linux7.9下面Hadoop3.1.3僞分布式安裝詳細操作

我用的是VM虛拟機，作業系統是RedHat7.9的系統進行JDK和Hadoop的安裝實施，

本過程隻是作測試和學習參考用。

1、添加hadoop新使用者

useradd -m hadoop -s /bin/bash # 添加hadoop使用者

passwd hadoop # 配置hadoop使用者的密碼

vi /etc/sudoers #編輯配置檔案 在root後一行加入 hadoop ALL=(ALL) ALL ,為hadoop添加管理者權限

2、配置免密登入(root)使用者操作

ssh-keygen -t rsa # 會有提示，都按回車就OK

cat ~/.ssh/id_rsa.pub >> ~/.ssh/authorized_keys # 加入授權

chmod 0600 ~/.ssh/authorized_keys #添權重限

配置完成後，執行ssh hadoop(主機名）指令可以不用輸入密碼即可登入為配置成功。

如查發現生成ssh-keygen報錯，這是沒有安裝openssh造成的，用yum安裝即可。

[root@hadoop dfs]# rpm -qa |grep openssh
openssh-server-7.4p1-21.el7.x86_64
openssh-clients-7.4p1-21.el7.x86_64
openssh-7.4p1-21.el7.x86_64           

3、配置JDK環境

下載下傳JDK的安裝包之後，将jdk安裝到/usr/local/jdk 這個目錄。

tar -zxvf /home/hadoop/download/jdk-8u212-linux-x64.tar.gz -C /usr/local/jdk/

添加環境變量

vi /etc/profile # 打開環境變量配置檔案，添加下面的配置

# java環境變量配置

export JAVA_HOME=/usr/local/jdk

export PATH=$JAVA_HOME/bin:$JAVA_HOME/jre/bin:$PATH

# 配置完成後 執行下面指令是配置生效

source /etc/profile

順便把HADOOP_HOME的環境變量也一起添加了

# hadoop環境變量配置

export HADOOP_HOME=/usr/local/hadoop

export PATH=$HADOOP_HOME/bin:$JAVA_HOME/sbin:$PATH

[root@hadoop download]# cat /etc/profile
unset i
unset -f pathmunge
export JAVA_HOME=/usr/local/jdk/jdk1.8.0_212
export PATH=$JAVA_HOME/bin:$PATH
export HADOOP_HOME=/usr/local/hadoop/hadoop-3.1.3
export PATH=$HADOOP_HOME/bin:$PATH
export PATH=$HADOOP_HOME/sbin:$PATH           

java環境是否配置成功，我們執行java -version 可以看到java相關的資訊

[root@hadoop download]# java -version
java version "1.8.0_212"
Java(TM) SE Runtime Environment (build 1.8.0_212-b10)
Java HotSpot(TM) 64-Bit Server VM (build 25.212-b10, mixed mode)           

4、安裝配置Hadoop

Hadoop NameNode格式化及運作測試，接下來對hadoop進行一些配置，使其能以僞分布式的方式運作。

進入到hadoop的配置檔案所在的目錄

cd /usr/local/hadoop/etc/hadoop           

1. 配置hadoop-env.sh

在該檔案内配置JAVA_HOME 所示:

vi /usr/local/hadoop/hadoop-3.1.3/etc/hadoop/hadoop-env.sh
###
# Generic settings for HADOOP
###

# Technically, the only required environment variable is JAVA_HOME.
# All others are optional.  However, the defaults are probably not
# preferred.  Many sites configure these options outside of Hadoop,
# such as in /etc/profile.d

# The java implementation to use. By default, this environment
# variable is REQUIRED on ALL platforms except OS X!
 export JAVA_HOME=/usr/local/jdk/jdk1.8.0_212

# Location of Hadoop.  By default, Hadoop will attempt to determine
# this location based upon its execution path.
 export HADOOP_HOME=/usr/local/hadoop/hadoop-3.1.3           

JAVA_HOME設定為我們自己的jdk安裝路徑即可

1、配置hdfs-site.xml

hdfs-site.xml的内容改成下面的配置。

<configuration>
<property>
    <name>dfs.replication</name>
    <value>1</value>
</property>
<property>
    <name>dfs.namenode.name.dir</name>
    <value>file:/home/hadoop_data/dfs/name</value>
</property>
<property>
    <name>dfs.datanode.data.dir</name>
    <value>file:/home/hadoop_data/dfs/data</value>
</property>
</configuration>           

dfs.replication # 為檔案儲存副本的數量
dfs.namenode.name.dir # 為hadoop namenode資料目錄，改成自己需要的目錄（不存在需建立）
dfs.datanode.data.dir # 為hadoop datanode資料目錄，改成自己需要的目錄（不存在需建立）           

