fs /dev/xvda1where xvda1 [no space] is the partition you expanded above- Prayer not needed
- Expend the EBS in the console or
aws ec2 modify-volume --size 150 --volume-id vol-1234567890abcdef0where 150 is the size in GB - Then ssh into the instance, check with
sudo lsblkanddf -h xfs_growfs /mntwhere /mnt is the mounted disk volumexfs_fsrto optimize Note: AWS takes a LONG time to optimize from the EBS, and depending on size, may not be done for several hours to half a day.
find /path/to/files* -mtime +5 -exec rm -i {} \;
In this case, 5. Note that there are spaces between rm, {}, and ; and I put an -i for safety in case one cut and pastes. I then hand-fix it to -rf or -r depending on what I need. Often I am removing log files over 60 days old to save space on really small systems.
If you are dealing with files, the difference is between mtime, ctime and atime is:
- atime "access time," is updated when the file's contents are read by an application or a command such as less, grep, or cat.
- ctime "change time," is when the file's property changes. It will always be changed when the mtime changes, but also when you change the file's permissions, name or location.
- mtime "modification time," is when the file was last modified. When you change the contents of a file, its mtime changes.
The easiest way to remember which is which is to read their alphabetical order:
- Atime can be updated alone
- Ctime will update atime
- Mtime will update both atime and ctime.
You can change the message you get when someone logs into sudo. Set lecture=always, never, or once and the message can be set to a text file
Defaults lecture=once,lecture_file=/etc/sudo_lecture.txt
# sudo --reset-timestamp # To reset this "once"
Defaults insults
# You can insult people who get the sudo password wrong, too
Suppose you have an output with columns, and the first, second, or nth column has a match you're looking for, and if that column matches, you print the entire line. You can use awk, which is designed for this:
awk '$2 == "LINUX" { print $0 }' test.txt # Show lines that have LINUX in the second column
iptables -vL | awk '$1 > 0 { print $0 }' # Show output of IP tables when the first column (packet hits) is greater than zero
grep -E -o "\b[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\.[A-Za-z]{2,6}\b" email_log.log
grep -E -o "(25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)\.(25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)\.(25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)\.(25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)" list_of_maybe_ips.txt
The keys are accessed using an exclamation point: ${!array[@]}, the values are accessed using ${array[@]}
for i in "${!array[@]}"
do
echo "key : $i"
echo "value: ${array[$i]}"
done
In standard bash arrays, ${!array[@]} will list indecies, like 0 1 2 ..n
If you create a file in /etc/sudoers.d/<username> and understand that this is a vulnerability you need to secure HARD... add the line:
<username> ALL=(ALL) NOPASSWD:ALL
You can substitute commands for files using parentheses. For example:
diff <(ls Directory/*.png) <(ls Backups/*.png)
python3 -m http.server [port number]
wget http://<hostip>:[port number]/filename
Launch the web server [default port 8000] and transfer files that way. Note: only do this on private networks.
cat test.json | python -mjson.tool
cat test.json | jq
Note, either python or jq have to be installed, respecively.
Surprisingly, the trick is to append a comma (,)
# Host and IP address
ansible all -i example.com,
ansible all -i 93.184.216.119,
# Requires 'hosts: all' in your playbook
ansible-playbook -i example.com, playbook.yml
$ # "Nothing's worth more than $20," according to Jim.
$ !:q
'# "Nothing'\''s worth more than $20," according to Jim.'
$ echo '# "Nothing'\''s worth more than $20," according to Jim.'
