Linux Tricks
From ben.goodacre.name/tech
Search for packages that contain a certain file
apt-file search filename
Stopwatch on your linux machine: time cat to srart, Ctrl+D to stop
Show startup items:chkconfig --list
Contents |
File searching/manipulation
Copy all files inc subdirs:cp -a /usr/local/foo/* /var/temp/barDelete all files inc subdirs:
rm -rf folderSCP a folder including sub-dirs:
scp -r folder/ 1.1.1.1:
Find files with indexer:
updatedb locate filename
Find files without indexer:
find /dir -name filename find /dir -name '*part of file name*'Delete everything older than 7 days:
find /directoryname -type f -mtime +7 -exec rm {} \;
Search text within files and print the lines:
find /dir -type f -exec grep "textinfile" {} \;
Search text within files and print only the filenames:
find /dir -type f | xargs grep -li "textinfile"
Search and replace over multiple files:
perl -pi -w -e 's/old/new/g;' *.php
Show files accessed this year:
ls -Rlua /dir|grep -v '\.$'|grep `date "+%Y"`
Du for each folder without showing subdirs:
for i in `find . -maxdepth 1 -type d`; do du -sh $i; done;
Remove duplicate files [1] The script below will find duplicate files (files with the same md5sum) in a specified directory and output a new shell script containing commented-out rm statements for deleting them. You can then edit this output to decide which to keep.
OUTF=rem-duplicates.sh;
echo "#! /bin/sh" > $OUTF;
find "$@" -type f -print0 |
xargs -0 -n1 md5sum |
sort --key=1,32 | uniq -w 32 -d --all-repeated=separate |
sed -r 's/^[0-9a-f]*( )*//;s/([^a-zA-Z0-9./_-])/\\\1/g;s/(.+)/#rm \1/' >> $OUTF;
chmod a+x $OUTF; ls -l $OUTF
Networking
Show listening ports and the processes using the ports:
sudo netstat -ltnup
Ascertain what line in the rouitng table a particular destination ip uses:
/sbin/ip ro get 1.1.1.1View routing table:
route -nAdd static route:
sudo ip route add networkaddress/cidr via next_hop_ip
Investigating a system / Troubleshooting
See also Linux System Troubleshooting
Show processes sorted by memory usage descending:
ps -e -orss=,args= | sort -b -k1,1n | pr -TW$COLUMNS
Text Manipulation
Strip out a single character from text:
;Strip colons from a MAC address, tr cannot be used when stripping a phrase echo 00:00:00:00:00:00 | tr -d ':' 000000000000
Strip out the phrase 'remove me' from text:
cat file|sed '/s/remove me//g'
Delete the first character of every line:
cat file|sed 's/^.//'
Find words in garbled text [2]
echo "Garbled Text" | grep -o -F -f /usr/share/dict/words | sed -e "/^.$/d"
Get columns 1 and 3 and seperate with commas:
cat file | awk -v OFS=',' '{print $1, $3}'
BASH
See Also BASH Tricks
Setup a portable prompt with user@host: pwd on systems that have an old prompt by default, like some BSD machines
PS1='\u@\h:\w\$ '
Add to bashrc for a scpnewestfile user@host:/dir command to automatically scp the most recen tly modified file to another host:
scpnewestfile()
{
scp `ls -ltr|tail -1|tr -s ' '|cut -d ' ' -f8` $1
}
Archiving
Create tar archive
tar -cvzf files.tar.gz file1 file2 file3 filen
Create bz2 archive of dir/
tar -c dir/ | bzip2 > dir.tar.bz2
Extract archive
bzip2 -dc dir.tar.bz2|tar -x
Copy dir/ with permissions to remote machine
( tar -c /dir/to/copy ) | ssh -C user@remote 'cd /where/to/ && tar -x -p'
Less
See Also: Less Tricks
/normal search text here/i
Redirecting Output
Redirect standard output and standard error to a file:
command >file 2>&1
Pipe standard output and standard error: tee in this example
script 2>&1 | tee file
SSH
SSH Auto complete:
SSH_COMPLETE=( $(cat ~/.ssh/known_hosts | \
cut -f 1 -d ' ' | \
sed -e s/,.*//g | \
uniq | \
egrep -v [0123456789]) )
complete -o default -W "${SSH_COMPLETE[*]}" ssh
