I have a script like that:
su lingcat -c PHPRC\=\/home\/lingcat\/etc\/php5\
PHP_FCGI_CHILDREN\=4\ \/usr\/bin\/php\-loop\.pl\ \/usr\/bin\/php5\-cgi\ \-b\
127\.0\.0\.1\:9006\ \>\>\/home\/lingcat\/logs\/php\.log\ 2\>\&1\ \<\/dev\/null\ \&\
echo\ \$\!\ \>\/var\/php\-nginx\/135488849520817\.php\.pid
This is working. But there is too many \ in the script, they make the code unreadable. So, I wrote a new shell script:
#!/bin/sh
case "$1" in
'start')
su biergaizi -c "PHPRC=/home/biergaizi/etc/php5 PHP_FCGI_CHILDREN=2
/usr/bin/php-loop.pl /usr/bin/php-cgi -b /var/run/virtualhost/php5-fpm-biergaizi.test.sock >>/home/biergaizi/logs/php.log 2>&1 </dev/null &
echo $! > /var/php-nginx/biergaizi.test.php.pid"
RETVAL=$?
;;
'stop')
su biergaizi -c "kill `cat /var/php-nginx/biergaizi.test.php.pid` ; sleep 1"
RETVAL=$?
;;
'restart')
$0 stop ; $0 start
RETVAL=$?
;;
*)
echo "Usage: $0 { start | stop }"
RETVAL=1
;;
esac
exit
But /var/php-nginx/biergaizi.test.php.pid is empty.
What's wrong?
The .pid file is empty, because $! gets substituted by the shell executing your script, instead of the shell executing the commands you pass through su. And as there is no recently started background command in your script, it substitutes an empty string. So, shell started by su executes simply echo > /var/php-nginx/biergaizi.test.php.pid.
To prevent that, quote your command passed to su using single quotes, instead of double quotes. It is better to do that to the "stop" command as well. Like this:
su biergaizi -c 'PHPRC=/home/biergaizi/etc/php5 PHP_FCGI_CHILDREN=2
/usr/bin/php-loop.pl /usr/bin/php-cgi -b /var/run/virtualhost/php5-fpm-biergaizi.test.sock >>/home/biergaizi/logs/php.log 2>&1 </dev/null &
echo $! > /var/php-nginx/biergaizi.test.php.pid'
And this:
su biergaizi -c 'kill `cat /var/php-nginx/biergaizi.test.php.pid` ; sleep 1'
See http://www.gnu.org/software/bash/manual/html_node/Quoting.html for details.
try this:
Escape $ from $!, before passing to su -c.
Related
I have no idea, how to execute this specific combined shell script command via ssh on a remote device.
#!/bin/bash
cmd=""
command="restart"
case "$command" in
restart)
cmd+="pkill -f fileA.py;"
cmd+="python3 -u fileA.py >> fileA.log &"
;;
*)
echo "Unknown command"
esac
cmd=$(ssh root#foobar $cmd)
The error that occurs is:
pkill -f fileA.py;python3 -u fileA.py >> fileA.log &
pkill: invalid user name: fileA.py
I know that the whole string is interpreted as one command, but that's not what I want to achive.
I appreciate any help.
Are you sure you pasted the right code? That pkill error messages comes from using the -u or
-U options. Please check that first.
Next, your lack of quotes is causing you problems. After the shell substitutes the variables, you have this:
cmd=$(ssh root#foobar pkill -f fileA.py;python3 -u fileA.py >> fileA.log &)
So you're killing the process on the remote system, and launching it on the local system.
I think you really need this:
case "$command" in
restart)
cmd="pkill -f fileA.py; nohup python3 -u fileA.py >> fileA.log & disown"
;;
*)
echo "Unknown command"
esac
cmd=$(ssh root#foobar bash -c "$cmd")
nohup and disown allow the backgrounded process to keep running after the shell exits.
I want to run the bash script, StartSomething.sh, as a specific user. I use runuser command for that. Also I want to know an exit code from this bash script. So I write an exit code to the file when the command is finished or interrupted. Here is the code:
runuser myuser -s /bin/bash -c "./StartSomething.sh --pidfile=${pidfile}; \
echo $? > ${statusfile};" &
sleep 5
pid=$(cat ${pidfile})
while ps -p ${pid} > /dev/null; do sleep 1; done
end=$(cat ${statusfile})
echo "End code: ${end}"
exit ${end}
Problem is that exit code is still 0, though bash script is interrupted. What can be wrong?
If I have separate file, start.sh, with this code:
./StartSomething.sh --pidfile=${pidfile}
echo $? > ${statusfile}
and runuser command look like this:
runuser myuser -s /bin/bash -c "./start.sh" &
everything is working fine. I want to use first example without separate file. Can someone tell me what can be wrong? Is there better solution for this problem?
If all you want to do is to run the program in the background, and wait for it to finish, I think you could also use wait to get the return value (runuser passes it through, unless something exceptional happens):
runuser myuser ./StartSomething.sh --pidfile=${pidfile} &
pid=$!
