Stopping CDH Services Using the Command Line
Important:
- If you use Cloudera Manager, do not use these command-line instructions.
- This information applies specifically to CDH 5.8.x. If you use a lower version of CDH, see the documentation for that version located at Cloudera Documentation.
To shut down all Hadoop Common system services (HDFS, YARN, MRv1), run the following on each host in the cluster:
$ for x in `cd /etc/init.d ; ls hadoop-*` ; do sudo service $x stop ; done
To verify that no Hadoop processes are running, run the following command on each host in the cluster:
# ps -aef | grep java
To stop system services individually, use the instructions in the table below.
Important: Stop services in the order listed in the table. (You can start services in
the reverse order.)
Order | Service | Comments | Instructions |
---|---|---|---|
1 | Hue |
sudo service hue stop |
|
2 | Impala |
sudo service impala-server stop sudo service impala-catalog stop sudo service impala-state-store stop |
|
3 | Oozie |
sudo service oozie stop |
|
4 | Hive | Exit the Hive console and ensure no Hive scripts are running. Stop the Hive server, HCatalog, and metastore daemon on each client. |
sudo service hiveserver2 stop sudo service hive-webhcat-server stop sudo service hive-metastore stop |
5 | Flume 1.x | There is no Flume master. |
sudo service flume-ng-agent stop |
6 | Sqoop 1 |
sudo service sqoop-metastore stop |
|
6 | Sqoop 2 |
sudo service sqoop2-server stop |
|
7 | Lily HBase Indexer (Solr/HBase Indexer) |
sudo service hbase-solr-indexer stop |
|
10 | Spark |
sudo service spark-worker stop sudo service spark-history-server stop sudo service spark-master stop |
|
8 | Sentry | Only present on a secure configuration. |
sudo service sentry-store stop |
9 | Solr Search |
sudo service solr-server stop |
|
10 | HBase | Stop the Thrift server and clients, followed by RegionServers and finally the Master. |
sudo service hbase-thrift stop sudo service hbase-rest stop sudo service hbase-regionserver stop sudo service hbase-master stop |
11 | MapReduce v1 | Stop the JobTracker service, then stop the TaskTracker service on all nodes where it is running. | For MRv1 HA setup:
sudo service hadoop-0.20-mapreduce-jobtrackerha stopFor Non-HA setup: sudo service hadoop-0.20-mapreduce-jobtracker stopFor all types of MRv1 setups: sudo service hadoop-0.20-mapreduce-tasktracker stop |
12 | YARN | Stop the JobHistory server, followed by the ResourceManager and each of the NodeManagers. |
$ sudo service hadoop-mapreduce-historyserver stop $ sudo service hadoop-yarn-resourcemanager stop $ sudo service hadoop-yarn-nodemanager stop |
13 | HDFS |
Stop HttpFS and the NFS Gateway (if present). Stop the Secondary NameNode, then the primary NameNode, followed by Journal nodes (if present) and then each of DataNodes. |
sudo service hadoop-httpfs stop sudo service hadoop-hdfs-nfs3 stop sudo service hadoop-hdfs-secondarynamenode stop sudo service hadoop-hdfs-namenode stop sudo service hadoop-hdfs-journalnode stop sudo service hadoop-hdfs-datanode stop |
14 | KMS (Key Management Server) | Only present if HDFS at rest encryption is enabled |
sudo service hadoop-kms-server stop |
15 | ZooKeeper |
sudo service zookeeper-server stop |
Page generated July 8, 2016.
<< Starting and Stopping HBase Using the Command Line | ©2016 Cloudera, Inc. All rights reserved | Migrating Data between a CDH 4 and CDH 5 Cluster >> |
Terms and Conditions Privacy Policy |