Background

• 本文主要介绍在 Kafka 中如何配置 Kerberos 认证，以及 java 使用 JAAS 来进行 Kerberos 认证连接。
• 本文演示为单机版。

所用软件版本

查看 Kerberos 版本命令：klist -V

软件名称	版本
jdk	1.8.0_202
kafka	2.12-2.2.1
kerberos	1.15.1

1、Kerberos Server 安装

• Kerberos 是一种由 MIT(麻省理工大学)提出的网络身份验证协议，它旨在通过使用密钥加密技术为客户端和服务器应用程序提供强身份验证。
• Kerberos 是一种基于加密 Ticket 的身份认证协议，主要由三个部分组成：Key Distribution Center (即KDC)、Client 和 Service：
  客户端会先访问两次 KDC，然后再访问目标服务，如：HTTP 服务、Zookeeper 服务、Kafka 服务等。

• 在线安装

yum install krb5-server krb5-workstation krb5-libs -y

• 配置主机名映射

在 /etc/hosts文件中新增本机映射（我这里的主机名是monkey）。

127.0.0.1 monkey

• 配置 krb5.conf

根据需要修改 /etc/krb5.conf文件，其中WLF.COM你可以改成你需要的，还有monkey是你的主机映射。

# Configuration snippets may be placed in this directory as well
includedir /etc/krb5.conf.d/

[logging]
 default = FILE:/var/log/krb5libs.log
 kdc = FILE:/var/log/krb5kdc.log
 admin_server = FILE:/var/log/kadmind.log

[libdefaults]
 dns_lookup_realm = false
 ticket_lifetime = 24h
 renew_lifetime = 7d
 forwardable = true
 rdns = false
 pkinit_anchors = FILE:/etc/pki/tls/certs/ca-bundle.crt
 default_realm = WLF.COM
 #default_ccache_name = KEYRING:persistent:%{uid}

[realms]
WLF.COM = {
 kdc = monkey
 admin_server = monkey
}

[domain_realm]
.monkey = WLF.COM
monkey = WLF.COM

• 配置 kdc.conf

• 修改/var/kerberos/krb5kdc/kdc.conf，kdc的专属配置文件。
• Java 使用 aes256-cts 验证方式需要安装额外的 jar 包，所以为了方便不用哈。

[kdcdefaults]
 kdc_ports = 88
 kdc_tcp_ports = 88

[realms]
 WLF.COM = {
  #master_key_type = aes256-cts
  acl_file = /var/kerberos/krb5kdc/kadm5.acl
  dict_file = /usr/share/dict/words
  admin_keytab = /var/kerberos/krb5kdc/kadm5.keytab
  max_file = 24h
  max_renewable_life = 7d
  supported_enctypes = aes128-cts:normal des3-hmac-sha1:normal arcfour-hmac:normal camellia256-cts:normal camellia128-cts:normal des-hmac-sha1:normal des-cbc-md5:normal des-cbc-crc:normal
 }

• 配置 kadm5.acl

• 修改权限相关配置文件 /var/kerberos/krb5kdc/kadm5.acl
• 其中前一个*号是通配符，表示像名为“abc/admin”或“xxx/admin”的人都可以使用此工具（远程或本地）管理kerberos数据库，后一个跟权限有关，*表示所有权限。WLF.COM是上面配置的realm。

*/admin@WLF.COM *

• 初始化KDC数据库

kdb5_util create -r WLF.COM -s
ll -a /var/kerberos/krb5kdc/

• 启动 Kerberos 服务

systemctl start krb5kdc kadmin
systemctl status krb5kdc kadmin

Kerberos 服务机器上可以使用 kadmin.local 来执行各种管理的操作！

2、Kafka 开启 Kerberos 认证

所有Kerberos 相关配置文件（java连接所需），我们都放在 Kafka 的 config/kerberos/目录下的(kerberos 目录需新建)，把krb5.conf也拷贝过去。

