Hadoop install:修订间差异
跳到导航
跳到搜索
无编辑摘要 |
无编辑摘要 |
||
第27行: | 第27行: | ||
====profile==== | ====profile==== | ||
<nowiki>#</nowiki> Java, 20201010, Adam | <nowiki>#</nowiki> Java, 20201010, Adam | ||
export JAVA_HOME=/usr/java/jdk1.8.0_361 | |||
export JAVA_HOME=/usr/java/jdk1.8.0_361 | export PATH=$PATH:$JAVA_HOME/bin | ||
<nowiki>#</nowiki> hadoop, 20201010, Adam | |||
export PATH=$PATH:$JAVA_HOME/bin | export HADOOP_HOME=/opt/hadoop | ||
export PATH=$PATH:$HADOOP_HOME/bin | |||
<nowiki>#</nowiki> hadoop, 20201010, Adam | export PATH=$PATH:$HADOOP_HOME/sbin | ||
export HADOOP_CONF_DIR=${HADOOP_HOME}/etc/hadoop | |||
export HADOOP_HOME=/opt/hadoop | |||
export PATH=$PATH:$HADOOP_HOME/bin | |||
export PATH=$PATH:$HADOOP_HOME/sbin | |||
export HADOOP_CONF_DIR=${HADOOP_HOME}/etc/hadoop | |||
===Hadoop 配置 === | ===Hadoop 配置 === | ||
第48行: | 第41行: | ||
<nowiki>#</nowiki> hadoop是守护线程 读取不到 /etc/profile 里面配置的JAVA_HOME路径 | <nowiki>#</nowiki> hadoop是守护线程 读取不到 /etc/profile 里面配置的JAVA_HOME路径 | ||
<nowiki>#</nowiki> /opt/hadoop/etc/hadoop/ | <nowiki>#</nowiki> /opt/hadoop/etc/hadoop/ | ||
<nowiki>#</nowiki> hadoop-env.sh, mapred-env.sh, yarn-env.sh | |||
<nowiki>#</nowiki> hadoop-env.sh, mapred-env.sh, yarn-env.sh | cp hadoop-env.sh hadoop-env.sh.20210409 | ||
cp mapred-env.sh mapred-env.sh.20210409 | |||
cp hadoop-env.sh hadoop-env.sh.20210409 | cp yarn-env.sh yarn-env.sh.20210409 | ||
cp mapred-env.sh mapred-env.sh.20210409 | echo ' | ||
<nowiki>#</nowiki> hdfs, 20210409, Adam | |||
export JAVA_HOME=/usr/java/jdk1.8.0_361' >> | |||
====Setup==== | ====Setup==== | ||
=====core-site.xml (Common组件)===== | =====core-site.xml (Common组件)===== | ||
<configuration> | <configuration> |
2023年2月17日 (五) 22:32的版本
ENV
USER
groupadd hadoop -g 1001
useradd hdfs -g hadoop -u 1001
Java
ln -s /usr/java/jdk1.8.0_361/jre/bin/java /usr/bin/java
# Java Error
java 以前使用 --version 来查看版本的;java8 以后变更为 -version。
$ java --version Unrecognized option: --version Error: Could not create the Java Virtual Machine. Error: A fatal exception has occurred. Program will exit.
