三台hadoop集群，分别是master、slave1和slave2。下面是这三台机器的软件分布：
- master：NameNode、ZK、HiveMetaSotre、HiveServer2、SentryServer
- slave1：DataNode、ZK
- slave2：DataNode、ZK
2 软件需求
- MySql
- mysql-jdbc.jar：mysql-connector-java-5.1.38-bin.jar
- Hadoop
- Hive1.1.0：apache-hive-1.1.0-bin.tar.gz
- Sentry1.6.0：apache-sentry-1.6.0-incubating-bin.tar.gz
注意：这里的Sentry1.6.0版本对应的是Hive1.1.0版本，目前还不支持Hive1.2及以上版本。
3 软件安装
3.1 MySql安装
- 在线安装：sudo apt-get install mysql-server。
- 修改文件：vim /etc/mysql/my.cnf，内容如下：
#bind-address = 127.0.0.1
bind-address = 0.0.0.0
- 重启mysql：service mysql restart
3.2 Hadoop安装
此处省略
3.2 Hive安装
此处省略
创建hive的元数据mysql数据库
create database hive;
CREATE USER hive IDENTIFIED BY 'hive';
GRANT all ON hive.* TO hive@'%' IDENTIFIED BY'hive';
flush privileges;
3.3 Sentry安装
1. 解压安装
tar -zxvf apache-sentry-1.6.0-incubating-bin.tar.gz -C /opt
2. 修改sentry的配置文件(sentry-site.xml)
mv apache-sentry-1.6.0-incubating-bin apache-sentry-1.6.0
cp mysql-connector-java-5.1.38-bin.jar /opt/apache-sentry-1.6.0/lib
cd /opt/apache-sentry-1.6.0/conf
vim sentry-site.xml
3. 修改sentry-sitem.xml文件
<?xml version="1.0" encoding="utf-8"?> <?xml-stylesheet type="text/xsl" href="configuration.xsl"?> <configuration> <property> <name>sentry.service.admin.group</name> <value>admin</value> </property> <property> <name>sentry.service.allow.connect</name> <value>hive,admin</value> </property> <property> <name>sentry.service.reporting</name> <value>JMX</value> </property> <property> <name>sentry.service.server.rpc-address</name> <value>192.168.70.110</value> </property> <property> <name>sentry.service.server.rpc-port</name> <value>8038</value> </property> <property> <name>sentry.store.group.mapping</name> <value>org.apache.sentry.provider.common.HadoopGroupMappingService</value> </property> <property> <name>sentry.hive.server</name> <value>server1</value> </property> <!-- 配置Webserver --> <property> <name>sentry.service.web.enable</name> <value>true</value> </property> <property> <name>sentry.service.web.port</name> <value>51000</value> </property> <property> <name>sentry.service.web.authentication.type</name> <value>NONE</value> </property> <property> <name>sentry.service.web.authentication.kerberos.principal</name> <value> </value> </property> <property> <name>sentry.service.web.authentication.kerberos.keytab</name> <value> </value> </property> <!--配置认证--> <property> <name>sentry.service.security.mode</name> <value>none</value> </property> <property> <name>sentry.service.server.principal</name> <value> </value> </property> <property> <name>sentry.service.server.keytab</name> <value> </value> </property> <!-- 配置Jdbc --> <property> <name>sentry.store.jdbc.url</name> <value>jdbc:mysql://192.168.70.110:3306/sentry</value> </property> <property> <name>sentry.store.jdbc.driver</name> <value>com.mysql.jdbc.Driver</value> </property> <property> <name>sentry.store.jdbc.user</name> <value>sentry</value> </property> <property> <name>sentry.store.jdbc.password</name> <value>sentry</value> </property> </configuration>
4. 配置环境变量
编辑vim /etc/profile
export SENTRY_HOME=/opt/apache-sentry-1.6.0
export PATH=${SENTRY_HOME}/bin:$PATH
使生效source /etc/profile
5. 在mysql中创建sentry数据库
create database sentry;
CREATE USER sentry IDENTIFIED BY 'sentry';
GRANT all ON sentry.* TO sentry@'%' IDENTIFIED BY 'sentry';
flush privileges;
6. 将mysql-jdbc的jar包拷贝到sentry/lib目录下
cp mysql-connector-java-5.1.38-bin.jar ${SENTRY_HOME}/lib
7. 替换hadoop中jline的jar包