# "Nothing's worth more than $20," according to Jim.
To update all your vagrant boxes:
vagrant box list | while read line; do BOX=$(echo $line | awk '{print $1}'); vagrant box update --box $BOX; done
To update and prune all your outdated boxes, even ones in use:
vagrant box list | while read line; do BOX=$(echo $line | awk '{print $1}'); vagrant box update --box $BOX; done && vagrant box prune -f
I have a lot of aliases for vagrant because I use it a LOT:
alias vagpro='vagrant provision'
alias vagredo='vagrant destroy -f && vagrant up'
alias vagrm='vagrant destroy -f'
alias vagssh='vagrant ssh'
alias vagstat='vagrant global-status'
alias vagupdate="vagrant box list | while read line; do BOX=$(echo $line | awk '{print $1}'); vagrant box update --box $BOX; done"
alias vagtrim="vagrant box list | while read line; do BOX=$(echo $line | awk '{print $1}'); vagrant box update --box $BOX; done && vagrant box prune -f"
alias vagwho='vagrant global-status'
alias vgdn='vagrant halt'
alias vgdown='vagrant halt'
alias vgssh='vagrant ssh'
alias vgup='vagrant up'
I have found myself in multiple jobs trying to figure out all kinds of cron jobs that are not standardized when I take over a server. Are they in /etc/crontab? Running as a crontab of a user? Something in one of the hours jobs? I developed these four lines of code to find them quickly.
# Any active jobs in /etc/crontab
grep -v "^#\|^$" /etc/crontab
# Any active jobs in the timed directories
for foo in hourly daily weekly monthly; do echo "==== Cron directory $foo>"; ls /etc/cron.$foo; done
# Any active jobs in the cron.d directory
ls /etc/cron.d/* # grep -v "^#\|^$"
# Any cron jobs in user crontabs
for foo in $(ls /var/spool/cron/crontabs/*); do echo "==== User Crontab $foo>"; grep -v "^#\|^$" $foo; done
sudo systemctl --type=service
sudo systemctl list-units --type=service
sudo systemctl --type=service --state=active
sudo systemctl --type=service --state=running
Your best bet is to run this as an alias:
alias sys_active='sudo systemctl --type=service --state=active'
alias sys_run='sudo systemctl --type=service --state=running'
timeout 8s ping 192.168.1.1
Run "uname -m" to confirm. If it says aarch64 then it is 64 bit. If it says armv7l then it is 32 bit. If it says armv61, it's also 32 bit.
This is helpful to find the highest level of TLS that you support.
openssl s_client -connect [domain]:443 -[tls1/tls1_1/tls1_2/tls1_3]
Sometimes, you have a long command that you need to edit by hand or swap some stuff out. Like say you had
ansible-playbook -T 60 --extra-vars="foo=bar bar=baz xyzzy=plugh" install-rsyslog-client.yml --vault-password-file=~/.ssh/vault_pass -l manny,mo,jack,cecil,beany,potato1,orangepi1,yakko,rpi-*
and the next command, substitute a lot of stuff no easy in one command, run:
fc
It will open up that command in your DEFAULT text editor. What, it's crummy old nano? Change it to vim. Open the .bashrc file, and replace both occurrences of program with the editor you want to set as the default editor:
export EDITOR='vim'
export VISUAL='vim'
Save and reload with "source .basrc"
Why was this so hard to find?
vcgencmd measure_temp
cat /sys/class/thermal/thermal_zone0/temp
After the post below, I needed a reminder when I look at this page again.
Handle Name Description
0 stdin Standard input
1 stdout Standard output
2 stderr Standard error
command &>file # ... OR
command > file-name 2>&1 # ... pipes both stdout and stderr in a file
command 2>&1 | less # ... it can also be piped
command 2> error.logr.log # put only stderr in a log
You can also pipe both standard output and standard error into the next stage of the pipeline like this:
if doesnotexist |& grep 'command not found' >/dev/null
then
echo "oops"
fi
This just seems super annoying, but you have to push all the output to stdout, then grep it, like this example [-f is otional "follow" which I use more often than not]:
docker logs -f 2434326a55f9 2>&1 | grep "No protocol specified"
set number
set relativenumber
You can also put the following two config directives in ~/.vimrc
csvkit provides in2csv, csvcut, csvjoin, csvgrep, etc.
You can hit alt+shift+3 which will put a "#" comment in front of it so you can go back to it later in your bash history
What does this box have? I knew a lot like dmidecde, dmesg, and kernel logs, but there are some new ones:
# Read BIOS settings
dd if=/dev/mem bs=1k skip=768 count=256 2>/dev/null | strings -n 8
lspci -tv # Show PCI devices
lsusb -tv # Show USB devices
lshal # Show a list of all devices with their properties
lsdev # Information about installed hardware (procinfo has to be installed)
lshw # Information about installed hardware
Use flock (part of the util-linux or flock suite of tools depending on distro).
