Help in creating script

Question

I am currently working on script with the following goals:

1. Script will check every 15-minute if there is AppA.log4 in /logs directory. There are numerous apps in /logs directory, for example AppA.log4, AppB.log4, AppC.log4
2. If there is AppA.log4, it will be renamed to AppA.log.<timestamp>
3. zip the renamed file and move to /backupDIR

So far, I have created working script however it is hardcoded.

Can you help improving this to be more dynamic? For example, AppA.log4 and AppB.log4 will be moved and zipped, leaving AppC.log4 in the /logs directory.

What if there's a separate file that lists all the App (AppA and AppB) that will be read by this script. How can I apply it here. Thank you in advance.

TIMESTAMP=`date "+%Y.%m.%d-%H.%M"`
LOGS="AppA.log.4"
APPLOGS_DIR="/logs/"
BACKUP_DIR="/backupDIR/"
cd $APPLOGS_DIR;
if [ -f "$LOGS" ];
then
        mv $LOGS $BACKUP_DIR;
                cd $BACKUP_DIR;
                mv $LOGS AppA_$TIMESTAMP;
        gzip AppA_$TIMESTAMP;
else
    echo "No log file to be backed up"
fi

Answer 1

you can use for loop and use string manipulation to generate output file name ${var%.*} based on input file name

TIMESTAMP=$(date "+%Y.%m.%d-%H.%M")
APPLOGS_DIR="./logs"
BACKUP_DIR="./backupDIR"
canonicalize relative paths
APPLOGS_DIR=$(realpath "$APPLOGS_DIR")
BACKUP_DIR=$(realpath "$BACKUP_DIR")
delete this line - just for demo
echo "AppA.log.4\nAppB.log.4" >> ./index.txt
read file names from list
LOGS=$(sort ./index.txt | uniq) || exit 127
cd "$APPLOGS_DIR" || exit 127
for log in $LOGS
  do
    if [ -f "$log" ]
      then
        mv -f $log "$BACKUP_DIR"
        cd "$BACKUP_DIR" || exit 127
        mv $log ${log%.}_${TIMESTAMP}
        gzip ${log%.}_${TIMESTAMP}
        cd "$APPLOGS_DIR"
      else
        echo "No log file to be backed up"
    fi
done

However, just archive & remove file can be done without cd

while IFS='' read -r log || [ "$log" ]
  do
    [ -f "$APPLOGS_DIR/$log" ] && \
    gzip -nc "$APPLOGS_DIR/$log" > "$BACKUP_DIR/${log%.*}_${TIMESTAMP}.${log##*.}.gz" && \
    rm "$APPLOGS_DIR/$log" || \
    echo "No '$log' file to be backed up"
done <<< $(sort ./index.txt | uniq)

or as one-liner (when output file names doesn't matter and file names does not contain whitespaces) use the -S flag for suffix and process the index.txt straight from gzip without any loop

cd "$APPLOGS_DIR" && \
gzip -S .${TIMESTAMP}.gz $(cat $OLDPWD/index.txt) && \
mv -f *.${TIMESTAMP}.gz "$BACKUP_DIR"

Answer 2

I modified what you provided to add a for loop where you can specify the various log prefixes. This assumes you are calling this script every 15 minutes from a job scheduler such as cron or autosys, if not then it can be wrapped in a while true loop with a 15 minute sleep.

TIMESTAMP=`date "+%Y.%m.%d-%H.%M"`
LOGS="AppA AppB"
# to pull log prefixes from a file replace with the next line
#LOGS="$(cat prefixes.txt)"
LOGSUFFIX=".log.4"
APPLOGS_DIR="/logs"
BACKUP_DIR="/backupDIR"
for LOG in ${LOGS}
do
    if [ -f "${APPLOGS_DIR}/${LOG}${LOGSUFFIX}" ]
    then
        mv "${APPLOGS_DIR}/${LOG}${LOGSUFFIX}" "${BACKUP_DIR}/${LOG}${TIMESTAMP}"
        gzip "${BACKUP_DIR}/${LOG}${TIMESTAMP}"
    else
        echo "No ${LOG}${LOGSUFFIX} file to be backed up"
    fi
done