Use awk to count log events by day, or by hour assuming each line starts with (date)(space)(hour)(colon)

count-log-events.sh

#!/bin/bash
# EXAMPLE 1, BY DAY (default delimiter space)
# otus.decare.com:/logs/foster/WebLogic/DDS_Prod/claimservices > ~/bin/count-log-events.sh " MapMessage received" *.out
# 2016-08-11 35306
#
# EXAMPLE 2, BY HOUR (delimiter colon)
# otus.decare.com:/logs/foster/WebLogic/DDS_Prod/claimservices > ~/bin/count-log-events.sh -F: " MapMessage received" *.out
# 2016-08-11 02 6501
# 2016-08-11 03 3529
# 2016-08-11 04 13221
# 2016-08-11 05 3344
# 2016-08-11 06 1307
# 2016-08-11 07 5941
# 2016-08-11 08 1465
# otus.decare.com:/logs/foster/WebLogic/DDS_Prod/claimservices >
USAGE="Usage: $0 [-F delim] pattern files..."
while getopts "F:" opt; do
    case $opt in 
        F ) delim="-F$OPTARG"
            shift;;
        * ) echo $USAGE
            exit 1
    esac
done

if [ $# -eq 0 ]; then
    echo $USAGE
    exit 1
fi

pattern=$1
shift
awk $delim -v pattern="$pattern" '$0 ~ pattern {counts[$1] = counts[$1] + 1; } END { for (val in counts) print val, counts[val]; }' "$@" | sort