Check kafka lag and queue growth with nagios/nrpe and KafkaOffsetMonitor

check_kafkalag.sh

#!/bin/bash

#inspired by https://github.com/vide/nagios-scripts/blob/master/check_kafka_lag.sh

function printHelp() {

    cat >&2 <<EOF
$@
--warning|-w Warning threshold
--critical|-c Critical threshold
--zk|-z Zookeper connection string
--group|-g Set Kafka group to monitor
--jar|-j path to the KafkaOffsetMonitor assembly
EOF

    exit 2

}

# main starts here
while [ $# -gt 0 ]
do
    case "$1" in
        --warning|-w) WARNING="$2"; shift 2;;
        --critical|-c) CRITICAL="$2"; shift 2;;
        --zk|-z) ZK="$2"; shift 2;;
        --group|-g) GROUP="$2"; shift 2;;
        --jar|-j) JAR="$2"; shift 2;;
        *) printHelp "Missing parameter" ;;
    esac
done

if [ -z "${WARNING}" ];
then
    printHelp "Please specify a warning threshold"
fi

if [ -z "${CRITICAL}" ];
then
    printHelp "Please specify a critical threshold"
fi

if [ -z "${ZK}" ];
then
    printHelp "Please specify a Zookeper connect string"
fi

if [ -z "${GROUP}" ];
then
    printHelp "Please specify a Kafka group"
fi


OFFSETS=`java -cp $JAR com.quantifind.kafka.offsetapp.OffsetGetterApp --group $GROUP --zk $ZK --sumPart true --onlyOffsets true 2> /dev/null | grep -e '\([1-9]\|[0-9]0\) *$'`
MAX=$(echo "$OFFSETS"|awk '$5 > max {max=$5; maxline=$0}; END{ print maxline}')
MAX_LAG=$(echo $MAX | cut -d ' ' -f5)
MAX_TOPIC=$(echo $MAX | cut -d ' ' -f1)

if [ -z "$MAX_LAG" ];
then
    echo "UNKNOW"
    exit 3
fi

if (( MAX_LAG >= ${CRITICAL} ))
then
    echo "$MAX_TOPIC HAS CRITICAL LAG|$MAX_TOPIC=$MAX_LAG"
    echo "$OFFSETS"
    exit 2
else
    if (( MAX_LAG >= ${WARNING} ))
    then
        echo "$MAX_TOPIC HAS WORRISOME LAG|$MAX_TOPIC=$MAX_LAG"
        echo "$OFFSETS"
        exit 1
    fi
    echo "LAG IS FINE|$MAX_TOPIC=$MAX_LAG"
    echo "$OFFSETS"
    exit 0
fi

check_kafkaqueue.sh

#!/bin/bash

#inspired by https://github.com/vide/nagios-scripts/blob/master/check_kafka_lag.sh

function printHelp() {

    cat >&2 <<EOF
$@
--warning|-w Warning threshold
--critical|-c Critical threshold
--zk|-z Zookeper connection string
--group|-g Set Kafka group to monitor
--topics|-t Set Kafka topics to monitor
--jar|-j path to the KafkaOffsetMonitor assembly
--dat|-d filname for dat file
EOF

    exit 2

}

# main starts here
while [ $# -gt 0 ]
do
    case "$1" in
        --warning|-w) WARNING="$2"; shift 2;;
        --critical|-c) CRITICAL="$2"; shift 2;;
        --zk|-z) ZK="$2"; shift 2;;
        --group|-g) GROUP="$2"; shift 2;;
        --topics|-t) TOPICS="$2"; shift 2;;
        --jar|-j) JAR="$2"; shift 2;;
        --dat|-d) DAT="$2"; shift 2;;
        *) printHelp "Missing parameter" ;;
    esac
done

if [ -z "${WARNING}" ];
then
    printHelp "Please specify a warning threshold"
fi

if [ -z "${CRITICAL}" ];
then
    printHelp "Please specify a critical threshold"
fi

if [ -z "${ZK}" ];
then
    printHelp "Please specify a Zookeper connect string"
fi

if [ -z "${GROUP}" ];
then
    printHelp "Please specify a Kafka group"
fi

if [ -z "${DAT}" ];
then
    printHelp "Please specify a temp filepattern"
fi

if [ -z "${TOPICS}" ];
then
    printHelp "Please specify a list of Kafka topics"
fi

DATFILE="/tmp/${DAT}_logsize.dat"

OLD_SIZE=`cat $DATFILE`
OFFSETS=`java -cp $JAR com.quantifind.kafka.offsetapp.OffsetGetterApp --group $GROUP --zk $ZK --sumPart true --onlyOffsets true --topics $TOPICS 2> /dev/null | awk '{print $2,$4}' | sort`
if [ -z "$OLD_SIZE" ];
then
    echo "$OFFSETS"| sort > $DATFILE;
    echo "UNKNOWN -- first run";
    exit 3;
else
    JOIN=`echo "$OFFSETS" | join -1 1 -2 1 -a 1 -e 0 - $DATFILE | awk '{print $1,$2-$3}'`
    echo "$OFFSETS" | sort > $DATFILE;
    MIN=$(echo "$JOIN" | awk '$2 >= min {min=$2; minline=$0}; END{ print minline}')
    MIN_SIZE=$(echo $MIN | cut -d ' ' -f2)
    MIN_TOPIC=$(echo $MIN | cut -d ' ' -f1)

if [ -z "$MIN_SIZE" ];
 then
     echo "UNKNOW"
     exit 3
fi

 if (( MIN_SIZE <= ${CRITICAL} ))
 then
     echo "$MIN_TOPIC HASN'T BEEN GROWING!|$MIN_TOPIC=$MIN_SIZE"
     echo "$JOIN"
     exit 2
 else
     if (( MIN_SIZE <= ${WARNING} ))
     then
         echo "$MIN_TOPIC HAS WORRISOME GROWTH|$MIN_TOPIC=$MIN_SIZE"
         echo "$JOIN"
         exit 1
     fi
     echo "KAFKA IS GOOD|$MIN_TOPIC=$MIN_SIZE"
     echo "$JOIN"
     exit 0
 fi
fi