1、配置core-site.xml

core-site.xml的内容改成下面的配置。

<configuration>
<property>
    <name>hadoop.tmp.dir</name>
    <value>file:/home/hadoop_data</value>
</property>
<property>
    <name>fs.defaultFS</name>
    <value>hdfs://hadoop:9000</value>
</property>
</configuration>           

hadoop.tmp.dir # hadoop 緩存目錄，更改為自己的目錄（不存在需建立）
fs.defaultFS # hadoop fs 監聽端口配置
mkdir /home/hadoop_data/dfs/name
mkdir /home/hadoop_data/dfs/data
cd /home/hadoop_data/ 
 chown -R hadoop:hadoop dfs && chmod  -R 777 dfs           

如果隻需要HDFS，配置就完成，如果需要用到Yarn，還需要做yarn相關的配置。

1、配置mapred-site.xml

mapred-site.xml的内容改成下面的配置。

<configuration>
<property>
    <name>mapreduce.framework.name</name>
    <value>yarn</value>
</property>
</configuration>           

1、配置yarn-site.xml

yarn-site.xml的内容改成下面的配置。

<configuration>
<!-- Site specific YARN configuration properties -->
<property>
    <name>yarn.nodemanager.aux-services</name>
    <value>mapreduce_shuffle</value>
</property>
</configuration>           

Hadoop格式化及啟動

現在hadoop基礎配置已經完成了，需要對Hadoop的namenode進行格式化，然後啟動hadoop dfs服務。

1、NameNode格式化

我們跳轉到hadoop的bin目錄，并執行格式化指令

cd /usr/local/hadoop/bin
./hdfs namenode -format           

執行結果如下圖所示，當exit status 為0時，則為格式化成功。

此時我們的hadoop已經格式化成功了，接下來我們去啟動我們hadoop。

進到hadoop下的sbin目錄

cd /usr/local/hadoop/sbin
./start-dfs.sh # 啟動HDFS
./start-yarn.sh # 啟動YARN           

執行./start-dfs.sh 如下圖所示：

[hadoop@hadoop ~]$ stop-dfs.sh
Stopping namenodes on [hadoop]
Stopping datanodes
Stopping secondary namenodes [hadoop]
[hadoop@hadoop ~]$ start-dfs.sh
Starting namenodes on [hadoop]
Starting datanodes
Starting secondary namenodes [hadoop]
[hadoop@hadoop ~]$ jps
48336 Jps
48002 DataNode
48210 SecondaryNameNode
46725 NodeManager
46621 ResourceManager
47886 NameNode
[hadoop@hadoop ~]$            

