Spark对接ranger

耀灵2年前 (2023-06-02)技术文章2074

1、包如图所示

https://dtstack-download.oss-cn-hangzhou.aliyuncs.com/insight/insight-4em/release/hadoop/spark/2.4.8_ranger2.2/Spark_2.4.8-dt_centos7_x86_64.tar

2、修改配置文件

[root@cdh01 conf]# vim ranger-spark-security.xml 
<configuration>

    <property>
        <name>ranger.plugin.spark.policy.rest.url</name>
        <value>http://172.16.106.151:6080</value>
    </property>

    <property>
        <name>ranger.plugin.spark.service.name</name>
        <value>hive_pro</value>   //这里定义名称要跟ranger页面对接服务名称要对应
    </property>

    <property>
        <name>ranger.plugin.spark.policy.cache.dir</name>
        <value>/etc/ranger/spark/policycache</value>
    </property>

    <property>
        <name>ranger.plugin.spark.policy.pollIntervalMs</name>
        <value>5000</value>
    </property>

    <property>
        <name>ranger.plugin.spark.policy.source.impl</name>
        <value>org.apache.ranger.admin.client.RangerAdminRESTClient</value>
    </property>

</configuration>

ranger-spark-audit.xml

[root@cdh01 conf]# vim ranger-spark-audit.xml 
<configuration>

    <property>
        <name>xasecure.audit.is.enabled</name>
        <value>true</value>
    </property>

    <property>
        <name>xasecure.audit.destination.db</name>
        <value>false</value>
    </property>

    <property>
        <name>xasecure.audit.destination.db.jdbc.driver</name>
        <value>com.mysql.jdbc.Driver</value>
    </property>

    <property>
        <name>xasecure.audit.destination.db.jdbc.url</name>
        <value>jdbc:mysql://172.16.106.151:3306/ranger</value>
    </property>

    <property>
        <name>xasecure.audit.destination.db.password</name>
        <value>123456</value>
    </property>

    <property>
        <name>xasecure.audit.destination.db.user</name>
        <value>ranger</value>
    </property>

</configuration>

spark-defaults.conf

[root@cdh01 conf]# vim spark-defaults.conf
spark.eventLog.enabled           false
spark.scheduler.mode             FAIR
spark.scheduler.allocation.file  /opt/dtstack/spark/spark_pkg/conf/fairscheduler.xml
#spark.shuffle.service.enabled true
export SPARK_LOCAL_DIRS=/data/spark_tmp/data



# 事件日志
spark.eventLog.enabled=true
spark.eventLog.compress=true

# 保存在hdfs上
spark.eventLog.dir=hdfs://nameservice1/tmp/spark-yarn-logs
spark.history.fs.logDirectory=hdfs://nameservice1/tmp/spark-yarn-logs
#spark.yarn.historyServer.address={{.historyserver_ip}}:18080

# 保存在本地
# spark.eventLog.dir=file://usr/local/hadoop-2.7.3/logs/
# spark.history.fs.logDirectory=file://usr/local/hadoop-2.7.3/logs/

#开启日志定时清除
spark.history.fs.cleaner.enabled=true
#日志有效时间
spark.history.fs.cleaner.maxAge={{.history_cleaner_maxAge}}
#日志检查时间
spark.history.fs.cleaner.interval={{.history_cleaner_interval}}
# 在配置文件底部添加下面的配置：
# # hive metastore的版本设置为 2.1.1
spark.sql.hive.metastore.version=2.1.1
#
# # 引用 hive2.1.1 相关的jar包
spark.sql.hive.metastore.jars=/opt/cloudera/parcels/CDH/lib/hive/lib/*
##开启ranger权限校验
spark.sql.extensions=org.apache.spark.ranger.security.api.RangerSparkSQLExtension