$ crontab -l
* * * * * /usr/bin/flock -w 0 /path/to/cron.lock /usr/bin/php /path/to/cron.php
You can also use fuser to see who and what are using them
$ fuser -v /path/to/cron.lock
USER PID ACCESS COMMAND
cron.lock: root 7836 f.... flock
root 7837 f.... php
There is a program ccze (not a default install) which replaces colorize but is much faster. Sadly, it doesn't page by default, so you have to modify the output to ascii and pipe it to less like so:
ccze -A < /mnt/log/mail.log | less -R
The programs ansible-lint and git hate trailing whitespace, so often I have to remove them from a file with
sed -i 's/\s*$//g' some_file.yml
If you need to escape linux telnet, you hit control + ] which is the escape key "^]". But you can also change it at the command prompt like:
telnet -e Q 192.168.0.1 25
Which would set the escape to "Q"
I kept getting errors in KDE/sddm a.k.a. Kubuntu that would pop up two dialoge boxes that said:
Configuration file "//.config/kwalletd5rc" not writable. Please contact your system administrator.
Configuration file "//.kde/share/config/kwalletdrc" not writable. Please contact your system administrator.
It did that at every fresh sddm login. For Kubuntu 14.04, 16.04, and even 18.04. Why is this happening when I don't even USE kwallet? I did a lot of research, and it turns out that this is due to some bug nobody has fixed for some reason when the pam.d modules are loaded, and they ask for /{$HOME}/[config stuff] and $HOME is not defined. So just disable these 4 lines in /etc/pam.d/sddm
# -auth optional pam_kwallet.so
# -auth optional pam_kwallet5.so
and
# -session optional pam_kwallet.so auto_start
# -session optional pam_kwallet5.so auto_start
When you restart, those boxes go away. I am not sure what this will do when you USE kwallet and its subsystem (probably nothing good), but that wasn't my problem.
I had a project that I had saved in multiple k3b files. Years and years went by. I had to see what files were saved and recreate burning audio CDs from them. First, I had to find CD blanks. But the next thing was "what the heck are k3b file data stored as?" I went through a lot of research, and found they are just zipped files with metadata in an XML format stored in them as "maindata.xml." I had to install the xsltproc and figure out the stylesheet to decode it (included in this repo). This is imperfect, but it works. I found out what files they had, and was able to recreate them in K3B or another burner program.
for foo in $(ls *.k3b); do
unzip -p $foo maindata.xml > "$foo.xml"
xsltproc k3b_stylesheet.xslt $foo.xml > "$foo.cd.txt"
rm $foo.xml
done
Sometimes I forget what this is called, like msconfig32 or wincfg32 or something. Always forget it's:
"C:\Program Files\Common Files\Microsoft Shared\MSInfo\msinfo32.exe"
grep -v "^#\|^$" filename.conf
Sometimes you get a weird box that doesn't want to or just can't install guest additions in VM, so you have to add some tweaks in the vagrantfile in the Vagrant.configure section (usually most of the file):
Vagrant.configure("2") do |config|
....
if Vagrant.has_plugin?("vagrant-vbguest")
config.vbguest.auto_update = false
end
....
end
sort -n -t . -k 1,1 -k 2,2 -k 3,3 -k 4,4 ./list_of_addresses.txt
For those times you want to weed out time stamps, this will print all but two first columns:
awk '{$1=$2=""; print $0}' somefile
Find the index where you want to start printing (index of $5 in $0) and print the substring of $0 starting at that index.
awk '{print substr($0,index($0,$5))}' fileName
When, for some reason, my local master branch has some changes I don't care about, to reset to origin/master
git fetch origin
git reset --hard origin/master
git clean -n -f # Clean localfiles (show only)
git clean -f # Clean localfiles (for reals)
git clean -d -f # Also clean local directories
If you haven't pulled recently, when branch is out for long time and you need to add changes made to origin/master since
git rebase master
sudo update-alternatives –config x-www-browser
Add timeout option to read command, but in case of interactive execution, user can easily skip waiting.