# do something else
wait $!
echo "it returned $?"
or
runuser myuser ./StartSomething.sh --pidfile=${pidfile} &
pid=$!
echo -n "waiting"
while kill -0 $pid 2>/dev/null; do
echo -n "."
sleep 1
done
echo
wait $!
echo "it returned $?"
There is problem with escaping special character $. Correct command:
runuser myuser -s /bin/bash -c "./StartSomething.sh --pidfile=${pidfile}; \
echo \$? > ${statusfile};" &
Replace $? with \$?.
I have a deployment script which works fine. At the end of it I added this block:
cat << EOF > /etc/init.d/uwsgi
#!/bin/bash
daemon=$APPVENV/bin/uwsgi
args="--emperor $APPCONF/uwsgi/app.ini --daemonize /var/log/emperor.log --emperor-pidfile $APPDIR/emperor.pid --gid `id -g $APPUSER`"
pid=$APPDIR/emperor.pid
case "$1" in
start)
echo "Starting uwsgi"
start-stop-daemon -p $pid --start --exec $daemon -- $args
;;
stop)
echo "Stopping script uwsgi"
start-stop-daemon --signal INT -p $pid --stop $daemon -- $args
;;
reload)
echo "Reloading conf"
kill -HUP $(cat $pid)
;;
*)
echo "Usage: /etc/init.d/uwsgi {start|stop|reload}"
exit 1
;;
esac
exit 0
EOF
Now when I run my deployment script it gets stuck here, and I just see a blinking cursor and it doesn't run any lines after it.
Have I done anything wrong with my formatting, as I know creating a file with cat was something very picky about how it was formatted, tabulated etc.
variables are still substituted in heredocs, also subshelling with $( ), which you do in your script here:
kill -HUP $(cat $pid)
your installer attempts to cat that file, with $pid probably empty, therefore it waits for input from standard input.
You want to escape those "$" to prevent expansion, like
\$(cat \$foo)
And of course with all those variables you don't want to get expanded during installation too.
I have a process that fails regularly & sometimes starts duplicate instances..
When I run:
ps x |grep -v grep |grep -c "processname"
I will get:
2
This is normal as the process runs with a recovery process..
If I get
0
I will want to start the process
if I get:
4
I will want to stop & restart the process
What I need is a way of taking the result of ps x |grep -v grep |grep -c "processname"
Then setup a simple 3 option function
ps x |grep -v grep |grep -c "processname"
if answer = 0 (start process & write NOK & Time to log /var/processlog/check)
if answer = 2 (Do nothing & write OK & time to log /var/processlog/check)
if answer = 4 (stot & restart the process & write NOK & Time to log /var/processlog/check)
The process is stopped with
killall -9 process
The process is started with
process -b -c /usr/local/etc
My main problem is finding a way to act on the result of ps x |grep -v grep |grep -c "processname".
Ideally, I would like to make the result of that grep a variable within the script with something like this:
process=$(ps x |grep -v grep |grep -c "processname")
If possible.
Programs to monitor if a process on a system is running.
Script is stored in crontab and runs once every minute.
This works with if process is not running or process is running multiple times:
#! /bin/bash
case "$(pidof amadeus.x86 | wc -w)" in
0) echo "Restarting Amadeus: $(date)" >> /var/log/amadeus.txt
/etc/amadeus/amadeus.x86 &
;;
1) # all ok
;;
*) echo "Removed double Amadeus: $(date)" >> /var/log/amadeus.txt
kill $(pidof amadeus.x86 | awk '{print $1}')
;;
esac
0 If process is not found, restart it.
1 If process is found, all ok.
* If process running 2 or more, kill the last.
A simpler version. This just test if process is running, and if not restart it.
It just tests the exit flag $? from the pidof program. It will be 0 of process is running and 1 if not.
#!/bin/bash
pidof amadeus.x86 >/dev/null
if [[ $? -ne 0 ]] ; then
echo "Restarting Amadeus: $(date)" >> /var/log/amadeus.txt
/etc/amadeus/amadeus.x86 &
fi
And at last, a one liner
pidof amadeus.x86 >/dev/null ; [[ $? -ne 0 ]] && echo "Restarting Amadeus: $(date)" >> /var/log/amadeus.txt && /etc/amadeus/amadeus.x86 &
This can then be used in crontab to run every minute like this:
* * * * * pidof amadeus.x86 >/dev/null ; [[ $? -ne 0 ]] && echo "Restarting Amadeus: $(date)" >> /var/log/amadeus.txt && /etc/amadeus/amadeus.x86 &
cccam oscam
I adopted the #Jotne solution and works perfectly! For example for mongodb server in my NAS
#! /bin/bash
case "$(pidof mongod | wc -w)" in
0) echo "Restarting mongod:"
mongod --config mongodb.conf
;;
1) echo "mongod already running"
;;
esac
I have adopted your script for my situation Jotne.
