Shulou(Shulou.com)06/03 Report--

Mainframe maintenance operation

The mainframe maintenance operations mainly include the offline and online operations of Active NameNode (active name node), Standby NameNode (standby name node), DataNode (data node), ResourceManager (resource manager) and Node Manager (node manager). These operations are aimed at the original node rather than the new node.

1. Active NameNode maintenance operation

1), offline

Execute the. / hadoop-daemon.sh stop namenode command under the $HADOOP_HOME/sbin directory, and the standby node automatically switches over to the primary node and stops the related processes of the primary node. At this time, the primary node has been offline, and the relevant maintenance operations can be carried out on the primary node.

2), go online

After the maintenance is complete, execute the. / hadoop-daemon.sh start namenode command under the $HAIDOOP_HOME/sbin directory to start the namenode process, and the namenode runs in the standby state.

2. Standby NameNode maintenance operation

1), offline

Execute the. / hadoop-daemon.sh stop namenode command under the $HADOOP_HOME/sbin directory. After the backup node is offline, you can perform relevant maintenance operations on the slave node.

2), go online

After the maintenance is completed, execute the. / hadoop-daemon.sh start namenode command under the $HAIDOOP_HOME/sbin directory to start the standby node, while the standby node is still running in the standby state.

3. DataNode maintenance operation

1), offline

Execute the. / hadoop-daemon.sh stop datanode command under the $HADOOP_HOME/sbin directory and wait for the DataNode to be logged off to perform related maintenance operations on the data node.

2), go online

After the maintenance operation is complete, execute the. / hadoop-daemon.sh start DataNode command in the $HADOOP_HOME/sbin directory to start the data node.

4. Active ResourceManager maintenance operation

1), offline

Execute the. / yarn-daemon.sh stop resourcemanager command under the $HADOOP_HOME/sbin directory of the node where the ResourceManager in the Active state is located, which closes the ResourceManager of the current node. Due to the failover feature of ResourceManager, Standby ResourceManager automatically switches to Active ResourceManager. When the Active ResourceManager is offline, you can carry out relevant maintenance operations on the node.

2), go online

After the maintenance is completed, execute the. / yarn-daemon.sh start resourcemanager command in the $HADOOP_HOME/sbin directory to start the resourcemanager process, and the resourcemanager runs in the standby state.

5. Standby ResourceManager maintenance operation

1), offline

Execute the. / yarn-daemon.sh stop resourcemanager command under the $HADOOP_HOME/sbin directory, wait for the backup node to be offline, and perform related maintenance operations on the data node.

2), go online

Execute the. / yarn-daemon.sh start resourcemanager command under the $HADOOP_HOME/sbin directory to start the standby node, and the resourcemanager is running in the standby state.

6. NodeManager maintenance operation

1), offline

Execute the. / yarn-daemon.sh stop nodemanager command under $HADOOP_HOME/sbin, wait for the nodemanage to be offline, and perform related maintenance operations on the nodemanager node.

2), go online

After the maintenance operation is complete, execute. / yarn-daemon.sh start nodemanager in the $HADOOP_HOME/sbin directory to start the NodeManager node.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.