read -p "Press any key to continue (auto continue in 30 seconds) " -t 30 -n 1
useradd - create a user, but not set any passwords or home directories:
good for scripts (useradd -D adds defaults)
userdel - delete user (userdel -f is a complete purge of home directory, mail spool, etc)
adduser - The interactive shell
if [ "$(whoami)" != 'root' ]; then
echo -e "\e[31;1m$0: ERROR: You have no permission to run $0 as non-root user.\e[0m"
exit 1;
fi
Some variables and packages in this Ubuntu example:
SUBNETINT="192.168.11.0/24"
apt-get update
apt-get upgrade -y
apt-get install -y netfilter-persistent
If ufw is on this system, disable it
ufw disable
Clear out any old rules
iptables -P INPUT ACCEPT
iptables -P INPUT ACCEPT
iptables -P FORWARD ACCEPT
iptables -F
iptables -Z
Make sure we don't hang up on ourselves, and keep ssh port open
iptables -A INPUT -m state --state ESTABLISHED,RELATED -j ACCEPT
iptables -A INPUT -p tcp --dport 22 -j ACCEPT
iptables -A INPUT -i lo -j ACCEPT
Set UDP ports for a VPN
iptables -A INPUT -p udp --dport 500 -j ACCEPT
iptables -A INPUT -p udp --dport 4500 -j ACCEPT
Set policy routing
iptables -A FORWARD --match policy --pol ipsec --dir in --proto esp -s ${SUBNETINT} -j ACCEPT
iptables -A FORWARD --match policy --pol ipsec --dir out --proto esp -s ${SUBNETINT} -j ACCEPT
iptables -t nat -A POSTROUTING -s ${SUBNETINT} -o eth0 -m policy --pol ipsec --dir out -j ACCEPT
iptables -t nat -A POSTROUTING -s ${SUBNETINT} -o eth0 -j MASQUERADE
iptables -t mangle -A FORWARD --match policy --pol ipsec --dir in -s ${SUBNETINT} -o eth0 -p tcp -m tcp --tcp-flags SYN,RST SYN -m tcpmss --mss 1361:1536 -j TCPMSS --set-mss 1360
Block all other connections
iptables -A INPUT -j DROP
iptables -A FORWARD -j DROP
Save that, needs package "iptables-persistent"
netfilter-persistent save
netfilter-persistent reload
netstat -peanut
ubuntu@vpn-strongswan:~$ netstat -peanut
(Not all processes could be identified, non-owned process info
will not be shown, you would have to be root to see it all.)
Active Internet connections (servers and established)
Proto Recv-Q Send-Q Local Address Foreign Address State User Inode PID/Program name
tcp 0 0 0.0.0.0:22 0.0.0.0:* LISTEN 0 15747 -
tcp 0 0 10.42.5.36:22 10.3.7.69:3151 ESTABLISHED 0 20138 -
tcp 0 0 10.42.5.36:22 10.3.7.69:36800 ESTABLISHED 0 24746 -
tcp 0 376 10.42.5.36:22 10.3.7.69:52220 ESTABLISHED 0 20293 -
tcp 0 0 10.42.5.36:22 10.3.7.69:62004 ESTABLISHED 0 23877 -
tcp 0 0 10.42.5.36:22 10.3.7.69:41994 ESTABLISHED 0 24438 -
tcp6 0 0 :::22 :::* LISTEN 0 15760 -
udp 0 0 0.0.0.0:68 0.0.0.0:* 0 14371 -
List open interface: lsof -i
vagrant@ubuntu-xenial:~$ sudo lsof -i
COMMAND PID USER FD TYPE DEVICE SIZE/OFF NODE NAME
dhclient 874 root 6u IPv4 11617 0t0 UDP *:bootpc
chronyd 1402 _chrony 2u IPv4 17457 0t0 UDP localhost:323
chronyd 1402 _chrony 3u IPv6 17458 0t0 UDP ip6-localhost:323
sshd 1424 root 3u IPv4 16654 0t0 TCP *:ssh (LISTEN)
sshd 1424 root 4u IPv6 16663 0t0 TCP *:ssh (LISTEN)