rm ${HADOOP_HOME}/share/hadoop/yarn/lib/jline-0.9.94.jar
cp ${SENTRY_HOME}/lib/jline-2.12.jar ${HADOOP_HOME}/share/hadoop/yarn/lib
8. 初始化sentry数据库
sentry --command schema-tool --conffile ${SENTRY_HOME}/conf/sentry-site.xml --dbType mysql --initSchema
9. 启动sentry服务
sentry --command service --conffile ${SENTRY_HOME}/conf/sentry-site.xml
4 Hive集成Sentry
1. 修改hive-site.xml配置文件
<?xml version="1.0" encoding="utf-8"?> <?xml-stylesheet type="text/xsl" href="configuration.xsl"?> <configuration> <property> <name>hive.sentry.conf.url</name> <value>file:///opt/apache-hive-1.1.0/conf/sentry-site.xml</value> </property> <!-- 配置开关控制列访问权限 --> <property> <name>hive.stats.collect.scancols</name> <value>true</value> </property> <!-- Hive Metastore集成Sentry --> <property> <name>hive.metastore.pre.event.listeners</name> <value>org.apache.sentry.binding.metastore.MetastoreAuthzBinding</value> </property> <property> <name>hive.metastore.event.listeners</name> <value>org.apache.sentry.binding.metastore.SentryMetastorePostEventListener</value> </property> <!-- Hive Server2集成Sentry --> <property> <name>hive.server2.session.hook</name> <value>org.apache.sentry.binding.hive.HiveAuthzBindingSessionHook</value> </property> <property> <name>hive.security.authorization.task.factory</name> <value>org.apache.sentry.binding.hive.SentryHiveAuthorizationTaskFactoryImpl</value> </property> <!-- 设置mysql-jdbc --> <property> <name>javax.jdo.option.ConnectionURL</name> <value>jdbc:mysql://192.168.70.110:3306/hive?createDatabaseIfNotExist=true</value> </property> <property> <name>javax.jdo.option.ConnectionDriverName</name> <value>com.mysql.jdbc.Driver</value> </property> <property> <name>javax.jdo.option.ConnectionUserName</name> <value>hive</value> </property> <property> <name>javax.jdo.option.ConnectionPassword</name> <value>hive</value> </property> </configuration>
2. 修改sentry-site.xml配置文件
注意：该文件是在hive/conf目录下的，不要和sentry/conf目录下的sentry-site.xml文件混淆！
<?xml version="1.0" encoding="utf-8"?> <?xml-stylesheet type="text/xsl" href="configuration.xsl"?> <configuration> <property> <name>sentry.service.client.server.rpc-address</name> <value>192.168.70.110</value> </property> <property> <name>sentry.service.client.server.rpc-port</name> <value>8038</value> </property> <property> <name>sentry.service.client.server.rpc-connection-timeout</name> <value>200000</value> </property> <!--配置认证--> <property> <name>sentry.service.security.mode</name> <value>none</value> </property> <property> <name>sentry.service.server.principal</name> <value> </value> </property> <property> <name>sentry.service.server.keytab</name> <value> </value> </property> <property> <name>sentry.provider</name> <value>org.apache.sentry.provider.file.HadoopGroupResourceAuthorizationProvider</value> </property> <property> <name>sentry.hive.provider.backend</name> <value>org.apache.sentry.provider.db.SimpleDBProviderBackend</value> </property> <property> <name>sentry.metastore.service.users</name> <value>hive</value> <!--queries made by hive user (beeline) skip meta store check--> </property> <property> <name>sentry.hive.server</name> <value>server1</value> </property> <property> <name>sentry.hive.testing.mode</name> <value>true</value> </property> </configuration>
3. 拷贝sentry/lib下的jar包到hive/lib目录下
cp ${SENTRY_HOME}/lib/sentry*.jar ${HIVE_HOME}/lib
cp ${SENTRY_HOME}/lib/shiro-*.jar ${HIVE_HOME}/lib
注意：shiro-*.jar一定也要拷贝过去，否则会报错。
