1、zookeeper搭建
Kafka集群依赖zookeeper,需要提前搭建好zookeeper
cd /usr/local/software
jdk-8u161-linux-x64.rpm
链接:https://pan.baidu.com/s/1i6iHIDJ 密码:bgcc
rpm -ivh jdk-8u161-linux-x64.rpm
vi /etc/profile
JAVA_HOME=/usr/java/jdk1.8.0_161
JRE_HOME=/usr/java/jdk1.8.0_161/jre
PATH=$PATH:$JAVA_HOME/bin:$JRE_HOME/bin
CLASSPATH=.:$JAVA_HOME/lib/dt.jar:$JAVA_HOME/lib/tools.jar:$JRE_HOME/lib
export JAVA_HOME JRE_HOME PATH CLASSPATH
source /etc/profile
echo $PATH
Step2:
# 下载zookeeper
#
tar -zxvf zookeeper-3.4.11.tar.gz
移动一下
mv zookeeper-3.4.11 /usr/local/zookeeper-3.4.11
Step3:重命名 zoo_sample.cfg文件
mv /usr/local/zookeeper-3.4.11/conf/zoo_sample.cfg /usr/local/zookeeper-3.4.11/conf/zoo.cfg
Step4:vi /usr/local/zookeeper-3.4.11/conf/zoo.cfg,修改
dataDir=/usr/local/zookeeper-3.4.11/data
Step5:创建数据目录
mkdir /usr/local/zookeeper-3.4.11/data
Step6:启动zookeeper:执行
/usr/local/zookeeper-3.4.11/bin/zkServer.sh start
Step7:检测是否成功启动:执行
/usr/local/zookeeper-3.4.11/bin/zkCli.sh
或者
yum install nc -y
echo stat| nc localhost 2181
================================================================================================================