sshd 2272 root 3u IPv4 37525 0t0 TCP 192.168.111.15:ssh->192.168.111.2:33696 (ESTABLISHED)
sshd 2334 vagrant 3u IPv4 37525 0t0 TCP 192.168.111.15:ssh->192.168.111.2:33696 (ESTABLISHED)
List open interface for just port 22: lsof -i :22
vagrant@ubuntu-xenial:~$ sudo lsof -i :22
COMMAND PID USER FD TYPE DEVICE SIZE/OFF NODE NAME
sshd 1424 root 3u IPv4 16654 0t0 TCP *:ssh (LISTEN)
sshd 1424 root 4u IPv6 16663 0t0 TCP *:ssh (LISTEN)
sshd 2272 root 3u IPv4 37525 0t0 TCP 192.168.111.15:ssh->192.168.111.2:33696 (ESTABLISHED)
sshd 2334 vagrant 3u IPv4 37525 0t0 TCP 192.168.111.15:ssh->192.168.111.2:33696 (ESTABLISHED)
This was handy with the Raspberry Pi when I ssh'd in and had to display an output on the attached screen
setsid sh -c 'exec command <> /dev/tty2 >&0 2>&1'
Win + shift + s
xinput list
You will get output like:
Virtual core pointer id=2 [master pointer (3)]
Virtual core XTEST pointer id=4 [slave pointer (2)]
SynPS/2 Synaptics TouchPad id=12 [slave pointer (2)]
Virtual core keyboard id=3 [master keyboard (2)]
Virtual core XTEST keyboard id=5 [slave keyboard (3)]
Power Button id=6 [slave keyboard (3)]
Video Bus id=7 [slave keyboard (3)]
Power Button id=8 [slave keyboard (3)]
Sleep Button id=9 [slave keyboard (3)]
Laptop_Integrated_Webcam_1.3M id=10 [slave keyboard (3)]
AT Translated Set 2 keyboard id=11 [slave keyboard (3)]
Dell WMI hotkeys id=13 [slave keyboard (3)]
Then type:
xinput set-prop 12 "Device Enabled" 0
fail2ban-client set ssh unbanip 123.123.123.123
Useful for things that need separate lines:
extra_lines: |
first line
second line
third line
Will display:
1. first line
2. second line
3. third line
Useful for things that need to be indented but in reality is all one line
extra_lines: >
first line
second line
third line
Will display:
first line second line third line
Other stuff:
--forks # The default is 5.
ansible-playbook -u ubuntu -i inventory/dev/ec2.py --vault-password-file=~/.ssh/ansible_vault_pass.txt dev-deploy.yml --private-key=~/.ssh/dev-key.pem --limit 192.168.5.4 --list-hosts
--check # Will do a dry run
--syntax-check # Will check yaml structure
--list-hosts # Will lists hosts it would run on
The order is L->R: is User/Owner, Group, Everyone/Other
-rwxr-xr-- 1 amrood users 1024 Nov 2 00:10 myfile
- User has read, write, execute permissions
- Group "users" has read, execute permissions
- Everyone else has read-only permissions
chmod 0 = no permissions
acl has to be enabled on disk first:
/dev/sdc1 /encrypt ext3 defaults,acl 0 0
Remount the disk if needed:
mount -a -o remount
set and read the acl:
setfacl -Rm "g:dev-interface:rw" /encrypt/sftp
getfacl /encrypt/sftp
Get IP address of running guest:
VBoxManage guestproperty get {UUID no brackets} "/VirtualBox/GuestInfo/Net/0/V4/IP"
Get list of vms/running vms:
VBoxManage list vms/runningvms
sudo update-alternatives --config editor
ssh-keygen -t ed25519
CTRL+T : transpose the character before the cursor with the one under the cursor, press Esc + t to transposes the two words before the cursor.