• 创建 keytab

cd /opt/kafka_2.12-2.2.1/
cp /etc/krb5.conf config/kerberos/cd
kadmin.local -q "add_principal -randkey kafka-server/monkey@WLF.COM"
kadmin.local -q "add_principal -randkey kafka-client@WLF.COM"
kadmin.local -q "xst -k config/kerberos/kafka-server.keytab kafka-server/monkey@WLF.COM"
kadmin.local -q "xst -k config/kerberos/kafka-client.keytab kafka-client@WLF.COM"

• Kafka 服务端配置

修改config/server.properties配置文件，新增如下内容：

listeners=SASL_PLAINTEXT://monkey:9092
security.inter.broker.protocol=SASL_PLAINTEXT
sasl.mechanism.inter.broker.protocol=GSSAPI
sasl.enabled.mechanisms=GSSAPI
sasl.kerberos.service.name=kafka-server

• KafkaServer 配置

新建config/kerberos/kafka-server-jaas.conf文件，内容如下：

KafkaServer {
   com.sun.security.auth.module.Krb5LoginModule required
   useKeyTab=true
   keyTab="/opt/kafka_2.12-2.2.1/config/kerberos/kafka-server.keytab"
   storeKey=true
   useTicketCache=false
   principal="kafka-server/monkey@WLF.COM";
};

• KafkaClient 配置

KafkaClient {
   com.sun.security.auth.module.Krb5LoginModule required
   useKeyTab=true
   keyTab="/opt/kafka_2.12-2.2.1/config/kerberos/kafka-client.keytab"
   storeKey=true
   useTicketCache=false
   principal="kafka-client@WLF.COM";
};

• 修改bin/kafka-server-start.sh，倒数第二行增加如下配置：

export KAFKA_OPTS="-Dzookeeper.sasl.client=false -Dzookeeper.sasl.client.username=zk-server -Djava.security.krb5.conf=/opt/kafka_2.12-2.2.1/config/kerberos/krb5.conf -Djava.security.auth.login.config=/opt/kafka_2.12-2.2.1/config/kerberos/kafka-server-jaas.conf"

• 修改bin/kafka-topics.sh、kafka-console-producer.sh、bin/kafka-console-consumer.sh ，倒数第二行增加如下配置：

export KAFKA_OPTS="-Djava.security.krb5.conf=/opt/kafka_2.12-2.2.1/config/kerberos/krb5.conf -Djava.security.auth.login.config=/opt/kafka_2.12-2.2.1/config/kerberos/kafka-client-jaas.conf"

• 新建 config/kerberos/client.properties 文件，内容如下：

security.protocol=SASL_PLAINTEXT
sasl.mechanism=GSSAPI
sasl.kerberos.service.name=kafka-server

• 附赠kafka操作脚本operate.sh

#/bin/bash

:<<!
【脚本说明】
1、此脚本用于操作kafka：启动、停止、查看运行状态、重启、查看日志、查询所有主题、创建主题、删除主题、订阅或消费主题数据；
2、建议把脚本放在kafka安装目录下；
3、适用单机版。
注意安装kafka修改配置文件：
# IP替换为Kafka所在主机的IP
sed -i '31 a listeners=PLAINTEXT://localhost:9092' config/server.properties
!

# kafka安装目录
KAFKA_HOME=/opt/kafka_2.12-2.2.1
# zookeeper地址
ZK_SERVER=monkey
# kafka地址
KAFKA_SERVER=monkey
# zk启动日志
LOG_ZK=$KAFKA_HOME/logs/zookeeper-run.log
# kafka启动日志
LOG_KAFKA=$KAFKA_HOME/logs/kafka-run.log
# sasl
CONF_SASL=config/kerberos/client.properties


# 操作
operate=$1
# 参数
param=$2

# 进程
pids=`ps -ef | egrep "Kafka|QuorumPeerMain" | egrep -v grep | awk '{print $2}'`

# 提示信息
msg='Please input params [<run>|<kil>|<res>|<sta>|<log> [zk]|<list>|<add> <{topic}>|<del> [{topic}]|<consume> <{topic}>|<produce> <{topic}>]'