2、下载Kafka
# mkdir -p /usr/local/software
# cd /usr/local/software
# wget http://mirror.bit.edu.cn/apache/kafka/1.0.0/kafka_2.12-1.0.0.tgz
# 百度云下载地址:
链接:https://pan.baidu.com/s/1Kp0uD_5YjGKOLkbW_igm2g 密码:v1q7
kafka_2.12-1.0.0.tgz //其中2.12-1.0.0为Scala的版本, kafka-1.0.0-src.tgz为kafka版本
3、解压
# tar zxf kafka_2.12-1.0.0.tgz -C /usr/local/
# cd /usr/local/ # mv kafka_2.12-1.0.0/ kafka/
4、配置
mkdir -p /usr/local/kafka/kafkaLogs
# vi /usr/local/kafka/config/server.properties
# broker的ID,集群中每个broker ID不可相同
broker.id=0
# 监听器,端口号和port一致即可
listeners=PLAINTEXT:/10.10.6.225/:9092
# Broker的监听端口
port=9092
# 必须填写当前服务器IP地址
host.name=10.10.6.225
# 必须填写当前服务器IP地址
advertised.host.name=10.10.6.225
# 暂未配置集群
zookeeper.connect=10.10.6.225:2181
# 消息持久化目录
log.dirs=/usr/local/kafka/kafkaLogs
# 可以删除主题
delete.topic.enable=true
# 关闭自动创建topic
auto.create.topics.enable=false
5、配置Kafka的环境变量
# vi /etc/profile
export KAFKA_HOME=/usr/local/kafka
export PATH=$PATH:$KAFKA_HOME/bin
# source /etc/profile
# vi /etc/hosts
# es为主机名 ,这里一定要注意,是主机名!!!!重要的话说三次!!!!!!!!
127.0.0.1 es
10.10.6.225 es
6、启动与停止Kafka
# kafka-server-start.sh -daemon $KAFKA_HOME/config/server.properties
官方推荐启动方式:
# /usr/local/kafka/bin/kafka-server-start.sh /usr/local/kafka/config/server.properties &
但这种方式退出shell后会自动断开
停止:
kafka-server-stop.sh
7、验证
# jps
2608 Kafka
2236 QuorumPeerMain
2687 Jps
看到Kafka的进程,说明Kafka已经启动
8、创建topic
创建名为test,partitions为3,replication为3的topic
# kafka-topics.sh --create --zookeeper 10.10.6.225:2181 --partitions 1 --replication-factor 1 --topic test
查看topic状态
# kafka-topics.sh --describe --zookeeper 10.10.6.225:2181 --topic test
Topic:test PartitionCount:1 ReplicationFactor:1 Configs:
Topic: test Partition: 0 Leader: 0 Replicas: 0 Isr: 0
删除topic
执行如下命令
# kafka-topics.sh --delete --zookeeper 10.10.6.225:2181 --topic test
9、测试使用Kafka
发送消息
# kafka-console-producer.sh --broker-list 10.10.6.225:9092 --topic test 输入以下信息: This is a message This is another message
接收消息
# kafka-console-consumer.sh --bootstrap-server 10.10.6.225:9092 --topic test --from-beginning
若看到上输入的信息说明已经搭建成功。
更复杂配置参考:
黄海添加于2018-02-11 夜
链接:https://pan.baidu.com/s/1i6HnIzr 密码:1soq
KafkaProducer.py
# http://kafka-python.readthedocs.io/en/master/
# 安装办法:
# C:\Users\Administrator>pip install kafka-python
# Collecting kafka-python
# Downloading kafka_python-1.4.1-py2.py3-none-any.whl (235kB)
# 100% |████████████████████████████████| 235kB 150kB/s
# Installing collected packages: kafka-python
# Successfully installed kafka-python-1.4.1
# http://blog.csdn.net/evankaka/article/details/52421314
from kafka import KafkaProducer
from Util.MySQLHelper import *
import json
producer = KafkaProducer(bootstrap_servers='10.10.6.225:9092', value_serializer=lambda v: json.dumps(v).encode('utf-8'))
db = MySQLHelper()
sql = "select ID,RESOURCE_ID_INT,RESOURCE_ID_CHAR,RESOURCE_TITLE,RESOURCE_TYPE_NAME,RESOURCE_FORMAT,RESOURCE_PAGE,CAST(CREATE_TIME AS CHAR) AS CREATE_TIME,DOWN_COUNT,FILE_ID,RESOURCE_TYPE,STRUCTURE_ID,PERSON_ID,PERSON_NAME,IDENTITY_ID from t_resource_info limit 100"
dt = db.query(sql)
print(len(dt))
for row in dt:
producer.send('t_resource_info', row)
producer.flush()
print('恭喜,完成!')
不依赖于MYSQL的数据提交:
import json
from kafka import KafkaProducer
import datetime
# kafka的服务器位置
kafka_servers = '10.10.6.194:9092'
# 日期的转换器
class DateEncoder(json.JSONEncoder):
def default(self, obj):
if isinstance(obj, datetime.datetime):
return obj.strftime('%Y-%m-%d %H:%M:%S')
elif isinstance(obj, datetime.date):
return obj.strftime("%Y-%m-%d")
else:
return json.JSONEncoder.default(self, obj)
# 黄海定义的输出信息的办法,带当前时间
def logInfo(msg):
i = datetime.datetime.now()
print(" %s %s" % (i, msg))
# 统一的topic名称
topicName = 'test'
dt=[{"id":1,"name":"刘备"},{"id":2,"name":"关羽"},{"id":3,"name":"张飞"}]
# kafka的生产者
producer = KafkaProducer(bootstrap_servers=kafka_servers)
# # 将字段大写转为小写
for row in dt:
new_dics = {}
for k, v in row.items():
new_dics[k.lower()] = v
jstr = json.dumps(new_dics, cls=DateEncoder)
producer.send(topic=topicName, partition=0, value=jstr.encode('utf-8'))
# 提交一下
producer.flush()
print('恭喜,完成!')
KafkaConsumer.py
from kafka import KafkaConsumer
import time
def log(str):
t = time.strftime(r"%Y-%m-%d_%H-%M-%S", time.localtime())
print("[%s]%s" % (t, str))
log('start consumer')
# 消费192.168.120.11:9092上的world 这个Topic,指定consumer group是consumer-20171017
consumer = KafkaConsumer('foobar', bootstrap_servers=['localhost:9092'])
for msg in consumer:
recv = "%s:%d:%d: key=%s value=%s" % (msg.topic, msg.partition, msg.offset, msg.key, msg.value)
log(recv)
如果是想读取kafka记得的所有消费记录:
from kafka import KafkaConsumer
import time
# kafka的服务器位置
kafka_servers = '10.10.6.194:9092'
# 统一的topic名称
topicName = 'test'
def log(str):
t = time.strftime(r"%Y-%m-%d_%H-%M-%S", time.localtime())
print("[%s]%s" % (t, str))
log('启动消费者...')
# auto_offset_reset='earliest' 这个参数很重要,如果加上了,就是kafka记录的最后一条位置,如果不加,就是以后要插入的数据了。
#consumer = KafkaConsumer(topicName, auto_offset_reset='earliest', bootstrap_servers=kafka_servers)
consumer = KafkaConsumer(topicName, bootstrap_servers=kafka_servers)
for msg in consumer:
recv = "%s:%d:%d: key=%s value=%s" % (msg.topic, msg.partition, msg.offset, msg.key, msg.value)
log(recv)