CTRL+X+backspace : delete all text from the beginning of line to the cursor.
CTRL+A : move to the beginning of line.
CTRL+B : move back one character
CTRL+D : if you've type something, Ctrl + d deletes the character under the cursor, else, it escapes the current shell.
CTRL+N : same as Down arrow.
CTRL+P : same as Up arrow.
CTRL+E : move to the end of line.
CTRL+F : move forward one character
CTRL+I : insert a tab
CTRL+K : cut everything after pointer
CTRL+K : delete all text from the cursor to the end of line.
CTRL+U : cut everything before pointer
CTRL+W : delete word before pointer
CTRL+Y : paste erased
CTRL+_ : undo typing.
CTRL+insert : copy
CTRL+left/right arrow : Go to beginning of word
SHIFT+insert: paste
CTRL+C : SIGTERM
CTRL+L : redraw screen (helpful if something messes up formatting)
CTRL+S : to stop output to terminal.
CTRL+Q : to resume output to terminal after CTRL+S
CTRL+R : Search history
set -o vi
vmstat
free -h
perf (command) # May need linux-tools installed
dstat -ta # Live stats on cpu/disk/network
Say you've got a utility, and you think it's trying to read a config file, but you don't know what path it's looking for.
strace -o [output file]
find /var/log -mmin -5
apt-get install socat
ncdu screen telnet vim-enhanced multitail lynx bind-utils curl wget openssh-clients lsof net-snmp rsync mlocate man zip unzip htop at yum-utils tree tmux policycoreutils-python boxes figlet sysstat tcpdump
semanage port -l | grep ssh
semanage port -a -t ssh_port_t -p tcp 222
In /etc/pam.d/sshd
session optional pam_exec.so /usr/local/bin/send-mail-on-ssh-login.sh
grep -C2 "foo" ./milefile.txt
command | xclip
For logs that get rotated, you may want tail -F or tail --follow=name so you start following the new log rather than the old file that got moved out.
tail -F rotating.log
tail -f /var/log/*.log is very useful!
[Enter] [~] [.]
More of these escape sequences can be listed with [Enter] [~] [?]
Supported escape sequences:
~. - terminate session
~B - send a BREAK to the remote system
~R - Request rekey (SSH protocol 2 only)
~# - list forwarded connections
~? - this message
~~ - send the escape character by typing it twice
(Note that escapes are only recognized immediately after newline.)
export HISTCONTROL=ignoredups:erasedups # no duplicate entries
export HISTSIZE=100000 # big big history
export HISTFILESIZE=100000 # big big history
shopt -s histappend # append to history, don't overwrite it
# Save and reload the history after each command finishes
export PROMPT_COMMAND="history -a; history -c; history -r; $PROMPT_COMMAND"
awk -F "=|,"
:w !sudo tee %
ctrl-x-e
export EDITOR=/usr/bin/vim (to change from nano, put in .bashrc to make permanent)
!!:gs/foo/bar
(As opposed to ^foo^bar, which only replaces the first occurrence of foo, this one changes every occurrence.)
while true; do echo -ne "\e[s\e[0;$((COLUMNS-27))H$(date)\e[u"; sleep 1; done &
[esc] *
UTF-8 to get proper lines and ncurses and remove â and so on
sed -e "s/.*Alex.*/###########/" deleteme.txt
sed -i '/pattern to match/d' deleteme.txt
[command] > /dev/null 2>&1
myarray=(one two three)
case "${myarray[@]}" in *"two"*) echo "found" ;; esac
stty cols 80
stty rows 50
TERM=vt100; export TERM
You need to have an ~/.xscreensaver file. There needs to be at least this line:
textMode: program
textLiteral: XScreenSaver
textFile:
textProgram: fortune
textURL: http://planet.debian.org/rss20.xml
yum grouplist # Will show you list.
