How to tell if script is still running

[littleIdiot]

Hi,

I have a script that, no matter how many catches and fail-safe's I put itno it, will sometimes hang due to the data I am dealing with. I have to Ctrl+C out of it when it hangs.

I have tried many ways of catching this before it happens but it is not possible in some instances.

I am looking for a way that will allow me to set this script off at night, let it run through the night, and if it fails or hangs, will be re-set off with the next input file.

Something like:

sh my_script input_file.1

if [script is still running]
 then
     sleep 600
 else
    sh my_script input_file.2
fi

This would need a list of input files to be worked on of course.


Any ideas, pointers in the right direction, or out and out solutions appreciated!

Cheers,
lil'

 

[rzs0502]

How about

<code>
#!/usr/bin/ksh

while [ `ps -ef | grep [p]rocess1 | wc -l -ne 0 ] ; do
sleep 600
done
sh process2.sh &
exit 0
</code>




&quot;If you always do what you've always done, you will always be where you've always been.&quot;

 

[mrn]

Couldn't you check and amend the data to make it conform?

Mike

"A foolproof method for sculpting an elephant: first, get a huge block of marble, then you chip away everything that doesn't look like an elephant."

http://www.airport-parking-site.co.uk/

 

[Ogzilal]

Hi,

Assume you have candidate files in /some/dir and your script exits with 0 at normal termination, otherwise non zero value.

cd /some/dir
for IN_FILE in $(ls)
do
   echo "input file : $INFILE"
   date
  /path/to/your_script $IN_FILE
  if [ $? -eq 0 ]
  then
      echo "$IN_FILE : OK
   else
      echo "$IN_FILE : KO
   fi
   date
done

 

[littleIdiot]

thanks for the replies - but none of them are a solution I'm afraid:

rzs0502
'ps -ef' is not possible as I may need to run two or more of these processes at once, but I am tryig not to. If that is the only way then so be it.

mrn
check and ammend the data ... I am dealing with 7TB of data, in several thousand files ... so no.

aau
I can't see how, if when my script hangs, this option will affect anything other than print out "[filename] : KO"
I need it to determine if the script has stopped or hung, not just see if it has exited.

My script works through a list of files - most files process cleanly, others fail/hang/bomb out etc. I was thinking along the lines of touching a 'flag file' every time my script itterates, and having a second script that works out the last time the flag file was touched - if it is over ... say 10 minutes, then set off the next batch of files??

Does that makes sense - something like that anyway ...

 

[mrn]

Could you do it with time, if it runs for over x minutes kill it?

or

Something like

#----------------------------------------------------------------
# Launch command in background and collect return code on completion
#----------------------------------------------------------------
( command < $commands > $command_log 2>&1 ; \
  echo $? > $command_return_code ) &
job_1_number=$!
tail -f $command_log | grep $search > $done_log &
job_2_number=$!
#-------------------------------
# Check done_log until complete
#-------------------------------
while [ `grep -c $search $done_log` = 0 ]
do
    sleep 1
done
#-----------------------------------------
# Tidy Up the jobs if hung and log result
#-----------------------------------------
if [ "`cat $command_return_code`" = "" ]
then
    kill -9 $job_1_number $job_2_number
    echo "command Complete at `s_date`"		   >> $process_log
else
    echo "command returned a [`cat $command_return_code`]" >> $process_log
    echo "Aborting $0 at `s_timestamp`"
    echo "========================================="
    cat $process_log
    echo "========================================="
    exit
fi

Mike

"A foolproof method for sculpting an elephant: first, get a huge block of marble, then you chip away everything that doesn't look like an elephant."

http://www.airport-parking-site.co.uk/