# 定制化shell输出
function custom_print(){
    echo -e "\033[5;34m ***** \033[0m"
    echo -e "\033[32m $@ ! \033[0m"
    echo -e "\033[5;34m ***** \033[0m"
}

function run(){
    rm -rf $LOG_ZK $LOG_KAFKA
    # 先启动zk
    echo "start zookeeper ..."
    nohup $KAFKA_HOME/bin/zookeeper-server-start.sh $KAFKA_HOME/config/zookeeper.properties > $LOG_ZK 2>&1 &
    sleep 5
    # 再启动kafka
    nohup $KAFKA_HOME/bin/kafka-server-start.sh $KAFKA_HOME/config/server.properties > $LOG_KAFKA 2>&1 &
}

function stop(){
    if [[ $pids ]]; then
        kill -9 $pids
        msg='Stopped success'
        custom_print $msg
    else
        msg='The service is already down'
        custom_print $msg
    fi
}

function restart(){
    if [[ $pids ]]; then
        kill -9 $pids
    fi
    run
    msg='Restart success'
    custom_print $msg
}

function status(){
    jps | egrep "Kafka|QuorumPeerMain"
    if [[ $pids ]]; then
        # 黄底蓝字
        msg='RUNNING'
        custom_print $msg
    else
        # 蓝底黑字
        echo -e "\033[5;34m ***** \033[0m"
        echo -e "\033[31m STOPPED ! \033[0m"
        echo -e "\033[5;34m ***** \033[0m"
    fi
}

function log(){
    if [[ -e $1 ]]; then
        tail -f $1
    else
        msg='No log has been generated yet'
        custom_print $msg
    fi
}

# 判断输入参数
if [[ $operate = "run" || $operate = "start" ]]; then
    if [[ $pids ]]; then
        msg='The service is already running'
        custom_print $msg
    else
        run
        msg='Start success'
        custom_print $msg
    fi
elif [[ $operate = "kil" || $operate = "stop" ]]; then
    stop
elif [[ $operate = "res" || $operate = "restart" ]]; then
    restart
elif [[ $operate = "sta" || $operate = "status" ]]; then
    status
elif [[ $operate = "log" ]]; then
    if [[ $param = "zk" ]]; then
        log $LOG_ZK
    else
        log $LOG_KAFKA
    fi
elif [[ $operate = "list" ]]; then
    $KAFKA_HOME/bin/kafka-topics.sh --zookeeper $ZK_SERVER:2181 --list
elif [[ $operate = "add" && ! -z $param ]]; then
    $KAFKA_HOME/bin/kafka-topics.sh --create --bootstrap-server $KAFKA_SERVER:9092 --replication-factor 1 --partitions 1 --topic $param
    msg="$param create success"
    custom_print $msg
elif [[ $operate = "del" ]]; then
    if [[ -z $param ]]; then
        topics=`$KAFKA_HOME/bin/kafka-topics.sh --zookeeper $ZK_SERVER:2181 --list`
        for topic in $topics; do
            if [[ $topic != "__consumer_offsets" ]]; then
                $KAFKA_HOME/bin/kafka-topics.sh --zookeeper $ZK_SERVER:2181 --delete --topic $topic> /dev/null
                msg="$topic delete success"
                custom_print $msg
            fi
        done
    else
        $KAFKA_HOME/bin/kafka-topics.sh --zookeeper $ZK_SERVER:2181 --delete --topic $param > /dev/null
        msg="$param delete success"
        custom_print $msg
    fi
elif [[ $operate = "consume" && ! -z $param ]]; then
    if [[ -z $CONF_SASL ]]; then
        $KAFKA_HOME/bin/kafka-console-consumer.sh --bootstrap-server $KAFKA_SERVER:9092 --from-beginning --topic $param
    else
        $KAFKA_HOME/bin/kafka-console-consumer.sh --bootstrap-server $KAFKA_SERVER:9092 --from-beginning --topic $param --consumer.config $CONF_SASL
    fi
elif [[ $operate = "produce" && ! -z $param ]]; then
    if [[ -z $CONF_SASL ]]; then
        $KAFKA_HOME/bin/kafka-console-producer.sh --broker-list $KAFKA_SERVER:9092 --topic $param 
    else
        $KAFKA_HOME/bin/kafka-console-producer.sh --broker-list $KAFKA_SERVER:9092 --topic $param --producer.config $CONF_SASL
    fi
else
    custom_print $msg
fi