#! /bin/bash
logfile="/var/oscamlog/oscam1check.log"
case "$(pidof oscam1 | wc -w)" in
0) echo "oscam1 not running, restarting oscam1: $(date)" >> $logfile
/usr/local/bin/oscam1 -b -c /usr/local/etc/oscam1 -t /usr/local/tmp.oscam1 &
;;
2) echo "oscam1 running, all OK: $(date)" >> $logfile
;;
*) echo "multiple instances of oscam1 running. Stopping & restarting oscam1: $(date)" >> $logfile
kill $(pidof oscam1 | awk '{print $1}')
;;
esac
While I was testing, I ran into a problem..
I started 3 extra process's of oscam1 with this line:
/usr/local/bin/oscam1 -b -c /usr/local/etc/oscam1 -t /usr/local/tmp.oscam1
which left me with 8 process for oscam1. the problem is this..
When I run the script, It only kills 2 process's at a time, so I would have to run it 3 times to get it down to 2 process..
Other than killall -9 oscam1 followed by /usr/local/bin/oscam1 -b -c /usr/local/etc/oscam1 -t /usr/local/tmp.oscam1, in *)is there any better way to killall apart from the original process? So there would be zero downtime?
If you changed awk '{print $1}' to '{ $1=""; print $0}' you will get all processes except for the first as a result. It will start with the field separator (a space generally) but I don't recall killall caring. So:
#! /bin/bash
logfile="/var/oscamlog/oscam1check.log"
case "$(pidof oscam1 | wc -w)" in
0) echo "oscam1 not running, restarting oscam1: $(date)" >> $logfile
/usr/local/bin/oscam1 -b -c /usr/local/etc/oscam1 -t /usr/local/tmp.oscam1 &
;;
2) echo "oscam1 running, all OK: $(date)" >> $logfile
;;
*) echo "multiple instances of oscam1 running. Stopping & restarting oscam1: $(date)" >> $logfile
kill $(pidof oscam1 | awk '{ $1=""; print $0}')
;;
esac
It is worth noting that the pidof route seems to work fine for commands that have no spaces, but you would probably want to go back to a ps-based string if you were looking for, say, a python script named myscript that showed up under ps like
root 22415 54.0 0.4 89116 79076 pts/1 S 16:40 0:00 /usr/bin/python /usr/bin/myscript
Just an FYI
The 'pidof' command will not display pids of shell/perl/python scripts. So to find the process id’s of my Perl script I had to use the -x option i.e. 'pidof -x perlscriptname'
I cannot get case to work at all.
Heres what I have:
#! /bin/bash
logfile="/home/name/public_html/cgi-bin/check.log"
case "$(pidof -x script.pl | wc -w)" in
0) echo "script not running, Restarting script: $(date)" >> $logfile
# ./restart-script.sh
;;
1) echo "script Running: $(date)" >> $logfile
;;
*) echo "Removed duplicate instances of script: $(date)" >> $logfile
# kill $(pidof -x ./script.pl | awk '{ $1=""; print $0}')
;;
esac
rem the case action commands for now just to test the script. the above pidof -x command is returning '1', the case statement is returning the results for '0'.
Anyone have any idea where I'm going wrong?
Solved it by adding the following to my BIN/BASH Script:
PATH=$PATH:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin
In case you're looking for a more modern way to check to see if a service is running (this will not work for just any old process), then systemctl might be what you're looking for.
Here's the basic command:
systemctl show --property=ActiveState your_service_here
Which will yield very simple output (one of the following two lines will appear depending on whether the service is running or not running):
ActiveState=active
ActiveState=inactive
And if you'd like to know all of the properties you can get:
systemctl show --all your_service_here
If you prefer that alphabetized:
systemctl show --all your_service_here | sort
And the full code to act on it:
service=$1
result=`systemctl show --property=ActiveState $service`
if [[ "$result" == 'ActiveState=active' ]]; then
echo "$service is running" # Do something here
else
echo "$service is not running" # Do something else here
fi
If you are using CentOS, no need to write a script and set cron job. Here is one of the smartest ways to ensure systemd services restart on failure.
Make following changes to /usr/lib/systemd/system/mariadb.service
Then under the [Service] section in the file, add the following 2 lines:
Restart=always
RestartSec=3
After saving the file we need to reload the daemon configurations to ensure systemd is aware of the new file
systemctl daemon-reload
Read the following link for the complete steps -
https://jonarcher.info/2015/08/ensure-systemd-services-restart-on-failure/
I'm trying to call a custom shell script through sh:
/bin/sh -c 'myscript.sh` >log.txt 2>&1 & echo $!
Output of this command is a PID of a created background process. I want to instruct /bin/sh to save return code of myscript.sh to some file. Is it possible?
echo $? >> /path/to/return_code
$? has the return code of the last statement in bash.
(/bin/sh -c "myscript.sh" >log.txt 2>&1 ; echo $? >somefile) & echo $!
(
/bin/sh -c 'myscript.sh` >log.txt 2>&1
echo $? > some_file
) &