還可以看日志看是不啟動報錯

[hadoop@hadoop logs]$ ls -rlt
total 2528
-rw-rw-r--. 1 hadoop hadoop      0 Mar 17 23:15 SecurityAuth-hadoop.audit
-rw-rw-r--. 1 hadoop hadoop    690 Mar 18 09:37 hadoop-hadoop-datanode-hadoop.out.5
-rw-rw-r--. 1 hadoop hadoop    690 Mar 18 09:38 hadoop-hadoop-secondarynamenode-hadoop.out.5
-rw-rw-r--. 1 hadoop hadoop    690 Mar 18 10:04 hadoop-hadoop-namenode-hadoop.out.5
-rw-rw-r--. 1 hadoop hadoop    690 Mar 18 10:04 hadoop-hadoop-datanode-hadoop.out.4
-rw-rw-r--. 1 hadoop hadoop   4124 Mar 18 10:05 hadoop-hadoop-secondarynamenode-hadoop.out.4
-rw-rw-r--. 1 hadoop hadoop    690 Mar 18 10:06 hadoop-hadoop-namenode-hadoop.out.4
-rw-rw-r--. 1 hadoop hadoop    690 Mar 18 10:06 hadoop-hadoop-datanode-hadoop.out.3
-rw-rw-r--. 1 hadoop hadoop    690 Mar 18 10:31 hadoop-hadoop-namenode-hadoop.out.3
-rw-rw-r--. 1 hadoop hadoop 121390 Mar 18 10:50 hadoop-hadoop-secondarynamenode-hadoop.out.3
-rw-rw-r--. 1 hadoop hadoop    690 Mar 18 10:51 hadoop-hadoop-namenode-hadoop.out.2
-rw-rw-r--. 1 hadoop hadoop    690 Mar 18 10:52 hadoop-hadoop-datanode-hadoop.out.2
-rw-rw-r--. 1 hadoop hadoop    690 Mar 18 10:52 hadoop-hadoop-secondarynamenode-hadoop.out.2
-rw-rw-r--. 1 hadoop hadoop    690 Mar 18 10:53 hadoop-hadoop-datanode-hadoop.out.1
-rw-rw-r--. 1 hadoop hadoop    690 Mar 18 10:54 hadoop-hadoop-secondarynamenode-hadoop.out.1
-rw-rw-r--. 1 hadoop hadoop   6151 Mar 18 10:55 hadoop-hadoop-namenode-hadoop.out.1
-rw-rw-r--. 1 hadoop hadoop   2215 Mar 18 14:42 hadoop-hadoop-resourcemanager-hadoop.out
-rw-rw-r--. 1 hadoop hadoop   2199 Mar 18 14:43 hadoop-hadoop-nodemanager-hadoop.out
-rw-rw-r--. 1 hadoop hadoop  41972 Mar 18 14:52 hadoop-hadoop-resourcemanager-hadoop.log
-rw-rw-r--. 1 hadoop hadoop  37935 Mar 18 15:42 hadoop-hadoop-nodemanager-hadoop.log
-rw-rw-r--. 1 hadoop hadoop    690 Mar 18 15:48 hadoop-hadoop-namenode-hadoop.out
-rw-rw-r--. 1 hadoop hadoop    690 Mar 18 15:48 hadoop-hadoop-datanode-hadoop.out
-rw-rw-r--. 1 hadoop hadoop    690 Mar 18 15:48 hadoop-hadoop-secondarynamenode-hadoop.out
-rw-rw-r--. 1 hadoop hadoop 970190 Mar 18 15:48 hadoop-hadoop-datanode-hadoop.log
drwxr-xr-x. 2 hadoop hadoop      6 Mar 18 15:48 userlogs
-rw-rw-r--. 1 hadoop hadoop 572169 Mar 18 15:49 hadoop-hadoop-namenode-hadoop.log
-rw-rw-r--. 1 hadoop hadoop 656741 Mar 18 15:49 hadoop-hadoop-secondarynamenode-hadoop.log
[hadoop@hadoop logs]$ pwd
/usr/local/hadoop/hadoop-3.1.3/logs
[hadoop@hadoop logs]$ tail -20f hadoop-hadoop-namenode-hadoop.log
2022-03-18 15:48:39,521 INFO org.apache.hadoop.hdfs.StateChange: BLOCK* registerDatanode: from DatanodeRegistration(192.168.10.248:9866, datanodeUuid=daafd206-fdfe-44cc-a1fc-8ac1279c5cda, infoPort=9864, infoSecurePort=0, ipcPort=9867, storageInfo=lv=-57;cid=CID-f921bab0-7c73-44ef-bc61-ea81d176ec82;nsid=311459717;c=1647534325420) storage daafd206-fdfe-44cc-a1fc-8ac1279c5cda
2022-03-18 15:48:39,523 INFO org.apache.hadoop.net.NetworkTopology: Adding a new node: /default-rack/192.168.10.248:9866
2022-03-18 15:48:39,523 INFO org.apache.hadoop.hdfs.server.blockmanagement.BlockReportLeaseManager: Registered DN daafd206-fdfe-44cc-a1fc-8ac1279c5cda (192.168.10.248:9866).
2022-03-18 15:48:39,889 INFO org.apache.hadoop.hdfs.server.blockmanagement.DatanodeDescriptor: Adding new storage ID DS-6dd0b7dc-1fd0-488c-924b-76de306ac2cb for DN 192.168.10.248:9866
2022-03-18 15:48:40,062 INFO BlockStateChange: BLOCK* processReport 0xf15bae747aec7666: Processing first storage report for DS-6dd0b7dc-1fd0-488c-924b-76de306ac2cb from datanode daafd206-fdfe-44cc-a1fc-8ac1279c5cda
2022-03-18 15:48:40,065 INFO BlockStateChange: BLOCK* processReport 0xf15bae747aec7666: from storage DS-6dd0b7dc-1fd0-488c-924b-76de306ac2cb node DatanodeRegistration(192.168.10.248:9866, datanodeUuid=daafd206-fdfe-44cc-a1fc-8ac1279c5cda, infoPort=9864, infoSecurePort=0, ipcPort=9867, storageInfo=lv=-57;cid=CID-f921bab0-7c73-44ef-bc61-ea81d176ec82;nsid=311459717;c=1647534325420), blocks: 0, hasStaleStorage: false, processing time: 3 msecs, invalidatedBlocks: 0
2022-03-18 15:49:44,972 INFO org.apache.hadoop.hdfs.server.namenode.FSNamesystem: Roll Edit Log from 192.168.10.248
2022-03-18 15:49:44,972 INFO org.apache.hadoop.hdfs.server.namenode.FSEditLog: Rolling edit logs
2022-03-18 15:49:44,973 INFO org.apache.hadoop.hdfs.server.namenode.FSEditLog: Ending log segment 85, 85
2022-03-18 15:49:44,973 INFO org.apache.hadoop.hdfs.server.namenode.FSEditLog: Number of transactions: 2 Total time for transactions(ms): 2 Number of transactions batched in Syncs: 84 Number of syncs: 2 SyncTimes(ms): 130 
2022-03-18 15:49:44,988 INFO org.apache.hadoop.hdfs.server.namenode.FSEditLog: Number of transactions: 2 Total time for transactions(ms): 2 Number of transactions batched in Syncs: 84 Number of syncs: 3 SyncTimes(ms): 144 
2022-03-18 15:49:44,990 INFO org.apache.hadoop.hdfs.server.namenode.FileJournalManager: Finalizing edits file /home/hadoop_data/dfs/name/current/edits_inprogress_0000000000000000085 -> /home/hadoop_data/dfs/name/current/edits_0000000000000000085-0000000000000000086
2022-03-18 15:49:44,992 INFO org.apache.hadoop.hdfs.server.namenode.FSEditLog: Starting log segment at 87
2022-03-18 15:49:45,514 INFO org.apache.hadoop.hdfs.server.namenode.TransferFsImage: Sending fileName: /home/hadoop_data/dfs/name/current/fsimage_0000000000000000083, fileSize: 533. Sent total: 533 bytes. Size of last segment intended to send: -1 bytes.
2022-03-18 15:49:45,641 INFO org.apache.hadoop.hdfs.server.namenode.TransferFsImage: Sending fileName: /home/hadoop_data/dfs/name/current/edits_0000000000000000084-0000000000000000084, fileSize: 1048576. Sent total: 1048576 bytes. Size of last segment intended to send: -1 bytes.
2022-03-18 15:49:45,744 INFO org.apache.hadoop.hdfs.server.namenode.TransferFsImage: Sending fileName: /home/hadoop_data/dfs/name/current/edits_0000000000000000085-0000000000000000086, fileSize: 42. Sent total: 42 bytes. Size of last segment intended to send: -1 bytes.
2022-03-18 15:49:46,668 INFO org.apache.hadoop.hdfs.server.common.Util: Combined time for file download and fsync to all disks took 0.00s. The file download took 0.00s at 0.00 KB/s. Synchronous (fsync) write to disk of /home/hadoop_data/dfs/name/current/fsimage.ckpt_0000000000000000086 took 0.00s.
2022-03-18 15:49:46,668 INFO org.apache.hadoop.hdfs.server.namenode.TransferFsImage: Downloaded file fsimage.ckpt_0000000000000000086 size 533 bytes.
2022-03-18 15:49:46,684 INFO org.apache.hadoop.hdfs.server.namenode.NNStorageRetentionManager: Going to retain 2 images with txid >= 83
2022-03-18 15:49:46,684 INFO org.apache.hadoop.hdfs.server.namenode.NNStorageRetentionManager: Purging old image FSImageFile(file=/home/hadoop_data/dfs/name/current/fsimage_0000000000000000081, cpktTxId=0000000000000000081)           