yum groupinstall "X Desktop Server" # Then you have to enclose it in quotes
ls /etc/cron* + cat
for user in $(cat /etc/passwd | cut -f1 -d:); do crontab -l -u $user; done
http://aa-project.sourceforge.net/
'dontLogTCPWrappersConnects' => 'true # To keep "snmpd[3458]: Connection from UDP: [127.0.0.1]:48911" from filling your logs'
yum groupinstall Xfce
yum install xorg-x11-fonts-Type1 xorg-x11-fonts-misc
/sbin/telinit 5
or startxfce4
-e will make the script stop at the first error (instead of coninuting along cheerfully)
-u will not allow unset/undefined variables (like perl strict)
-x debug mode
for srv in server{1..5}; do echo "$srv";done
server1
server2
server3
server4
server5
for srv in $(seq -w 5 10); do echo "server${srv}";done
server05
server06
server07
server08
server09
server10
len() in bash
$ var='Hello, World!'
$ echo "${#var}"
13
left() in bash
$ var='Hello, World!'
#${string:position:length}
$ echo "${var:0:5}"
Hello
right() in bash
$ var='Hello, World!'
#${string:position:length}
$ echo "${var:7:${#var}}"
World!
#or a litte more dynamic.. (the 6 most right chars)
echo "${var:$((${#var}-6)):${#var}}"
World!
mid() in bash
$ var='Hello, World!'
#${string:position:length}
$ echo "${var:4:4}"
o, W
string replace first in bash (substitute)
$ var='Hello, World!'
#${string/substring/replacement}
$ echo "${var/o/a}"
Hella, World!
string replace all in bash (substitute)
$ var='Hello, World!'
#${string//substring/replacement}
$ echo "${var//o/a}"
Hella, Warld!
string to lower case
$ string="A FEW WORDS"
$ echo "${string,}"
a FEW WORDS
$ echo "${string,,}"
a few words
$ echo "${string,,[AEIUO]}"
a FeW WoRDS
string to upper case (all caps)
$ string="a few words"
$ echo "${string^}"
A few words
$ echo "${string^^}"
A FEW WORDS
$ echo "${string^^[aeiou]}"
A fEw wOrds
Extract string replacing "cut" or "awk" fields
$ STRING="username:homedir:shell"
$ echo "$STRING"|cut -d ":" -f 3
shell
$ echo "${STRING##*:}"
shell
Extract the value after equal character
$ VAR="myClassName = helloClass"
$ echo ${VAR##*= }
helloClass
Extract text in round brackets:
$ VAR="Hello my friend (enemy)"
$ TEMP="${VAR##*\(}"
$ echo "${TEMP%\)}"
enemy
Set a default value if variable undefined or null
$ echo ${VAR:-sushi}
sushi
$ VAR=sashimi; echo ${VAR:-sushi}
sashimi
Give an error if a variable is undefined or null
$ echo ${VAR:?wtf null error}
bash: VAR: wtf null error
$ VAR=cookie;echo ${VAR:?wtf null error}
cookie
Remove a prefix or a suffix of a pattern FROM a variable
$ x=shark.png
$ echo ${x%.png}
shark
$ echo ${x#shark}
.png
Find in a variable if it contains substring
mystring='My string';
if [[ $mystring =~ "My" ]]; then
echo "It's there!"
fi
tail -f application.log | grep -i -P "(error|warning|failure)"
rsync -WavP --human-readable --progress
Generate various SSL certificate, when a CSR has to be sent to a signing authority, which also creates new key:
Regular cert:
foo=example.com; \
openssl req -new -newkey rsa:2048 -nodes -out "$foo.csr" \
-keyout "$foo.key" \
-subj "/C=US/ST=YourState/L=Your City/O=Company Name/CN=$foo"
Wildcard cert:
foo=example.com; \
openssl req -new -newkey rsa:2048 -nodes -out "star_$foo.csr" \
-keyout "star_$foo.key" \
-subj "/C=US/ST=YourState/L=Your City/O=Company Name/CN=*.$foo"
openssl genrsa 2048 > host.key
openssl req -new -x509 -nodes -sha256 -days 3650 -key host.key > host.cert
#[enter *.domain.com for the Common Name]
openssl x509 -noout -fingerprint -text < host.cert > host.info
cat host.cert host.key > host.pem
chmod 400 host.key host.pem
openssl rsa -in private.key -text -noout
The top line of the output will display the key size.
