I've written two shell scripts where script A calls script B, passing two parameters.

Script B takes about a minute to complete (loading and processing data).

When running script B once from script A, everything works ok. However, when running script B more than once from script A, things start to mess up:

e.g with Script A in the format below, the second run starts before the first run has finished.

ScriptB Param1A Param2A
ScriptB Param1B Param2B

Can anyone advise how I can introduce a check to see if script B has completed before starting it again with new parameters?

For info I'm using hp/ux 11.

I'm new to Unix and new to this forum, so thanks for your help in advance.

Use a lock file that is generated when the first instance of the script is run. The second script will look for the presence of the lock file. If it exists, the second script could either wait (for a specified period of time) or it could exit. After the first script is done, it would of course clean up (remove) the lock file.
I had created a function awhile back to create a lock file based upon the process number of the shell. The function takes two arguments, one for the name of the log file and the other is $$. The script creates a lock file in the /tmp directory. This file holds the process ID of the shell that created it. If you attempt to execute the script while another instance was running, this function will prevent the second instance from running. Works well for me. Feel free to use if you so choose.


CreateLockFile () { unset PROCESS_ID unset ID LFILE=${TEMP_PATH}/${1}_${TODAYS_DATE}.LCK ID=$2 if [ -r ${LFILE} ] then OPID=`cat ${LFILE}` 2> /dev/null if [ -z ${OPID} -eq 0 ] #Make Sure OPID contains a value then exit ${FAILURE} "ERROR-APP-->: `basename ${LFILE}` exists but contains no Process ID" | tee -a ${INLOG} else PROCESS_ID=`ps -p ${OPID} | grep ADD-SCRIPT-NAME-HERE | awk -F" " '{print $1}' 2> /dev/null` if [ ${PROCESS_ID} ] #Lock File is there, check if process is actually running then echo "WARNING-->: ${1} Script Is Currently Running [PID=${OPID}], Exiting. ${DATE_TIME}" | tee -a ${INLOG} exit ${SUCCESS} else echo "INFO-->: Old Lock File with PID= [ ${OPID} ] Exists But Process Is Not Running. " >> ${INLOG} echo "INFO-->: Overwriting Old PID with New PID Value of [ ${ID} ] " >> ${INLOG} echo "$ID" > ${LFILE} fi fi else echo "$ID" > ${LFILE} if [ $? -ne 0 ] then exit ${FAILURE} "ERROR-APP-->: Could Not Create Lock File - Exiting " | tee -a ${INLOG} fi fi }

CreateLockFile () {

unset PROCESS_ID
unset ID
LFILE=${TEMP_PATH}/${1}_${TODAYS_DATE}.LCK
ID=$2

if [ -r ${LFILE} ]
 then
    OPID=`cat ${LFILE}` 2> /dev/null
    if [ -z ${OPID} -eq 0 ] #Make Sure OPID contains a value
     then
	 exit ${FAILURE} "ERROR-APP-->: `basename ${LFILE}` exists but contains no Process ID" | tee -a ${INLOG}
     else
        PROCESS_ID=`ps -p ${OPID} | grep ADD-SCRIPT-NAME-HERE | awk -F" " '{print $1}'  2> /dev/null`

        if [ ${PROCESS_ID} ]  #Lock File is there, check if process is actually running
         then
           echo "WARNING-->: ${1} Script Is Currently Running [PID=${OPID}], Exiting. ${DATE_TIME}" | tee -a ${INLOG}
           exit ${SUCCESS}
        else
	  echo "INFO-->: Old Lock File with PID= [ ${OPID} ] Exists But Process Is Not Running. " >> ${INLOG}
	  echo "INFO-->: Overwriting Old PID with New PID Value of [ ${ID} ] " >> ${INLOG}
          echo "$ID" > ${LFILE}
        fi 
     fi
else
  echo "$ID" > ${LFILE}

    if [ $? -ne 0 ]
      then 
        exit ${FAILURE} "ERROR-APP-->: Could Not Create Lock File - Exiting " | tee -a ${INLOG}
    fi
fi
}
        

Thanks a lot Google.

That works nicely.