3、java 程序连接 Kafka

• 防火墙放行88/udp端口

• kdc服务默认端口是88。

firewall-cmd --zone=public --add-port=88/udp --permanent
firewall-cmd --reload

• 引入maven依赖

<dependency>
    <groupId>org.apache.kafka</groupId>
    <artifactId>kafka-clients</artifactId>
    <version>2.2.1</version>
</dependency>

• 程序示例

注意：需要修改 kafka-client-jaas.conf配置文件中配置的kafka-client.keytab路径！

package com.cloudansys;

import org.apache.kafka.clients.consumer.ConsumerRecord;
import org.apache.kafka.clients.consumer.ConsumerRecords;
import org.apache.kafka.clients.consumer.KafkaConsumer;
import org.apache.kafka.clients.producer.KafkaProducer;
import org.apache.kafka.clients.producer.ProducerRecord;
import org.apache.kafka.clients.producer.RecordMetadata;

import java.time.Duration;
import java.util.Collections;
import java.util.Properties;
import java.util.concurrent.Future;

public class TestKafkaKerberos {

    public static void main(String[] args) {
        // 消费者
        testConsumer();

        // 生产者
        testProducer();
    }

    private static void testConsumer() {
        System.setProperty("java.security.auth.login.config", "F:\\test\\kerberos\\kafka-client-jaas.conf");
        System.setProperty("java.security.krb5.conf", "F:\\test\\kerberos\\krb5.conf");
        Properties props = new Properties();
        props.put("bootstrap.servers", "monkey:9092");
        props.put("group.id", "test_group");
        props.put("enable.auto.commit", "true");
        props.put("auto.commit.interval.ms", "1000");
        props.put("key.deserializer", "org.apache.kafka.common.serialization.StringDeserializer");
        props.put("value.deserializer", "org.apache.kafka.common.serialization.StringDeserializer");
        // sasl
        props.put("sasl.mechanism", "GSSAPI");
        props.put("security.protocol", "SASL_PLAINTEXT");
        props.put("sasl.kerberos.service.name", "kafka-server");
        KafkaConsumer<String, String> consumer = new KafkaConsumer<>(props);

        String topic = "test";
        consumer.subscribe(Collections.singletonList(topic));
        while (true) {
            try {
                ConsumerRecords<String, String> records = consumer.poll(Duration.ofMillis(1000));
                for (ConsumerRecord<String, String> record : records) {
                    System.out.printf("offset = %d, partition = %d, key = %s, value = %s%n",
                            record.offset(), record.partition(), record.key(), record.value());
                }
            } catch (Exception e) {
                e.printStackTrace();
            }
        }
    }

    private static void testProducer() {
        // JAAS配置文件路径和Kerberos配置文件路径
        System.setProperty("java.security.auth.login.config", "F:\\test\\kerberos\\kafka-client-jaas.conf");
        System.setProperty("java.security.krb5.conf", "F:\\test\\kerberos\\krb5.conf");
        // kafka属性配置
        Properties props = new Properties();
        props.put("bootstrap.servers", "monkey:9092");
        props.put("key.serializer", "org.apache.kafka.common.serialization.StringSerializer");
        props.put("value.serializer", "org.apache.kafka.common.serialization.StringSerializer");
        // kerberos安全认证
        props.put("security.protocol", "SASL_PLAINTEXT");
        props.put("sasl.mechanism", "GSSAPI");
        props.put("sasl.kerberos.service.name", "kafka-server");

        String topic = "test";
        String msg = "this is a test msg";
        KafkaProducer<String, String> kafkaProducer = new KafkaProducer<>(props);
        ProducerRecord<String, String> record = new ProducerRecord<>(topic, msg);
        // 发送消息记录
        Future<RecordMetadata> future = kafkaProducer.send(record);
        try {
            RecordMetadata metadata = future.get();
            System.out.printf("Message sent to Kafka topic=%s, partition=%d, offset=%d\n", metadata.topic(), metadata.partition(), metadata.offset());
        } catch (Exception e) {
            e.printStackTrace();
        }
        kafkaProducer.close();
    }
}