Private-Key: (2048 bit)
To view the key size from a certificate:
openssl x509 -in public.pem -text -noout | grep "RSA Public Key"
RSA Public Key: (2048 bit)
openssl x509 -noout -modulus -in /etc/ssl/certs/example.crt | openssl md5;\
openssl rsa -noout -modulus -in /etc/ssl/private/example.key | openssl md5
openssl x509 -noout -in server_file.crt -enddate
echo | openssl s_client -servername https_hostname.com -connect hostname.com:443 2>/dev/null | openssl x509 -noout -enddate
echo | openssl s_client -connect pop3_hostname.com:995 2>/dev/null | openssl x509 -noout -enddate
echo | openssl s_client -connect imap_hostname.com:993 2>/dev/null | openssl x509 -noout -enddate
echo | openssl s_client -connect smtps_hostname.com:465 2>/dev/null | openssl x509 -noout -enddate
echo | openssl s_client -starttls smtp -connect starttls_hostname.com:587 2>/dev/null | openssl x509 -noout -enddate
for foo in $(seq 0 2); do
color="[3"
color+=$(($foo + 1))
color+="m"
echo -e "\e$color-------\nprod-sessions-slave$foo\n"
ssh root@prod-sessions-slave$foo.shrm.org 'df -h'
done
echo -e "\e[0m"
for foo in server-cache-{00..12..2}; do
echo $foo
done
rsync -auvz source/ dest/
rsync -auvz source/ dest
rsync -auvz source/* dest
will take all the files and subdirectories from source/ and put it into dest/
function displaytime {
local T=$1
local D=$((T/60/60/24))
local H=$((T/60/60%24))
local M=$((T/60%60))
local S=$((T%60))
(( $D > 0 )) && printf '%d days ' $D
(( $H > 0 )) && printf '%d hours ' $H
(( $M > 0 )) && printf '%d minutes ' $M
(( $D > 0 || $H > 0 || $M > 0 )) && printf 'and '
printf '%d seconds\n' $S
}
If you happen to have the servernames listed in a textfile, or if you need to prefix your ssh connections with a username, you could create that server list file like this (with or without the username prefix, depending on how you wish to connect):
glarson@ndi-web-dev
glarson@ndi-portal-dev
glarson@cache-serv001.fishbowl.punkadyne.net
and so on. This way, the tab completion will include your username prefix. And to utilize this particular file, you would enter something like this (of course, pointing to your particular server list, not mine):
complete -W "`cat /home/glarson/serverlist.txt`" ssh
and again, be careful with the quotes and back-ticks.
#!/bin/bash
file="/home/glarson/list-o-servers.txt"
while IFS= read -r line
do
# display $line or do somthing with $line
printf '%s\n' "$line"
done <"$file"
#!/bin/bash
file="/etc/passwd"
while IFS=: read -r f1 f2 f3 f4 f5 f6 f7
do
# display fields using f1, f2,..,f7
printf 'Username: %s, Shell: %s, Home Dir: %s\n' "$f1" "$f7" "$f6"
done <"$file"
1. passwd -l $USERACCT # -u is unlock, -e expire and set user to change on login
2. chage -E 0 $USERACCT # Setting it to -1 will enable no expiration, otherwise, YYYY-MM-DD
3. usermod -e 0 $USERACCT # YYYY-MM-DD format otherwise
4. usermod -s /sbin/nologin $USERACCT # Not recommended but can be useful for other purposes
semanage port -l | grep ssh
semanage port -a -t ssh_port_t -p tcp 222
sudo dd if=/dev/zero of=/swapfile bs=1024 count=4194304
sudo mkswap /swapfile
sudo chmod 600 /swapfile
sudo swapon /swapfile
sudo echo "/swapfile none swap sw 0 0" >> /etc/fstab
sudo echo "vm.swappiness=50" >> /etc/sysctl.conf
Some links: