#1.測試類
想執行hadoop的測試類,我們必須先編譯hadoop
mac下編譯hadoop-2.7.4
然後檢視測試類
org.apache.hadoop.hdfs.server.namenode.TestINodeAttributeProvider
- 1
然後直接執行testDelegationToProvider這個測試類方法,就可以執行,並且在
@Override
public void checkPermission(String fsOwner, String supergroup,
UserGroupInformation ugi, INodeAttributes[] inodeAttrs,
INode[] inodes, byte[][] pathByNameArr, int snapshotId, String path,
int ancestorIndex, boolean doCheckOwner, FsAction ancestorAccess,
FsAction parentAccess, FsAction access, FsAction subAccess,
boolean ignoreEmptyDir) throws AccessControlException {
System.out.println("=========="+fsOwner);
System.out.println("=========="+supergroup);
System.out.println("=========="+ugi);
System.out.println("=========="+inodeAttrs);
System.out.println("=========="+inodes);
System.out.println("=========="+pathByNameArr);
System.out.println("=========="+snapshotId);
System.out.println("=========="+path);
System.out.println("=========="+ancestorIndex);
System.out.println("=========="+doCheckOwner);
System.out.println("=========="+ancestorAccess);
System.out.println("=========="+parentAccess);
System.out.println("=========="+access);
System.out.println("=========="+subAccess);
System.out.println("=========="+ignoreEmptyDir);
}
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10
- 11
- 12
- 13
- 14
- 15
- 16
- 17
- 18
- 19
- 20
- 21
- 22
- 23
- 24
- 25
方法中列印出我們需要的內容
/**
fsOwner = lcc
supergroup = supergroup
ugi = u1 (auth:SIMPLE)
inodeAttrs = [Lorg.apache.hadoop.hdfs.server.namenode.INodeAttributes;@6a93a149
inodes = [Lorg.apache.hadoop.hdfs.server.namenode.INode;@31d3da26
pathByNameArr = [[B@35308ff8
snapshotId = 2147483646
path = /tmp/foo
ancestorIndex = 1
doCheckOwner = false
ancestorAccess = null
parentAccess = null
access = READ_EXECUTE
subAccess = null
ignoreEmptyDir = false
*/
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10
- 11
- 12
- 13
- 14
- 15
- 16
- 17
但是注意看方法
@Before
public void setUp() throws IOException {
CALLED.clear();
Configuration conf = new HdfsConfiguration();
String name = MegrezHdfsAuthorizer.class.getName();
conf.set(DFSConfigKeys.DFS_NAMENODE_INODE_ATTRIBUTES_PROVIDER_KEY,
MyAuthorizationProvider.class.getName());
// name);
conf.setBoolean(DFSConfigKeys.DFS_NAMENODE_ACLS_ENABLED_KEY, true);
EditLogFileOutputStream.setShouldSkipFsyncForTesting(true);
miniDFS = new MiniDFSCluster.Builder(conf).build();
}
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10
- 11
- 12
這個方法設定了兩個重要的屬性
public static final String DFS_NAMENODE_INODE_ATTRIBUTES_PROVIDER_KEY = "dfs.namenode.inode.attributes.provider.class";
public static final String DFS_NAMENODE_ACLS_ENABLED_KEY = "dfs.namenode.acls.enabled";
- 1
- 2
2.不修改原始碼,實現自定義許可權
新建一個maven專案,在/Users/lcc/IdeaProjects/hadoop-hdfs/src/main/java/
目錄下新建一個包package org.apache.hadoop.hdfs.server.namenode;
(這個包位置很重要,一定要是和org.apache.hadoop.hdfs.server.namenode.INodeAttributeProvider
這個類一樣的目錄),然後新建類MyHdfsAuthorizer.java
書寫程式碼
package org.apache.hadoop.hdfs.server.namenode;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import org.apache.hadoop.fs.permission.FsAction;
import org.apache.hadoop.security.UserGroupInformation;
import java.security.AccessControlException;
import java.util.concurrent.ConcurrentHashMap;
import java.util.concurrent.ConcurrentMap;
/**
* HDFS 鑑權具體的實現
*
* @author hulb
* @date 2018/7/11 下午3:17
* <p>
* cp hdfs-auth/target/hdfs-auth-spark2.0.jar /Users/hulb/opt/third/hadoop/share/hadoop/hdfs/lib/
* <p>
* 實現思路:
* 還需要加一個許可權列表的快取。
*/
public class MyHdfsAuthorizer extends INodeAttributeProvider {
private static final Log LOG = LogFactory.getLog(MegrezHdfsAuthorizer.class);
@Override
public void start() {
System.out.println("MegrezHdfsAuthorizer 啟動");
}
@Override
public void stop() {
System.out.println("MegrezHdfsAuthorizer 停止");
}
/**
* 必須重寫的方法
* @param deafultEnforcer
* @return
*/
@Override
public AccessControlEnforcer getExternalAccessControlEnforcer(
AccessControlEnforcer deafultEnforcer) {
return new MegrezAccessControlEnforcer();
}
/**
* cc@lcc ~$ hdfs dfs -ls /user/hive /user/lcc
*
* @param pathElements 這裡的fullPath是我們操作Hdfs時候的全路徑
* 比如: hdfs dfs -ls /
* 那麼全路徑是
* /
* /hadoop
* /ranger
* 比如: hdfs dfs -ls /
* 那麼全路徑是
* /hadoop/
* /hadoop/tempdata
* @param inode 路徑:/hadoop/tempdata
* getGroupName : supergroup
* getUserName : lcc
* getAccessTime : 0
* getAclFeature : null
* getFsPermission : rwxr-xr-x
* getFsPermissionShort : 493