這就是namenode啟動成功，如查格式化兩次就會出現datanode啟動不成功，這是clusterID兩次不緻造成的，可以進入

-rw-rw-r--. 1 hadoop hadoop 229 Mar 18 15:48 VERSION
drwx------. 4 hadoop hadoop  54 Mar 18 15:48 BP-301391941-192.168.10.248-1647534325420
[root@hadoop current]# cat VERSION 
#Fri Mar 18 15:48:38 CST 2022
storageID=DS-6dd0b7dc-1fd0-488c-924b-76de306ac2cb
clusterID=CID-f921bab0-7c73-44ef-bc61-ea81d176ec82
cTime=0
datanodeUuid=daafd206-fdfe-44cc-a1fc-8ac1279c5cda
storageType=DATA_NODE
layoutVersion=-57
[root@hadoop current]# pwd
/home/hadoop_data/dfs/data/current
修改clusterID和namenode節點的clusterID一樣後，在重新開機hadoop服務           

現在僞分布式hadoop叢集已經部署成功了，如果啟動hadoop的時候遇到了問題，可以檢視對應的log檔案檢視是由什麼問題引起的。

一般的問題如，未設定JAVA_HOME hadoopdata目錄不存在，或者無權限等等。

現要在可以進入hadoop元件hdfs的UI界面:

現要在可以進入hadoop元件yarn的UI界面:

可能安裝過成會遇到各樣的問題，可是檢視日志和搜尋或是去官網站都可以找到解決的答案，我隻是把做的過程記錄出來了，以備後需。