4. 修改Hive在HDFS上的文件权限
hadoop fs -chown -R hive:hive /user/hive/warehouse
hadoop fs -chmod -R 770 /user/hive/warehouse
由于HDFS上文件是hive：hive权限，所以linux用户上能
5. 启动Hive Metastore服务
hive --service metastore -hiveconf hive.root.logger=DEBUG,console
6. 启动Hive Server2服务
hive --service hiveserver2 -hiveconf hive.root.logger=DEBUG,console
5 Hive权限验证
5.1 准备测试数据
- 创建测试数据文件：vim /tmp/events.csv，内容如下：
10.1.2.3,US,android,createNote
10.200.88.99,FR,windows,updateNote
10.1.2.3,US,android,updateNote
10.200.88.77,FR,ios,createNote
10.1.4.5,US,windows,updateTag
- 然后，在hive中运行下面sql语句（注意：以hive用户启动hive命令）：
- create database sensitive;
- create table sensitive.events (ip STRING, country STRING, client STRING, action STRING) ROW FORMAT DELIMITED FIELDS TERMINATED BY ',';
- load data local inpath '/tmp/events.csv' overwrite into table sensitive.events;
- create database filtered;
- create view filtered.events as select country, client, action from sensitive.events;
- create view filtered.events_usonly as select * from filtered.events where country = 'US';
- 在master上通过 beeline 连接 hiveserver2，运行下面命令创建角色和组：
beeline -u "jdbc:hive2://master:10000/" -n admin -p admin -d org.apache.hive.jdbc.HiveDriver
- 在beelin中执行下面的 sql 语句创建 role、group等：
- create role admin_role;
- GRANT ALL ON SERVER server1 TO ROLE admin_role;
- GRANT ROLE admin_role TO GROUP admin;
- create role test_role;
- GRANT ALL ON DATABASE filtered TO ROLE test_role;
- use sensitive;
- GRANT SELECT(ip) on TABLE sensitive.events TO ROLE test_role;
- GRANT ROLE test_role TO GROUP test;
上面创建了两个角色：
- admin_role，具有管理员权限，可以读写所有数据库，并授权给 admin 和 hive 组（对应操作系统上的组）
- test_role，只能读写 filtered 数据库，和只能读取sensitive库中events 表中的ip字段，并授权给 test 组。
因为系统上没有test用户和组，所以需要手动创建：useradd test
5.2 测试
5.2.1 测试admin_role角色
使用admin用户访问beelin：
beeline -u "jdbc:hive2://master:10000/" -n admin -p admin -d org.apache.hive.jdbc.HiveDriver
admin用户属于admin_role组，具有管理员权限，可以查看所有角色：
0: jdbc:hive2://master:10000/> show roles;
+-------------+--+
| role |
+-------------+--+
| admin_role |
| test_role |
+-------------+--+
2 rows selected (0.251 seconds)
查看所有权限：
0: jdbc:hive2://master:10000/> show grant role admin_role;
+-----------+--------+------------+---------+-----------------+-----------------+------------+---------------+-------------------+----------+--+
| database | table | partition | column | principal_name | principal_type | privilege | grant_option | grant_time | grantor |
+-----------+--------+------------+---------+-----------------+-----------------+------------+---------------+-------------------+----------+--+
| * | | | | admin_role | ROLE | * | false | 1461507543582000 | -- |
+-----------+--------+------------+---------+-----------------+-----------------+------------+---------------+-------------------+----------+--+
1 row selected (0.111 seconds)
0: jdbc:hive2://master:10000/> show grant role test_role;
+------------+---------+------------+---------+-----------------+-----------------+------------+---------------+-------------------+----------+--+
| database | table | partition | column | principal_name | principal_type | privilege | grant_option | grant_time | grantor |
+------------+---------+------------+---------+-----------------+-----------------+------------+---------------+-------------------+----------+--+
| sensitive | events | | ip | test_role | ROLE | select | false | 1461558337008000 | -- |
| filtered | | | | test_role | ROLE | * | false | 1461557354579000 | -- |