* getLocalNameBytes : [B@50af74e1
* getModificationTime : 1530179365668
* getXAttrFeature : null
* isDirectory : true
* toString : tempdata
* @return
*/
@Override
public INodeAttributes getAttributes(String[] pathElements,
INodeAttributes inode) {
// for(int i=0;i<pathElements.length;i++){
// System.out.println(pathElements[i].toString());
// }
return inode;
}
public class MegrezAccessControlEnforcer implements AccessControlEnforcer {
private ConcurrentMap<String, UserAction> allPass =
new ConcurrentHashMap<String, UserAction>();
/**
* 必須重寫的方法
* 改方法會被系統自動呼叫
* /
@Override
public void checkPermission(String fsOwner, String supergroup,
UserGroupInformation ugi, INodeAttributes[] inodeAttrs,
INode[] inodes, byte[][] pathByNameArr, int snapshotId, String path,
然後修改$HADOOP_HOME/etc/hadoop/hdfs-site.xml
<configuration>
<property>
<name>dfs.replication</name>
<value>1</value>
</property>
<property>
<name>dfs.permissions</name>
<value>false</value>
</property>
<property>
<name>www.fengshen157.com/ dfs.permissions</name>
<value>true</value>
</property>
<property>
<name>dfs.permissions.umask-mode</name>
<value>077</value>
</property>
<property>
<name>dfs.permissions.enabled</name>
<value>true</value>
</property>
<property>
<name>dfs.namenode.acls.enabled</name>
<value>true</value>
lcc@lcc ~$ cp ~/IdeaProjects/spark-authorizer/hadoop_hdfs/target/hdfs-auth-spark2.0.jar /Users/lcc/soft/hadoop/hadoop/share/hadoop/hdfs/lib/
- 1
然後重啟叢集
lcc@lcc hadoop$ sbin/start-all.sh
This script is Deprecated. Instead use start-dfs.sh and start-yarn.sh
18/07/16 15:44:59 WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
Starting namenodes on [lcc]
lcc: starting namenode, logging to /Users/lcc/soft/hadoop/hadoop/logs/hadoop-lcc-namenode-lcc.out
localhost: starting datanode, logging to /Users/lcc/soft/hadoop/hadoop/logs/hadoop-lcc-datanode-lcc.out
Starting secondary namenodes [0.0.0.0]
0.0.0.0: starting secondarynamenode, logging to /Users/lcc/soft/hadoop/hadoop/logs/hadoop-lcc-secondarynamenode-lcc.out
18/07/16 15:45:17 WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
starting yarn daemons
starting resourcemanager, logging to /Users/lcc/soft/hadoop/hadoop/logs/yarn-lcc-resourcemanager-lcc.out
localhost: starting nodemanager,www.yongshiyule178.com logging to /Users/lcc/soft/hadoop/hadoop/logs/yarn-lcc-nodemanager-lcc.out
lcc@lcc hadoop$
然後測試
lcc@lcc ~$ hdfs dfs -mkdir /lcc6
18/07/16 14:27:36 WARN util.NativeCodeLoader:www.taohuayuan178.com/ Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
- 1
- 2
然後檢視日誌
lcc@lcc ~$ tail -100 /Users/lcc/soft/hadoop/hadoop/logs/hadoop-lcc-namenode-lcc.out
MegrezHdfsAuthorizer 啟動
呼叫了本地重寫的getAttributes方法
呼叫了本地重寫的getAttributes方法
呼叫了本地重寫的getAttributes方法
呼叫了本地重寫的getAttributes方法
ax memory size (kbytes, -m) unlimited
open files (-n) 256
pipe size (512 bytes, -p) www.leyouzaixan.cn 1
stack size (kbytes, -s) 8192
cpu time (seconds, -t) unlimited
max user processes (-u) 709
virtual memory (kbytes, -v) unlimited
然後他媽的尷尬了,這裡竟然每呼叫到內部類checkPermission方法
@Override
public void checkPermission(String fsOwner, String supergroup,
UserGroupInformation ugi, INodeAttributes[] inodeAttrs,
INode[] inodes, byte[][] pathByNameArr, int snapshotId, String path,
int ancestorIndex, boolean doCheckOwner, FsAction ancestorAccess,
FsAction parentAccess, FsAction access, FsAction subAccess,
boolean ignoreEmptyDir) throws AccessControlException {
這個問題已經解決,原來是使用命令不會呼叫,實用程式呼叫就可以了
比如
hdfs dfs -mkdir /lcc12
- 1
用程式呼叫就可以了
<dependencies>
<dependency>
<groupId>junit</groupId>
<artifactId>junit</artifactId>
<version>3.8.1</version>
<scope>test</scope>
</dependency>
<dependency>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-hdfs</artifactId>
<version>2.7.4</version>
</dependency>
<dependency>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-client</artifactId>
<version>2.7.4</version>
</dependency>
<dependency>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-common</artifactId>
<version>2.7.4</version>
</dependency>
</dependencies>
package com.lcc.hadoop.test;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.Path;
import java.net.URI;
public class MyMkdir {
public static void main(String[] args)throws Exception{
FileSystem fs = FileSystem.get(new URI("hdfs://lcc:9000"),new Configuration(),"root");
//測試建立一個資料夾,在HDFS上建立一個leitao資料夾,原根目錄下使沒有這個檔案的
boolean flag = fs.mkdirs(new Path("/leitao3"));
這是一個大坑,具體為撒不一樣,可能是命令和FileSystem內部實現是不一樣的