$ java -version java version "1.8.0_361" Java(TM) SE Runtime Environment (build 1.8.0_361-b09) Java HotSpot(TM) 64-Bit Server VM (build 25.361-b09, mixed mode)
hadoop
ln -s /opt/hadoop-3.3.0 /opt/hadoop
profile
# Java, 20201010, Adam export JAVA_HOME=/usr/java/jdk1.8.0_361 export PATH=$PATH:$JAVA_HOME/bin # hadoop, 20201010, Adam export HADOOP_HOME=/opt/hadoop export PATH=$PATH:$HADOOP_HOME/bin export PATH=$PATH:$HADOOP_HOME/sbin export HADOOP_CONF_DIR=${HADOOP_HOME}/etc/hadoop
Hadoop 配置
配置 Hadoop 环境脚本文件中的 JAVA_HOME 参数
# hadoop是守护线程 读取不到 /etc/profile 里面配置的JAVA_HOME路径
# /opt/hadoop/etc/hadoop/ # hadoop-env.sh, mapred-env.sh, yarn-env.sh cp hadoop-env.sh hadoop-env.sh.20210409 cp mapred-env.sh mapred-env.sh.20210409 cp yarn-env.sh yarn-env.sh.20210409 echo ' # hdfs, 20210409, Adam export JAVA_HOME=/usr/java/jdk1.8.0_361' >>
Setup
core-site.xml (Common组件)
<configuration> <property> <name>fs.defaultFS</name> <value>hdfs://g2-hdfs-01:9000</value> </property> <property> <name>io.file.buffer.size</name> <value>131072</value> </property> <property> <name>hadoop.tmp.dir</name> <value>/u01/hdfs/tmp</value> </property> </configuration>
hdfs-site.xml (HDFS组件)
- 修改 dfs.replication 不影响历史文件的备份数。修改历史文件备份数:hadoop fs -setrep -R 3 <path>
<configuration> <property> <name>dfs.namenode.http-address</name> <value>g2-hdfs-01:9870</value> </property> <property> <name>dfs.namenode.secondary.http-address</name> <value>g2-hdfs-02:9870</value> </property> <property> <name>dfs.namenode.name.dir</name> <value>file:/u01/hdfs/dfs/nn</value> </property> <property> <name>dfs.datanode.data.dir</name> <value>file:/u01/hdfs/dfs/dn</value> </property> <property> <name>dfs.webhdfs.enabled</name> <value>true</value> </property> <property> <name>dfs.permissions</name> <value>false</value> </property> </configuration> -- del <property> <name>dfs.replication</name> <value>3</value> </property> <property> <name>dfs.blocksize</name> <value>268435456</value> </property> <property> <name>dfs.namenode.handler.count</name> <value>100</value> </property>
mapred-site.xml
<configuration> <property> <name>mapreduce.framework.name</name> <value>yarn</value> </property> </configuration> -- del <property> <name>mapreduce.jobhistory.address</name> <value>g2-hdfs-01:10020</value> </property> <property> <name>mapreduce.jobhistory.webapp.address</name> <value>g2-hdfs-01:19888</value> </property> <property> <name>mapreduce.application.classpath</name> <value>$HADOOP_MAPRED_HOME/share/hadoop/mapreduce/*:$HADOOP_MAPRED_HOME/share/ hadoop/mapreduce/lib/*</value> </property>
yarn-site.xml
<configuration> <property> <name>yarn.resourcemanager.hostname</name> <value>g2-hdfs-01</value> </property> <property> <name>yarn.nodemanager.aux-services</name> <value>mapreduce_shuffle</value> </property> <property> <name>yarn.resourcemanager.webapp.address</name> <value>g2-hdfs-01:8088</value> </property> <property> <name>yarn.scheduler.maximum-allocation-mb</name> <value>32768</value> </property> <property> <name>yarn.nodemanager.vmem-check-enabled</name> <value>false</value> </property> <property> <name>yarn.nodemanager.env-whitelist</name> <value>JAVA_HOME,HADOOP_COMMON_HOME,HADOOP_HDFS_HOME,HADOOP_CONF_DIR,CLASSPATH_PREPE ND_DISTCACHE,HADOOP_YARN_HOME,HADOOP_MAPRED_HOME</value> </property> </configuration>
<property> <name>yarn.resourcemanager.webapp.address</name> <value>hadoop01/192.168.44.5:8088</value> <description>配置外网只需要替换外网ip为真实ip,否则默认为 localhost:8088</description> </property> yarn.resourcemanager.hostname 指定yarn的ResourceManager管理界面的地址,不配的话,Active Node始终为0 yarn.scheduler.maximum-allocation-mb 每个节点可用内存,单位MB,默认8182MB yarn.nodemanager.aux-services reducer获取数据的方式 yarn.nodemanager.vmem-check-enabled false = 忽略虚拟内存的检查
# workers
g2-hdfs-01
g2-hdfs-02
..
INIT
# chown chown -R hdfs:hadoop /opt/hadoop* # nn /opt/hadoop/bin/hdfs namenode -format
Start
## Start : root(发现需要用 hdfs 停服务,root不可停) /opt/hadoop/sbin/start-dfs.sh /opt/hadoop/sbin/start-yarn.sh /opt/hadoop/sbin/stop-dfs.sh /opt/hadoop/sbin/stop-yarn.sh http://mc0:9870 # hdfs http://mc0:8088 # yarn ## 单节点启停 # /opt/hadoop/bin hdfs --daemon start datanode hdfs --daemon start namenode yarn --daemon start nodemanager yarn --daemon stop resourcemanager