+------------+---------+------------+---------+-----------------+-----------------+------------+---------------+-------------------+----------+--+
2 rows selected (0.111 seconds)
hive用户可以查看所有数据库、访问所有表：
0: jdbc:hive2://master:10000/> show databases;
+----------------+--+
| database_name |
+----------------+--+
| default |
| filtered |
| sensitive |
+----------------+--+
3 rows selected (0.142 seconds)
0: jdbc:hive2://master:10000/> use filtered;
No rows affected (0.144 seconds)
0: jdbc:hive2://master:10000/> select * from filtered.events;
+-----------------+----------------+----------------+--+
| events.country | events.client | events.action |
+-----------------+----------------+----------------+--+
| US | android | createNote |
| FR | windows | updateNote |
| US | android | updateNote |
| FR | ios | createNote |
| US | windows | updateTag |
+-----------------+----------------+----------------+--+
5 rows selected (0.458 seconds)
0: jdbc:hive2://master:10000/> select * from sensitive.events;
+---------------+-----------------+----------------+----------------+--+
| events.ip | events.country | events.client | events.action |
+---------------+-----------------+----------------+----------------+--+
| 10.1.2.3 | US | android | createNote |
| 10.200.88.99 | FR | windows | updateNote |
| 10.1.2.3 | US | android | updateNote |
| 10.200.88.77 | FR | ios | createNote |
| 10.1.4.5 | US | windows | updateTag |
+---------------+-----------------+----------------+----------------+--+
5 rows selected (0.731 seconds)
5.2.2 测试test_role角色
使用test用户访问beeline：
beeline -u "jdbc:hive2://master:10000/" -n test -p test -d org.apache.hive.jdbc.HiveDriver
test用户不是管理员，是不能查看所有角色的：
0: jdbc:hive2://master:10000/> show roles;
ERROR : Error processing Sentry command: Access denied to test.Please grant admin privilege to test.
Error: Error while processing statement: FAILED: Execution Error, return code 1 from org.apache.hadoop.hive.ql.exec.SentryGrantRevokeTask. SentryAccessDeniedException: Access denied to test (state=08S01,code=1)
test用户可以列出所有数据库：
0: jdbc:hive2://master:10000/> show databases;
+----------------+--+
| database_name |
+----------------+--+
| default |
| filtered |
| sensitive |
+----------------+--+
test用户可以filtered库：
0: jdbc:hive2://master:10000/> use filtered;
No rows affected (0.256 seconds)
0: jdbc:hive2://master:10000/> select * from events;
+-----------------+----------------+----------------+--+
| events.country | events.client | events.action |
+-----------------+----------------+----------------+--+
| US | android | createNote |
| FR | windows | updateNote |
| US | android | updateNote |
| FR | ios | createNote |
| US | windows | updateTag |
+-----------------+----------------+----------------+--+
5 rows selected (0.199 seconds)
test用户只能查看sensitive库中events表中的ip字段，其他的都都无法查看：
0: jdbc:hive2://master:10000/> use sensitive;
No rows affected (0.112 seconds)
0: jdbc:hive2://master:10000/> select * from events;
Error: Error while compiling statement: FAILED: SemanticException No valid privileges
User test does not have privileges for QUERY (state=42000,code=40000)
0: jdbc:hive2://master:10000/> select ip from events;
+---------------+--+
| ip |
+---------------+--+
| 10.1.2.3 |
| 10.200.88.99 |
| 10.1.2.3 |
| 10.200.88.77 |
| 10.1.4.5 |
+---------------+--+
5 rows selected (0.176 seconds)
原文出处：http://www.shsxt.com/it/Big-data/825.html
上海尚学堂大数据教研组创作