//Browsing archives (Page 6 of 13)
ls | xargs rm
Sometime there are so many files in a directory than the rm command doesn't work
On this case the best option is to use ls in conjuntion with xargs
pokes a random line from a file
Alternatively use the --get switch to return them.
Values can be found in /usr/share/zoneinfo
Latest tips by RSS
Click here to subscribe
Follow Shell-Fu on Twitter
Click here to follow
Follow Shell-Fu on identi.ca
Click here to follow
Ever been on a machine that just wouldn't respond? As soon as you're root, lower the priority of the offending process ID(s) by using the 'renice' command:
# renice -19 <PID>
You can use the 'uniq' command to handle duplicate lines in a file:
$ cat tst a a b c c d
$ uniq -d tst //show duplicated lines a c
$ uniq -u tst //show unique lines b d
$ uniq tst //remove duplicates a b c d
Use the following to see the commands you use most often based on your shell history:
history | awk '{print $2}' | sort | uniq -c | sort -rn | headIf you want to tail the errors on another terminal, just push them to a fifo:
On your other terminal:
$ mkfifo cmderror
$ mycommand 2> cmderror
On your other terminal:
$ tail -f cmderror
Longish oneliner (I actually wrote it in one line first) for giving you somewhat (mount list is never good enough) accurate sum of your file systems' totals.
df | egrep -v "(Filesystem|\/dev$|shm$|dvd|cdrom)" | awk '{totalu += $2 ; totalf += $4} END {print "Total space in devices: " (totalu/1024/1024) " GB\nFree space total: " (totalf/1024/1024) " GB"}'
ls | xargs rm
Sometime there are so many files in a directory than the rm command doesn't work
[root@server logs]# rm * bash: /bin/rm: Argument list too long
On this case the best option is to use ls in conjuntion with xargs
[root@server logs]# ls | xargs rm
cat file.txt | sort -R | tail -1
pokes a random line from a file
alias vim='mplayer rocky-theme.ogg& && vim'
Play your favourite tune while you code! :)
Play your favourite tune while you code! :)
Make a backup of existing files, afterwards copy new files from somedir:
1. Go to proddir
1. Go to proddir
ls /update-200805/ |xargs -n1 -I xxx cp xxx xxx.`date +%Y%m%d` ; cp /update-200805/* .
/usr/sbin/scutil --set ComputerName "COMPUTER NAME HERE"-or-
/usr/sbin/scutil --set LocalHostName "LOCALHOST NAME HERE"
Alternatively use the --get switch to return them.
/usr/sbin/systemsetup -settimezone "TIMEZONE HERE"
Values can be found in /usr/share/zoneinfo
/usr/sbin/systemsetup -setnetworktimeserver "TIMESERVER HERE"
I use the following to list non-system users. It should be portable though won't work on systems without the getent command.
alias lsusers='getent passwd | tr ":" " " | awk "\$3 >= $(grep UID_MIN /etc/login.defs | cut -d " " -f 2) { print \$1 }" | sort'You might want to get rid of the awk, sort, uniq and grep depending on how much info you need.
for dir in $(find /proc -maxdepth 1 -type d -name "*[0-9]");do ls -l $dir/fd/ | awk '{print $11}';done | sort | uniq | grep "^/var/"
file -N * | awk -F":" '{type[$2]++}END{ for (i in type) print type[i],i }'
The for loop has a problem with entries with spaces, whether it's file names or entries in a text file. Common solutions include changing the IFS variable to exclude spaces (so that the for loop will only use tabs and line breaks to separate entries) and piping the entries to a 'while read line; do' command. However, setting and resetting IFS each time you want to include/exclude spaces is kinda messy and the 'while read' thing means you can't make any normal use of the read command within the loop. For a once-per-line loop that doesn't present these problems, try
i=0
numberoflines=$(cat mytextfile.txt | wc -l)
while [ $i -lt $numberoflines ]; do
let i++
contentoflinenumberi="$(sed -n ${i}p mytextfile.txt)"
read -p "Do you wish to see line ${i}? (y/n) " reply
if [ "$reply" = "y" ]; then
echo "$contentoflinenumberi"
fi
done
ps -eo pid,args | awk '/sleep/ && !/awk/{ system("kill -9 "$1)}'
# awk 'BEGIN{FS=""}{for(i=1;i<=NF;i+=2){ r=r $i $(i+1)":"}}END{sub(/:$/,"",r);print r}' file
During an ftp session, get multiple files like this:
prompt
mget *.tar.gz
I guess anyone who's administered several remote boxes has had the unfortunate problem of (when not thinking straight) taking down the network card on a machine you have no physical access to. The result being that the ssh session you used to connect dies. The typical mistake is to do something like (as root):
The unfortunate result being that the first statement disconnects your session and hangs up the chain resulting in the network not coming back up. A nice way around this is to use the bash "disown" builtin command, ie:
In this case you launch a backgrounded task that is disconneced from the session (meaning the ssh session dying won't kill the process) which sleeps for 5 seconds (to give the down a chance to happen) then configures the network card as appropriate and brings it back up. As soon as this launches and is disowned, then immediately takes the network card down. If the configuration change keeps the IP address the same, you'll find that after 5 seconds your bash prompt just comes back and the session resumes.
ifconfig eth0 down; ifconfig eth0 inet 123.4.5.6; ifconfig eth0 up
The unfortunate result being that the first statement disconnects your session and hangs up the chain resulting in the network not coming back up. A nice way around this is to use the bash "disown" builtin command, ie:
(sleep 5; ifconfig eth0 inet 123.4.5.6; ifconfig eth0 up)& disown -h $! ; ifconfig eth0 down
In this case you launch a backgrounded task that is disconneced from the session (meaning the ssh session dying won't kill the process) which sleeps for 5 seconds (to give the down a chance to happen) then configures the network card as appropriate and brings it back up. As soon as this launches and is disowned, then immediately takes the network card down. If the configuration change keeps the IP address the same, you'll find that after 5 seconds your bash prompt just comes back and the session resumes.
This piece of code lists the size of every file and subdirectory of the current directory, much like du -sch ./* except the output is sorted by size, with larger files and directories at the end of the list. Useful to find where all that space goes.
du -sk ./* | sort -n | awk 'BEGIN{ pref[1]="K"; pref[2]="M"; pref[3]="G";} { total = total + $1; x = $1; y = 1; while( x > 1024 ) { x = (x + 1023)/1024; y++; } printf("%g%s\t%s\n",int(x*10)/10,pref[y],$2); } END { y = 1; while( total > 1024 ) { total = (total + 1023)/1024; y++; } printf("Total: %g%s\n",int(total*10)/10,pref[y]); }'
Grep file(s) and highlight the grep matches using less...
grep query *.php | less +/query
Add one number per line from stdin until a 0 is found, then print the result.
dc -e '[+pq]sQ0[?d0=Q+lXx]dsXx'
lynx -dump http://www.spantz.org | grep -A999 "^References$" | tail -n +3 | awk '{print $2 }'
Mass-renaming files using find and sed:
(this example will rename all .php3 files to .php)
find -name "*.php3" | sed 's/\(.*\).php3$/mv "&" "\1.php"/' | sh
(this example will rename all .php3 files to .php)