一、配置集群
100
101
102
HDFS
namenode、
datanode
datanode
Secondary namenode(2nn)、datanode
YARN
nodemanager
resourcemanager、nodemanager
nodemanager
二、修改配置:
切换路径:cd /opt/module/hadoop-3.1.3/etc/hadoop/
(1)配置core-site.xml,在cd /opt/module/hadoop-3.1.3/etc/hadoop/路径下,
vi core-site.xml
fs.defaultFS hdfs://hadoop1000:9820 hadoop.data.dir /opt/module/hadoop-3.1.3/data
(2)配置hdfs-site.xml,在cd /opt/module/hadoop-3.1.3/etc/hadoop/路径下,
vi hdfs-site.xml
dfs.namenode.name.dir file://${hadoop.data.dir}/name dfs.datanode.data.dir file://${hadoop.data.dir}/data dfs.namenode.checkpoint.dir file://${hadoop.data.dir}/namesecondary dfs.client.datanode-restart.timeout 30 dfs.namenode.http-address hadoop1000:9870 dfs.namenode.secondary.http-address hadoop1002:9868
(3)配置yarn-site.xml,在cd /opt/module/hadoop-3.1.3/etc/hadoop/路径下,
vi yarn-site.xml
yarn.nodemanager.aux-services mapreduce_shuffle yarn.resourcemanager.hostname hadoop1001 yarn.nodemanager.env-whitelist JAVA_HOME,HADOOP_COMMON_HOME,HADOOP_HDFS_HOME,HADOOP_CONF_DIR,CLASSPATH_PREPEND_DISTCACHE,HADOOP_YARN_HOME,HADOOP_MAPRED_HOME
三、ssh免密登录
免密登录之前 每次远程访问主机都需要输密码:
1、在每个节点生成公钥和私钥,并拷贝
Hadoop1000 生成公钥和私钥:
[root@hadoop1001] ssh-keygen -t rsa
然后敲(三个回车)
将公钥拷贝到要免密登录的目标机器上
[root@hadoop1000] ssh-copy-id hadoop1000 [root@hadoop1000] ssh-copy-id hadoop1001 [root@hadoop1000] ssh-copy-id hadoop1002
Hadoop1001:生成公钥和私钥
[root@hadoop1001] ssh-keygen -t rsa
然后敲(三个回车)
将公钥拷贝到要免密登录的目标机器上
[root@hadoop1001] ssh-copy-id hadoop1000 [root@hadoop1001] ssh-copy-id hadoop1001 [root@hadoop1001] ssh-copy-id hadoop1002
Hadoop1002: 生成公钥和私钥
[root@hadoop1002] ssh-keygen -t rsa
然后敲(三个回车)
将公钥拷贝到要免密登录的目标机器上
[root@hadoop1002] ssh-copy-id hadoop1000 [root@hadoop1002] ssh-copy-id hadoop1001 [root@hadoop1002] ssh-copy-id hadoop1002
远程访问主机 命令:ssh hadoop1001(设置免密登录后,不用输密码直接进入hadoop1001)
登出:exit 路径切换到etc:cd ..
四、复制文件
复制文件到hadoop1001:
scp -r hadoop/ root@hadoop1001:/opt/module/hadoop-3.1.3/etc/
复制文件到hadoop1002:
scp -r hadoop/ root@hadoop1002:/opt/module/hadoop-3.1.3/etc/
五、格式化namenode (namenode启动不了需要删除data文件和logs)
命令:hdfs namenode -format
六、集群单点启动
Hadoop1000:
hdfs --daemon start namenode
hdfs --daemon start datanode
yarn --daemon start nodemanager
hadoop1001:
yarn --daemon start resourcemanager
hdfs --daemon start datanode
yarn --daemon start nodemanager
hadoop1002:
hdfs --daemon start secondarynamenode
hdfs --daemon start datanode
yarn --daemon start nodemanager
欢迎分享,转载请注明来源:内存溢出
评论列表(0条)