3.1 Create a file
cd bin jump to the bin folder in
touch start-kafka-cluster.sh - a key to start a new file
touch stop-kafka-cluster.sh - a new button to stop file
3.2 write Start
Start-Kafka used to live-Cluster .sh content:
! # / bin / bash
Brokers = "hadoop01 hadoop02 hadoop03"
KAFKA_HOME = "/ Home / hadoop / kafka_2.12-2.3.0"
KAFKA_NAME = "kafka_2.12-2.3.0"
echo "INFO: to the Begin Kafka Cluster ... Start "
for $ Broker in Brokers
do
echo" the INFO: Starting KAFKA_NAME $ {ON} $ {...} Broker "
SSH $ {} -C Broker" Source / etc / Profile; $ {SH} KAFKA_HOME {} $ KAFKA_HOME -daemon /bin/kafka-server-start.sh /config/server.properties "
IF [[$ -eq 0]?]; the then
echo" the INFO: [Broker $ {}] the Start successfully "
fi
done
echo "INFO:Kafka cluster starts successfully !"
3.2 编写stop
stop-kafka-cluster.sh内容：
#!/bin/bash
brokers="hadoop01 hadoop02 hadoop03"
KAFKA_HOME="/home/hadoop/kafka_2.11-2.3.0"
KAFKA_NAME="kafka_2.11-2.3.0"
echo "INFO : Begin to stop kafka cluster ..."
for broker in $brokers
do
echo "INFO : Shut down ${KAFKA_NAME} on ${broker} ..."
ssh ${broker} "source /etc/profile;bash ${KAFKA_HOME}/bin/kafka-server-stop.sh"
if [[ $? -ne 0 ]]; then
echo "INFO : Shut down ${KAFKA_NAME} on ${broker} is down"
fi
done
注1：如 cat -v start-kafka-cluster.sh 或 cat -v start-kafka-cluster.sh met each end of the line there is "^ M", then use the cat filename | tr -d '\ r'> newfilename command
Note 2: Modify kafka-server-stop.sh (because the file itself can not really stop to stop kafka process after installation):
#PIDS = $ (PS AX | grep -i 'kafka \ .Kafka' | grep the Java | grep -v grep | awk 'Print $ {}. 1')
the PIDS = $ (PS AX | grep -i 'Kafka \ .Kafka' | awk '{}. 1 Print $' | xargs -SIGTERM the kill)
the implementation process is as follows:
[hadoop@hadoop01 ~]$ jps --查看kafka是否启动
8736 NodeManager
8593 ResourceManager
8083 DataNode
7942 NameNode
21336 Jps
8330 SecondaryNameNode
13595 Kafka
[hadoop@hadoop01 ~]$ cd bin
[hadoop@hadoop01 bin]$ sh stop-kafka-cluster.sh --启动stop脚本
INFO : Begin to stop kafka cluster ...
INFO : Shut down kafka_2.12-2.3.0 on hadoop01 ...
No kafka server to stop
INFO : Shut down kafka_2.12-2.3.0 on hadoop01 is down
INFO : Shut down kafka_2.12-2.3.0 on hadoop02 ...
No kafka server to stop
INFO : Shut down kafka_2.12-2.3.0 on hadoop02 is down
INFO : Shut down kafka_2.12-2.3.0 on hadoop03 ...
No kafka server to stop
INFO : Shut down kafka_2.12-2.3.0 on hadoop03 is down
INFO : kafka cluster shut down completed!
[hadoop@hadoop01 bin]$ jps --查看是否停止成功
8736 NodeManager
8593 ResourceManager
8083 DataNode
21428 Jps
7942 NameNode
8330 SecondaryNameNode
[hadoop@hadoop01 bin]$ sh start-kafka-cluster.sh --启动start脚本
INFO : Begin to start kafka cluster ...
INFO : Starting kafka_2.12-2.3.0 on hadoop01 ...
INFO:[hadoop01] Start successfully
INFO : Starting kafka_2.12-2.3.0 on hadoop02 ...
INFO:[hadoop02] Start successfully
INFO : Starting kafka_2.12-2.3.0 on hadoop03 ...
INFO:[hadoop03] Start successfully
INFO: Cluster Soho starts Kafka used to live successfully!
[Hadoop @ hadoop01 bin] $ JPS - to see whether the node 1 successfully started
8736 the NodeManager
8593 the ResourceManager
8083 DataNode
21923 Jps
7942 the NameNode
21847 Kafka used to live
8330 SecondaryNameNode
[hadoop @ hadoop01 bin] $ SSH hadoop02 - View node 2 launch OK
Last the Login: Wed Oct 2 09:21:59 2019 from Gateway
[hadoop @ hadoop02 ~] $ JPS
10564 the NodeManager
10888 QuorumPeerMain
15563 Jps
10431 DataNode
15439 Kafka used to live