注：
適用於kafka叢集線上擴縮容；分割槽重新分配

cat kafka_partition_reassignment.sh

#!/bin/bash
KAFKA_BIN="/app/kafka_2.12-2.1.0/bin"
zk_server="127.0.0.1:2181"
reassignment_json_file="part.json"
reassignment_borker_ids="200,201,202,203,204,205,206,207,208" #填寫叢集brokerID，擴容的話增加ID，縮容減少ID即可。

topic_list=`$KAFKA_BIN/kafka-topics.sh --list --zookeeper $zk_server|grep -v __consumer_offsets` # kafka叢集所有topic分割槽分配

#topic_list=$1 #單個topic 分割槽分配使用該引數

for n in `echo $topic_list`
do
    cat >move.json <<EOF
{"topics": [{"topic": "$n"}],
 "version":1
}
EOF
    $KAFKA_BIN/kafka-reassign-partitions.sh --zookeeper $zk_server --topics-to-move-json-file move.json  --broker-list  "$reassignment_borker_ids"  --generate|tail -1 >$reassignment_json_file
    $KAFKA_BIN/kafka-reassign-partitions.sh --zookeeper $zk_server --reassignment-json-file $reassignment_json_file --execute
    sleep 10
    for n in `seq 10000`
    do
       process_num=`$KAFKA_BIN/kafka-reassign-partitions.sh --zookeeper $zk_server --reassignment-json-file $reassignment_json_file --verify|awk '{print $NF}'|grep "progress"|wc -l`
       if [ "$process_num" -eq 0 ];then
          break
       else
          $KAFKA_BIN/kafka-reassign-partitions.sh --zookeeper $zk_server --reassignment-json-file $reassignment_json_file --verify
          sleep 10
       fi
     done
done
echo "檢視進度:$KAFKA_BIN/kafka-reassign-partitions.sh --zookeeper $zk_server --reassignment-json-file $reassignment_json_file --verify"